Design and Operation of Farm Irrigation Systems 2nd Edition

Design and Operation of Farm Irrigation Systems 2nd edition

Glenn J. Hoffman Robert G. Evans Marvin E. Jensen Derrel L. Martin Ronald L. Elliott

Copyright © 2007 by the American Society of Agricultural and Biological Engineers. All rights reserved. Manufactured in the United States of America.

This publication may not be reproduced in whole or in part by any means (with the exception of short quotes for the purpose of review) without the permission of the publisher. For information, contact: ASABE, 2950 Niles Rd., St. Joseph, MI 49085-9659 USA. Phone: 269 429 0300 Fax: 269 429 3852 E-mail: [email protected]

Copy editing by Peg McCann Cover design by Melissa Miller Production assistance from Marcia Stults McCavit and Paticia Howard Cover photographs courtesy of USDA Natural Resources Conservation Service and University of Nebraska

ASABE is an educational and scientific organization dedicated to the advancement of engineering applicable to agricultural, food, and biological systems. For more information visit www.asabe.org.

The American Society of Agricultural and Biological Engineers is not responsible for statements and opinions advanced in its meetings or printed in its publications. They represent the views of the individual to whom they are credited and are not binding on the Society as a whole.

International Standard Book Number (ISBN) 1-892769-64-6 Library of Congress Card Number (LCCN) 2007934701 ASABE Order Number 801P1007

PREFACE Irrigated agriculture has played an important role in food production during the past century and will become even more important as the global population continues to increase. The internet has hastened the dissemination of new irrigation technology and water management guidelines developed by specialists. Personal computers have facilitated complex calculations and control of automated irrigation systems. However, there is still a need for irrigation system designers and system operators to have a comprehensive book on farm irrigation systems readily available. This monograph has been widely disseminated since first published in 1980 and reprinted in 1982. This second edition is expected to meet an important need for several future decades. Publication of this monograph culminates a decade of effort. Planning the second edition first started in 1994 when a group met informally at a meeting of the American Society of Agricultural Engineers to discuss the need for an updated version of the first irrigation monograph. Ronald Elliott and Marvin Jensen began the organizational and editorial processes. Leading experts in their respective areas were asked to write various chapters. In 2000, because Elliott’s new assignment restricted his time for this work and Jensen’s time was limited because of his involvement in several water use studies, work on this edition was delayed. In 2002, the SW-24 Group chaired by Robert Evans established a new editorial committee consisting of Glenn Hoffman, Robert Evans, Gary Clark, and Derrel Martin to share the load of this monumental task. The committee met in Denver in November, 2002, to review the status of the revised chapters, to add new chapters, set new target dates, and to make review assignments. Elliott indicated that he did not wish to continue an active role and in August, 2005, Gary Clark assumed a new position and no longer had time to continue as an editor. The remaining work load was redistributed to Evans, Hoffman, Jensen, and Martin. This edition provides the latest technology in the design of surface, sprinkler, and microirrigation systems along with basic information about soils and current information on estimating crop water requirements. New chapters have been added on planning systems, environmental issues, efficiency and uniformity, chemigation, and use of wastewater for irrigation. As is the case with all ASABE monographs, the Society is indebted to many individuals who contributed significantly to the planning, writing, reviewing, editing, and publishing of this book. The sharing of their knowledge, time, and patience in the production of this book is greatly appreciated. All of the editors are extremely grateful for the opportunity to work with so many dedicated and enthusiastic engineers and scientists in completing this project. The assistance of the ASABE staff, especially Peg McCann, in editing and producing this monograph is especially appreciated. The Editors: Glenn J. Hoffman Robert G. Evans Marvin E. Jensen Derrel L. Martin August, 2007 Ronald L. Elliott

THE AUTHORS Richard G. Allen, University of Idaho Research and Extension Center, 3793 North 3600 East, Kimberly, ID 83341 James E. Ayars, USDA-ARS Water Management Research Laboratory, 9611 South Riverbend Ave., Parlier, CA 93648 Evan W. Christen, CSIRO Land and Water, Griffith, New South Wales 2680, Australia Allan W. Clark, Clark Brothers, Inc., 19772 South Elgin, Dos Palos, CA 93620 Albert J. Clemmens, USDA-ARS Arid-Land Agricultural Research Center, 21881 North Cardon Lane, Maricopa, AZ 85238 Allen R. Dedrick, USDA-ARS National Program Staff, Beltsville, MD (retired). Current address: 608 West Villa Rita Dr., Phoenix, AZ 85023 Harold R. Duke, USDA-ARS Water Management Research, Fort Collins, CO (retired). Current address: 1047 Greenfield Court, Fort Collins, CO 80524 Keith O. Eggleston, Water Quality, U.S. Bureau of Reclamation, Denver Federal Center, P.O. Box 25007 (D-5724), Denver, CO 80225 (retired) Dean E. Eisenhauer, Department of Biological Systems Engineering, 232 Chase Hall, University of Nebraska, Lincoln, NE 68583 Abd El-Ghani M. El-Gindy, Agricultural Mechanization Department, Faculty of Agriculture, Ain-Shams University, Cairo, Egypt Ronald L. Elliott, Biosystems and Agricultural Engineering Department, 111 Ag Hall, Oklahoma State University, Stillwater, OK 74078 Robert G. Evans, USDA-ARS Northern Plains Agricultural Research Laboratory, 1500 North Central Avenue, Sidney, MN 59270 Robert O. Evans, Biological and Agricultural Engineering Department, P.O. Box 7625, North Carolina State University, Raleigh, NC 27695 Delmar D. Fangmeier, University of Arizona, Tucson (retired). Current address: 848 West Safari Dr., Tucson, AZ 85704 James L. Fouss, USDA-ARS Soil and Water Research Unit, 4115 Gourrier Ave., Baton Rouge, LA 70808 Ronald J. Gaddis, A B Consulting Co., Inc., Lincoln, NE (retired). Current address: 8100 Sanborn Dr., Lincoln, NE 68505 Leland A. Hardy, H & R Engineering, Inc., 690 Loring Dr. NW, Salem, OR 97304 Dale F. Heermann, USDA-ARS Water Management Research, Natural Resources Research Center, 2150 Centre Ave., Building D, Suite 320, Fort Collins, CO 80526 (retired) Glenn J. Hoffman, Department of Biological Systems Engineering, University of Nebraska, Lincoln, NE (retired). Current address: 9203 N. Crown Ridge, Fountain Hills, AZ 85268 Sagit R. Ibatullin, Water Economy Research Institute, 12 Kolbasshy Koygeldy Str., 480022 Taraz City, Kazakhstan Marvin E. Jensen, USDA-ARS National Program Staff, Fort Collins, CO (retired). Current address: 1207 Springwood Dr., Fort Collins, CO 80525 Dennis C. Kincaid, USDA-ARS, Kimberly, ID (retired). Current address: 3849B North 3700 East, Hansen, ID 83334 Larry G. King, Department Agricultural and Biological Systems Engineering, Washington State University, Pullman, WA (retired). Current address: 19855 East Silver Creek Lane, Queen Creek, AZ 85242

vi

Contents

E. Gordon Kruse, USDA-ARS Water Management Research, Fort Collins, CO (retired). Current address: 4740 Player Dr., Fort Collins, CO 80525 Joseph M. Lord, Jr., JMLord, Inc., 267 North Fulton St., Fresno, CA 93701 William M. Lyle, Texas A&M University, Lubbock, TX (retired). Current address: Box 1679, Hilltop Lakes, TX 77871 Mark Madison, CH2M Hill, 2020 SW 4th Ave., Portland, OR 97201 Derrel L. Martin, Department of Biological Systems Engineering, 243 Chase Hall, University of Nebraska, Lincoln, NE 68583 Anne M. S. McFarland, Texas Institute for Applied Environmental Research, Tarleton State University, 201 St. Felix Street, Stephenville, TX 76401 Marshall J. McFarland, Agricultural Research and Extension Center, Stephenville, TX (retired). Current address: 1025 Darren Drive, Stephenville, TX 76401 Alan W. Moore, Cameron County Drainage District 5, 301 East Pierce St., Harlingen, TX 78550 Luis S. Pereira, Technical University of Lisbon, Lisbon, Portugal. Current address: Instituto Superior de Agronomia, Departmento Engenharia Rural, Tapada da Ajuda, Lisboa Codex 1399, Portugal William O. Pruitt, University of California, Davis, CA (retired). Current address: 804 West 8th St., Davis, CA 95616 John A. Replogle, USDA-ARS Arid-Land Agricultural Research Center, 21881 North Cardon Lane, Maricopa, AZ 85238 Matt A. Sanderson, USDA-ARS Pasture Systems and Watershed Management Research Unit, Building 3702, Curtin Rd., University Park, PA 16802 Joseph Shalhevet, Institute of Soil, Water and Environmental Science, Agricultural Research Organization, Bet Dagan, Israel (retired). Current address: 14 Einstein Street, Rehovot 76470, Israel Allen G. Smajstrala, Department of Agricultural and Biological Engineering, University of Florida, Gainesville, FL (deceased) Roger E. Smith, UDSA-ARS, Fort Collins, CO. Current address: Colorado State University, 2150 Centre Ave., Building D, Fort Collins, CO 80526 Kenneth H. Solomon, BioResource and Agricultural Engineering Department, California Polytechnic State University, San Luis Obispo, CA (retired). Current address: 190 Kodiak St., Morro Bay, CA 93442 Dean D. Steele, Department of Agricultural and Biosystems Engineering, North Dakota State University, 1221 Albrecht Blvd., P.O. Box 5626, Fargo, ND 58105 Theodor S. Strelkoff, USDA-ARS Arid-Land Agricultural Research Center, 21881 North Cardon Lane, Maricopa, AZ 85238 Thomas J. Trout, USDA-ARS Water Management Research, Natural Resources Research Center, 2150 Centre Ave., Building D, Suite 320, Fort Collins, CO 80526 Ted W. van der Gulik, Resource Management Branch, BC Ministry of Agriculture and Lands, 1767 Angus Campbell Rd., Abbotsford, BC V3G 2E5, Canada Wynn R. Walker, College of Engineering, Utah State University, Logan, UT 84322 Arthur W. Warrick, Department of Soils, Water and Environmental Science, Shantz Building, P.O. Box 210038, University of Arizona, Tucson, AZ 85721 Lyman S. Willardson, Utah State University, Logan, UT (deceased) James L. Wright, USDA-ARS Northwest Irrigation and Soils Research Laboratory, Kimberly, ID (retired) I-Pai Wu, Department of Biosystems Engineering, University of Hawaii, Honolulu, HI (retired)

CONTENTS Preface .................................................................................................iii The Authors..........................................................................................v Chapter 1 Introduction

1

1.1 Overview ................................................................................................................. 1 1.2 Worldwide Irrigation Development......................................................................... 2 1.3 Irrigation Development in the United States ........................................................... 7 1.4 Issues Facing Irrigated Agriculture ....................................................................... 12 1.5 Future Directions ................................................................................................... 23 References ................................................................................................................... 30

Chapter 2 Sustainable and Productive Irrigated Agriculture

33

2.1 Introduction ........................................................................................................... 33 2.2 Role of Irrigation in Food and Fiber Production ................................................... 37 2.3 Crop Production and Irrigation Water Requirements ............................................ 43 2.4 System Design and Increasing Competition for Renewable Water Supplies ........ 46 2.5 Irrigation Water Management During Droughts ................................................... 49 2.6 Other Agricultural Purposes and Benefits of Irrigation......................................... 49 2.7 System Design and Operation Challenges............................................................. 50 2.8 Summary ............................................................................................................... 51 References ................................................................................................................... 52

Chapter 3 Planning and System Selection

57

3.1 Introduction ........................................................................................................... 57 3.2 Planning for Irrigation ........................................................................................... 61 3.3 Irrigation System Selection ................................................................................... 66 References ................................................................................................................... 75

Chapter 4 Environmental Considerations

76

4.1 Introduction ........................................................................................................... 76 4.2 Water Storage, Diversion, and Consumption ........................................................ 80 4.3 Groundwater Quality ............................................................................................. 85 4.4 Surface Water Runoff............................................................................................ 93 References ................................................................................................................. 100

Chapter 5 Efficiency and Uniformity

108

5.1 Introduction ......................................................................................................... 108

viii

Contents

5.2 Irrigation Scheme Physical Models ..................................................................... 109 5.3 Irrigation Performance Parameter Definitions..................................................... 111 5.4 Summary ............................................................................................................. 118 References ................................................................................................................. 118

Chapter 6 Soil Water Relationships

120

6.1 Introduction ......................................................................................................... 120 6.2 Water-Holding Characteristics of Soils ............................................................... 120 6.3 Soil Hydraulic Conductivity................................................................................ 135 6.4 Water Movement in Soil ..................................................................................... 139 6.5 Complicating Factors........................................................................................... 150 List of Symbols ......................................................................................................... 154 References ................................................................................................................. 155

Chapter 7 Controlling Salinity

160

7.1 Introduction ......................................................................................................... 160 7.2 Quantifying Salinity Hazards .............................................................................. 164 7.3. Crop Tolerance ................................................................................................... 168 7.4. Leaching ............................................................................................................. 178 7.5 Salinity Impacts on Irrigation Design.................................................................. 187 7.6 Salinity Management Practices ........................................................................... 193 7.7 Summary and Conclusions .................................................................................. 201 References ................................................................................................................. 201

Chapter 8 Water Requirements

208

8.1 Introduction ......................................................................................................... 208 8.2 Definitions ........................................................................................................... 209 8.3 Direct Measurements........................................................................................... 211 8.4 Estimation of Reference ET ................................................................................ 212 8.5 Estimating ET for Crops...................................................................................... 227 8.6 Evapotranspiration Coefficients for Landscapes ................................................. 258 8.7 Estimating Kc from Fraction of Cover ................................................................ 265 8.8 Effect of Irrigation Method on Kc ....................................................................... 266 8.9 Effects of Surface Mulching on Kc ..................................................................... 266 8.10 Precipitation Runoff .......................................................................................... 267 8.11 Other Water Requirements ................................................................................ 270 8.12 Effective Rainfall............................................................................................... 272 8.13 Design Requirements......................................................................................... 272 8.14 Annual Irrigation Water Requirements ............................................................. 277 List of Symbols ......................................................................................................... 278 References ................................................................................................................. 281

Design and Operation of Farm Irrigation Systems

Chapter 9 Drainage Systems

ix

289

9.1 Introduction ......................................................................................................... 289 9.2 Environmental Considerations ............................................................................ 291 9.3 Drainage System Requirements........................................................................... 293 9.4 Approaches to Design of Subsurface Drainage Systems ..................................... 293 9.5 Determining Design Variables ............................................................................ 313 9.6 Drainage System Materials.................................................................................. 315 9.7 Construction Methods and Equipment ................................................................ 317 9.8 Drainage System Operation and Maintenance..................................................... 318 9.9 Drainage System Performance Evaluation .......................................................... 318 References ................................................................................................................. 319

Chapter 10 Land Forming for Irrigation

320

10.1 Introduction ....................................................................................................... 320 10.2 System Layout................................................................................................... 322 10.3 Soil Survey and Allowable Cuts........................................................................ 324 10.4 Topographic Survey .......................................................................................... 324 10.5 Earthwork Analyses........................................................................................... 329 10.6 Types of Equipment .......................................................................................... 334 10.7 Field Operating Procedures ............................................................................... 339 10.8 Cost and Contracting ......................................................................................... 344 10.9 Safety................................................................................................................. 344 References ................................................................................................................. 345

Chapter 11 Delivery and Distribution Systems

347

11.1 Introduction ....................................................................................................... 347 11.2 Irrigation Water Delivery .................................................................................. 348 11.3 Farm Water Distribution Systems ..................................................................... 369 References ................................................................................................................. 387

Chapter 12 Pumping Systems

392

12.1 Introduction ....................................................................................................... 392 12.2 Pump Components and Characteristics ............................................................. 394 12.3 Selecting a Pump ............................................................................................... 404 12.4 Power Units ....................................................................................................... 419 12.5 Pump Controls ................................................................................................... 425 12.6 Economic Considerations .................................................................................. 426 12.7 Impact of Pumping System Modification.......................................................... 430 12.8 Maintenance and Testing................................................................................... 432 List of Symbols ......................................................................................................... 434 References ................................................................................................................. 434

x

Chapter 13 Hydraulics of Surface Systems

Contents

436

13.1 Introduction ....................................................................................................... 436 13.2 Basic Concepts of Surface Irrigation Hydraulics .............................................. 438 13.3 Variables That Influence the Surface Irrigation Process ................................... 441 13.4 Conservation Laws for Mass and Momentum................................................... 451 13.5 Hydrologic Modeling of the Surface Irrigation Process .................................... 452 13.6 Hydrodynamic Modeling of the Surface Irrigation Process .............................. 461 13.7 Estimation of Field Parameters.......................................................................... 479 References ................................................................................................................. 491

Chapter 14 Design of Surface Systems

499

14.1 Introduction ....................................................................................................... 499 14.2 Design Considerations and Approaches ............................................................ 500 14.3 Sloping-Furrow Irrigation.................................................................................. 509 14.4 Border-Strip Irrigation....................................................................................... 518 14.5 Level-Basin and Level-Furrow Irrigation.......................................................... 524 14.6 Surface Irrigation System Headworks and Control of Inflow ........................... 528 References ................................................................................................................. 530

Chapter 15 Hydraulics of Sprinkler and Microirrigation Systems

532

15.1 Introduction ....................................................................................................... 532 15.2 Hydraulics of Pipe Systems ............................................................................... 533 15.3 Irrigation System Valves ................................................................................... 547 15.4 Water Distribution to the Soil............................................................................ 554 15.5 Redistribution in the Soil................................................................................... 554 15.6 Summary ........................................................................................................... 554 List of Symbols ......................................................................................................... 555 References ................................................................................................................. 556

Chapter 16 Design and Operation of Sprinkler Systems

557

16.1 Introduction ....................................................................................................... 557 16.2 Components of Sprinkler Systems..................................................................... 558 16.3 Design Fundamentals ........................................................................................ 560 16.4 Application Uniformity ..................................................................................... 576 16.5 Solid Set Systems .............................................................................................. 580 16.6 Periodically Moved Laterals.............................................................................. 587 16.7 Center Pivots ..................................................................................................... 595 16.8 Lateral Move Systems ....................................................................................... 613 16.9 Low Energy Precision Application (LEPA) Systems........................................ 618 16.10 Travelers.......................................................................................................... 621 16.11 Auxiliary Uses of Sprinkler Systems............................................................... 626 16.12 Safety............................................................................................................... 626

Design and Operation of Farm Irrigation Systems

xi

16.13 Summary ......................................................................................................... 627 List of Symbols ......................................................................................................... 627 References ................................................................................................................. 628

Chapter 17 Microirrigation Systems

632

17.1 Introduction ....................................................................................................... 632 17.2 Microirrigation Systems .................................................................................... 638 17.3 Design Factors ................................................................................................... 642 17.4 Hydraulics of Emitters and Emitter Design Variation....................................... 647 17.5 Microirrigation Design ...................................................................................... 652 17.6 Designing the System Control Head.................................................................. 656 17.7 Installation ......................................................................................................... 658 17.8 Maintenance ...................................................................................................... 658 17.9 Management ...................................................................................................... 661 17.10 Scheduling Microirrigation.............................................................................. 664 17.11 Plugging of Microirrigation Systems............................................................... 666 17.12 Subsurface Drip Irrigation ............................................................................... 668 17.13 Subirrigation .................................................................................................... 675 17.14 Microirrigation in Nurseries and Greenhouses ................................................ 675 References ................................................................................................................. 677

Chapter 18 Water Table Control Systems

684

18.1 Introduction ....................................................................................................... 685 18.2 Management of Soil Water by Water Table Control ......................................... 686 18.3 System Design and Operation in Humid Regions ............................................. 691 18.4 System Design and Operation in Arid Regions ................................................. 703 18.5 Documentation of System Design and Installation............................................ 717 18.6 Summary ........................................................................................................... 717 References ................................................................................................................. 718

Chapter 19 Chemigation

725

19.1 Introduction ....................................................................................................... 725 19.2 Backflow Prevention and Safety ....................................................................... 727 19.3 Injection Systems............................................................................................... 732 19.4 Injection System Calibration ............................................................................. 739 19.5 Irrigation System Considerations ...................................................................... 741 19.6 Calculating Injection Rates................................................................................ 750 References ................................................................................................................. 752

Chapter 20 Wastewater and Reclaimed Water Irrigation

754

20.1 Introduction ....................................................................................................... 754 20.2 Constituents and Characteristics of Wastewater and Reclaimed Water ............ 756

xii

Contents

20.3 Nutrients in Wastewater and Reclaimed Water ................................................. 764 20.4 Health Concerns ................................................................................................ 765 20.5 Crops Suitable for Irrigation with Wastewater and Reclaimed Water............... 769 20.6 Design of Nitrogen Loading Rate...................................................................... 770 20.7 Irrigation Systems Using Wastewater ............................................................... 774 20.8 Design and Operation of Systems to Protect Human Health ............................. 779 20.9 Monitoring......................................................................................................... 781 References ................................................................................................................. 783

Chapter 21 Evaluating Performance

790

21.1 Introduction ....................................................................................................... 790 21.2 Management Variables ...................................................................................... 792 21.3 Measures of Performance .................................................................................. 794 21.4 Field Evaluation Methods.................................................................................. 798 21.5 Economics ......................................................................................................... 799 21.6 Analysis and Interpretation................................................................................ 800 21.7 Mobile Labs....................................................................................................... 801 21.8 Outline for Evaluations...................................................................................... 802 21.9 Conclusions ....................................................................................................... 802 References ................................................................................................................. 803

Appendix A Glossary............................................................................................804 Appendix B Annotated Bibliography of Irrigation Standards...........................838 Index ................................................................................................851

CHAPTER 1

INTRODUCTION Glenn J. Hoffman (University of Nebraska, Lincoln, Nebraska) Robert G. Evans (USDA-ARS, Sidney, Montana) Abstract. This chapter introduces the second edition of “Design and Operation of Farm Irrigation Systems.” The distribution and development of irrigation in the United States and throughout the world is reviewed. Issues facing irrigated agriculture including environmental concerns, sustainability, and changes in water policy are discussed. The chapter concludes with a discussion of options for future directions and some needs for irrigation research and education. Keywords. Environmental concerns, History, Irrigation, Sustainability, Water policy, Worldwide development.

1.1 OVERVIEW This monograph updates and complements ASAE Monograph 3, “Design and Operation of Farm Irrigation Systems,” which was edited by M. E. Jensen and published by the American Society of Agricultural Engineers in 1980. The 1980 monograph has been the most requested and most successful of all ASAE monographs over the intervening years. More than 8500 copies of the first edition have been distributed worldwide. Significant developments and additions to our understanding and knowledge of irrigation along with major changes in the environmental, ecological, sociological, and political scenes attest to the need for a second edition. This monograph provides current research results and state-of-the-art knowledge on all aspects of irrigation on the farm. It is intended to provide design procedures and scientific guidance for the management of water resources and irrigation systems. This edition provides updated material for many of the chapters covered in the first edition. Other chapters are significantly different or completely new. Chapters that are built upon the subjects of the first edition are system selection, soil-water relationships, salinity control, water requirements, drainage systems, land forming, delivery systems, pumps, hydraulics, design of surface systems, design of sprinkler and microirrigation systems, evaluation of performance, and irrigation management. New or substantially different subjects include efficiency and uniformity, environmental considerations, sustainability of irrigation, water table control, chemigation, wastewater and reclaimed irrigation water, and site-specific management. By the number of new

2

Chapter 1 Introduction

subjects alone, the reader should be convinced of the many new issues in irrigation that need to be addressed today. This monograph should benefit practicing professionals, farmers, researchers, and governing agencies. It provides technical data, design procedures, and management strategies to benefit all who make decisions about the design and operation of farm irrigation systems.

1.2 WORLDWIDE IRRIGATION DEVELOPMENT At the beginning of a new millennium it is appropriate to review the history of irrigation and its impact on civilization and the global food supply. Irrigation was developed to counter both short-term and long-term drought on crop production in arid and semiarid areas whenever a reliable supplemental water supply was nearby. The prudent use of water enables the consistent production of food, fiber, and landscaping at levels and in locations where it would not otherwise be possible. An outstanding account of the history of irrigation has been given by Postel (1999), and much of the following historical discussion has been extracted from her writings. It is important to note the impact irrigation has had on civilization from its earliest beginnings to the present. Historians, through archeological evidence, credit early irrigation with the assurance of dependable food supplies; thereby allowing some members of the society to pursue activities other than hunting, nomadic grazing, or farming. Among the accomplishments of these early irrigation-based societies are writing, the wheel, water-lifting devices, yokes for draft animals, sailboats, palaces, pyramids, temples, ceramics, fine textiles, handicrafts, the Hammurabi code of law, fired bricks, cities, and various forms of accounting, taxation, management, and government. 1.2.1 Historical Development Six thousand years ago, settlers in Mesopotamia (which is now Iraq, and part of what is referred to as the Fertile Crescent) dug ditches and diverted water to their fields from the Euphrates River and initiated the practice of irrigation. Irrigation transformed both the land and society like no other previous activity by producing a dependable food supply, frequently so abundant that many people were able to pursue non-farm activities. The food surpluses had to be stored and distributed, which led to new forms of centralized management. This need for management led to stratified societies and centralized control. With some of the populace not needed in farming, historians credit these early settlers, called Sumerians, with the development of writing and the wheel as well as the creation of sailboats, water-lifting devices, and yokes for harnessing animals. With time, as the range of activities expanded, these early societies increased in population, which led to the first true cities. The Sumerians were followed by other irrigation-based civilizations in Mesopotamia. Among these were the Babylonians, who are renowned for building magnificent palaces and developing Hammurabi’s historic code of law which dealt, in part, with irrigation (Postel, 2000). Irrigated agriculture endured in Mesopotamia through several civilizations. Archeological evidence, however, indicates that around 2400 B.C. soil salination had reached detrimental levels. In those early civilizations, wheat was the preferred cereal for food, and it was grown in preference to barley. Wheat, however, is less salt tolerant than barley. Over time wheat’s share of the harvest dropped to less than two percent in some areas and barley production dominated, and by 1700 B.C. wheat was no longer cultivated. By then, even barley yields were declining. This archeological evidence

Design and Operation of Farm Irrigation Systems

3

has led historians to blame salination as the major contributor to the decline of civilization in the Fertile Crescent of present-day Iraq. To this day, the southern portions of the Fertile Crescent have never fully recovered from the demise caused by salinity. Elsewhere in the world, early societies dependent upon irrigation arose in the Indus River Valley in Pakistan, the Yellow River basin of China, and the Nile River Valley of Egypt. Much later, irrigation-based cultures developed in the western hemisphere. Central Mexico, coastal Peru, and the American Southwest each saw the rise and fall of an advanced society built on irrigated agriculture. The early civilization along the Indus River about 3500 B.C. also depended on irrigation. The complex, hierarchical society that developed by about 2300 B.C. lasted less than 500 years. Although conquest by invaders was probably the direct cause of the collapse of this society, instability caused by salination, siltation, flooding, and possibly climate change also contributed to its demise. Ancient Chinese efforts to control and use the water resources of the Yellow River in the north China plain began about 4000 years ago. The Yellow, however, proved to be a difficult river to tame. It emerges from the Loess Plateau carrying some 1.4 billion Mg of silt each year, making channel alterations and flooding a common occurrence. The Yellow has breached its dikes more than 1500 times over the last two millennia. Heavy silt deposits have elevated the river above its surrounding plain necessitating the continual raising and reinforcing of dikes to prevent flooding. Ultimately a breach of the dike occurs and a change in the course of the river results. The river has changed course about once every century. These unpredictable shifts have been devastating to irrigated agriculture and human settlement. In sharp contrast to the other early civilizations built upon irrigation in the eastern hemisphere, the Egyptian civilization in the Nile River Valley has endured for 5000 years without interruption. Egypt’s civilization has survived warfare and conquests and widespread disease. Only in recent times has the sustainability of Egyptian agriculture come into question. In response to a 20-fold increase in population over the last two centuries, Egypt replaced its time-tested agriculture based on the Nile’s natural flow rhythms with more intensified irrigation and flood management that requires complete control of the river. The natural flow rhythm of the Nile was once dominated by an annual flood that was relatively benign, predictable, and timely. With nearly flawless predictability the river rose in southern Egypt in early July and reached flood stage in the vicinity of Aswan by mid-August. The flood continued to surge northward and reached the northern end of the valley by the end of September. At its peak, the flood covered the floodplain to a depth of 1.5 meters. By late November most of the valley was drained. Egyptian farmers then had well-watered fields that had been fertilized by the rich silt carried from Ethiopia and deposited across the floodplain. Crops were then planted as the mild winter began and harvested in the spring just in time for the cycle to repeat. Egyptian irrigators did not experience many of the troublesome problems of other early civilizations. Fertility was renewed each year by the silt-laden floodwater and the inundation prior to planting pushed whatever salts had accumulated down below the root zone. In contrast to other ancient civilizations, early Egyptian society did not centrally manage irrigation works. Irrigation was carried out on a local rather than a regional or national scale. Despite the existence of many civil and criminal codes in ancient Egypt, no evidence exists of written water law. Apparently, water management was

4

Chapter 1 Introduction

neither complex nor contentious, and oral traditions of common law withstood all tests over a considerable length of time. The many political disruptions at the state level, which included numerous conquests, did not greatly impact the system’s operation or maintenance, probably because the politicians had no control over the flows of the Nile. Though later in time and smaller in size, the irrigation-based civilizations that developed in the western hemisphere also shaped cultural developments. The geographic cradle of civilization in the west was Central America. Settled villages evolved about 2000 B.C., when the productivity of domesticated corn reached a level that could support stable communities. Around 300 B.C., canals were used for irrigation throughout the Tehuacan Valley southeast of Mexico City. Along the Peruvian coast, an irrigated crop could be grown in four months leaving time to construct pyramids and temples and to develop fine textiles, ceramics and handicrafts. A centralized bureaucracy was formed to manage the large irrigation system and to control the distribution of water. In North America, the Hohokam culture thrived for more than 1000 years along the Gila and Salt rivers of south central Arizona but then the culture disappeared suddenly around 1400 A.D., most likely due to prolonged severe drought. Archeologists have documented more than 500 km of main irrigation canals, with many of them linked in networks. In the 16th century, Leonardo da Vinci studied the Amo watershed in northern Italy. His studies led to a better understanding of the relationship between a river’s catchment area and its flow. This study helped establish the fundamentals of hydrology and river basin management (Newson, 1992). Several hundred years passed before the principles of hydraulics and mechanized water control technologies were developed that together transformed irrigation from an art to a science. 1.2.2 Recent Developments and Trends Over many centuries, irrigated lands have become essential to the world’s food supply. These lands now constitute approximately 20% of the world’s total cultivated farmland but produce about 40% of the food and fiber. Irrigated agricultural activities also provide considerable food and foraging areas for migratory and local birds as well as other wildlife. In short, irrigation underpins our modern world society and lifestyles. In 1800, the worldwide irrigated area totaled about 8 million ha. The irrigated area increased five-fold during the 19th century, mainly because during the latter half of the century much of the scientific and technical foundation for irrigation was developed. In the 20th century, global irrigation grew from 40 to more than 270 million ha, an almost seven-fold increase (Figure 1.1). Currently India and China, with nearly the same amount of irrigated cropland (57 and 55 million ha, respectively), together account for about 40% of the world’s irrigated land. Ten countries collectively account for two-thirds of the world total (Table 1.1). It is estimated that from 36% to 47% of the world’s food is produced by irrigated production (Gleick, 1998; Postel, 1999). The total amount of irrigated land in a country depends upon its size, arable cultivated land, and climatic conditions. Thus, it is interesting to note the proportion of each country’s arable land that is irrigated. The percent of arable land that is irrigated within a country and by continent is given in Table 1.1. More than half of the arable land is under irrigation in Pakistan, Iraq, Japan, Bangladesh, and Iran; nearly or all of

Design and Operation of Farm Irrigation Systems

5

300

Millions of Hectares

250

200

150

100

50

0 1800

1825

1850

1875

1900

1925

1950

1975

2000

Years

Figure 1.1. Increase in irrigated area worldwide from 1800 to 2000 (adapted from Postel, 1999, and FAOSTAT, 2005).

the arable land in Uzbekistan and Egypt is irrigated. Thirty-eight percent of all the arable land in Asia is irrigated. Brazil has seen rapid growth in recent years. In contrast, in Africa, a continent desperate for food, only 7% of the arable land is irrigated. On average, irrigation is practiced on 20% of the arable land in the world (Table 1.1). Worldwide irrigated agriculture increased significantly during the second half of the 20th century with large government investments, major financial support from international donors and lenders, and the spread of improved pumping technologies. This period coincided with numerous, large-scale, government-sponsored surface water irrigation projects in many countries and the proliferation of both private and public groundwater wells in others. In China, waterworks were constructed on major Huai, Hai, and Yellow River basins, mostly for rice and small grain production. By 1990, more than 4 million ha were irrigated from the Yellow River. With improved pumping and well-drilling technologies and the availability of electricity, China turned to groundwater to irrigate the North China Plain. The number of wells increased from 110,000 in 1961 to more than 2 million in the mid-1980s (Postel, 1999). Likewise, irrigation by tubewells increased dramatically in India and Pakistan during the last half of the 20th century. Irrigation from tubewells expanded from 100,000 ha in 1961 to 11.3 million ha in 1985 (World Bank, 1991). In this same time period, canal building for surface irrigation by the Indian government doubled the surface irrigated area. In the Indus River basin, two large storage dams and corresponding construction of irrigation canals transformed the basin into the world’s largest contiguous irrigation network, covering 14 million ha.

6

Chapter 1 Introduction

Table 1.1. Irrigated areas in the top 25 countries, by continents, and in the world in 2002 (adapted from FAOSTAT, 2005). These totals are estimates because definitions of irrigated land vary by country. Arable Land Irrigated Irrigated Area Region of the World (percent) (million hectares) Country Egypt 100 3.4 Uzbekistan 96 4.3 Pakistan 83 17.9 Iraq 60 3.5 Japan 59 2.6 Bangladesh 58 4.6 Iran 50 7.5 Viet Nam 45 3.0 China 38 54.9 India 35 57.2 Italy 34 2.8 Romania 33 3.1 Thailand 31 5.0 Afghanistan 30 2.4 Spain 28 3.8 Mexico 25 6.3 Indonesia 23 4.8 Turkey 20 5.2 France 14 2.6 United States 13 22.5 Kazakhstan 11 2.4 Ukraine 7 2.3 Brazil 5 2.9 Australia 5 2.5 Russia 4 4.6 Continent Asia 38 193.9 North & Central America 12 31.4 Europe 9 25.2 South America 9 10.5 Africa 7 12.9 Oceania 6 2.8 World 20 276.7

In the former Soviet Union, leaders expanded irrigation in areas with otherwise favorable climatic conditions that lacked adequate rainfall for crop production. They concentrated on two key areas—the Central Asian republics, which accounted for about 40% of the Soviet irrigated area before the nation’s breakup, and the southeastern European region, including parts of Russia and Ukraine. With irrigation water drawn from the rivers feeding into the Aral Sea, central Asia became a major cotton growing region. However, the Aral Sea is a small fraction of its previous size because

Design and Operation of Farm Irrigation Systems

7

of declining inflows due to greatly increased upstream diversions for irrigation. While large scale irrigation in southeastern Europe protected this important grain-producing region from drought there has been severe damage to the region’s natural ecosystems. Presently, chronic scarcity of water is experienced or expected in large parts of Africa and the Middle East, the northern part of China, the former Soviet Union and the Central Asian republics, parts of India and Mexico, the western part of the U.S., and northeast Brazil. A region is said to be water stressed when its renewable water supplies drop below about 1700 cubic meters per capita (Gleick, 1998; Postel, 1999). With supplies below this level, it becomes difficult for a country to mobilize enough water to satisfy all the food, household, and industrial needs of its population. For example, about 1000 Megagrams (Mg) of water are required to produce 1 Mg of grain. With grain being the staple of the human diet, water-stressed countries import grain to satisfy food requirements and reserve their water for other uses. Collectively, 32 of the 34 water-stressed countries in Africa and Asia import about 50 million Mg of grain annually, about a quarter of the total traded internationally (Postel, 2000). During the first quarter of the 21st century, the global irrigation base is expected to grow at less than 1% a year, down from the annual growth rate of more than 3% for the last half of the 20th century. This has largely been the result of diminishing water supplies and increased demands from other sectors. In most areas, the best and easiest sites are already developed. A 1995 study by the World Bank reported that irrigation development costs on more than 190 of their funded projects averaged about $4,800 per hectare (Jones, 1995).

1.3 IRRIGATION DEVELOPMENT IN THE UNITED STATES Irrigation has been practiced in the Southwest region of the U.S. for two millennia. Today, irrigation is practiced in every state of the union with the total irrigated agricultural cropland area exceeding 21 million ha (21.3 million ha using the USDA-NASS, 2004, estimate; 22.5 million ha using a different methodology, as listed in Table 1.1) with a farm gate value exceeding $47 billion per year. This represents 49% of the market value of all crops from 18% of all harvested U.S. croplands. Nationally, per hectare sales from harvested irrigated lands average more than 4.5 times the sales from nonirrigated land (Gollehon, 2002). However, this is a somewhat misleading figure because much higher input, labor, and equipment costs often result in net returns similar to that for rainfed or dryland producers, especially for field crops. This is usually not the case for tree and vine crops, vegetables, and other highvalue crops. It is estimated that there are also approximately 16.4 (± 3.6) million ha of managed turf in the U.S. (Milesi et al., 2005) that is not included in any of the estimates of cropland referenced in this chapter (or elsewhere). Turfgrass areas include lawns, airports, institutional facilities, military bases, schools, parks, golf courses, athletic fields, churches, cemeteries, etc. These areas are expected to continue growing, and much of this land is irrigated to some degree, almost totally with sprinklers. The large area of turfgrass makes it the largest irrigated crop in the U.S. Milesi et al. (2005) estimated that irrigated turfgrass amounted to three to four times the area of irrigated corn (about 4.3 million ha), the largest irrigated agricultural crop. Applied water and other inputs are high because turf is generally managed for appearance rather than profitability. The amount of irrigated turfgrass is extremely significant because it indicates that ur-

8

Chapter 1 Introduction

ban irrigators have a large potential for water and energy savings and pollution abatement that is comparable to irrigated agriculture. Furthermore, the political and economic base clearly lies with the urban users and they will have a major voice in future land use and water policy issues. Thus, including turf, there is a total of about 37 million ha of irrigated land in the US. The impact of turfgrass is quite significant because urban users are often competing for the same water supplies, especially in the 17 western states (not including Hawaii and Alaska) that contain about 30% of the total turf area in the U.S. In addition, urbanization often expands on to irrigated agricultural lands, and much of the associated water is often lost to agriculture as it is then used for urban landscapes. 1.3.1 Historical Development of Irrigation in the U.S. Irrigation by indigenous people of the Southwest is known to have existed as early as 100 B.C. Early irrigation was practiced in the Salt River Valley, on the Colorado Plateau, and along several watercourses elsewhere in the West. Corn, beans, squash, milo, peaches, and other crops were grown through an intricate network of ditches and canals. Some of the early immigrants arriving in North America, particularly from the Mediterranean area, brought with them a heritage of irrigation as part of farming. For example, Spanish colonists irrigated extensively at the missions established along the Pacific Coast beginning in the 1760s. In the mid-1800s, irrigation development in the U.S. expanded along with the settlement of the West. Most of the early projects were accomplished by private enterprise. In 1847, an advance party of Mormons settled in the Salt Lake Valley of Utah and began diverting water to grow potatoes. Early Mormon activities set the stage for other private irrigation ventures, and by about 1875, the center for irrigation innovation and development had shifted to Colorado and California. By 1890, irrigation was practiced on more than 400,000 ha in California, 350,000 ha in Colorado, and more than 100,000 ha in Utah. About this time, irrigation became a central theme of the federal government’s strategy to encourage settlement of the West. The Desert Land Act of 1877 and the Carey Act of 1894 were intended to stimulate private and state development of irrigated land. The federal government became involved in irrigation development with the passage of the Reclamation Act of 1902. The Bureau of Reclamation formed by this act provided engineering expertise and financial capital to develop large irrigation projects throughout the West. The irrigated area in the U.S. increased from about 1 million ha in the 1880s to about 8 million ha by the middle of the 20th century. At this time, the major irrigated regions were the Southwest (2.2 million ha), the Mountain States (2.5 million ha), and the Pacific Northwest (1.4 million ha) (U.S. Department of Commerce, 1983). The drought years of the 1950s stimulated irrigation development in the southern Great Plains. Then, with the advent of center pivot sprinkler irrigation systems and with groundwater readily available, irrigation expanded rapidly in the central Great Plains during the 1960s and 1970s. During the same time period, irrigation increased markedly in the southeastern states. The total irrigated area declined in the 1980s because of depressed farm commodity prices, increased energy costs, and declining water resources. In the 1990s the total irrigated area recovered and data for 2003 shows the total irrigated area to be about 21.3 million ha (USDA-NASS, 2004) Recent growth in U.S. irrigation has been mostly in the southeastern areas of the country, primarily in the Mississippi River delta. The 1980s and 1990s saw growth in

Design and Operation of Farm Irrigation Systems

9

the lower Mississippi River areas of Missouri, Mississippi, and Louisiana; the 1960s and 1970s had significant growth in Georgia, South Carolina, Alabama, and Arkansas. Mississippi, Arkansas, and Missouri added about 1.2 million ha of irrigated land since 1980. This essentially doubled irrigation in these states (USDA-NASS, 2004). 1.3.2 Current Status of U.S. Irrigation The “official” survey of agricultural irrigation in the U.S. is typically taken as part of the U.S. Census of Agriculture. The most recent results are in the 2003 USDA Farm and Ranch Irrigation Survey as part of the 2002 Census of Agriculture (USDA-NASS, 2004). The 2002 Census of Agriculture reported 21.3 million ha of irrigated land. The USGS (Hutson et al., 2004) estimated about 25 million ha of total irrigated farmland in the U.S. in 2000. The differences between the two values are due to different estimating methodologies. Data reported in the figures and tables in this section are from the 2002 Census. The approximate agricultural cropland area irrigated in each state is indicated in Figure 1.2. Except for Iowa and North Dakota, the states west of the Mississippi River all have in excess of 100,000 ha of irrigated land. In comparison, only seven states east of the Mississippi had more that 100,000 ha of irrigation. Other interesting comparisons can be made from the U.S. agricultural census data. The source of water for farm irrigation is summarized in the 2003 survey and indicates that 26% of the water used for irrigation comes from surface sources off the farm, while about 60% of the irrigation water comes from groundwater delivered from wells, and only 14% comes from on-farm surface sources (Table 1.2). Nationally, thermoelectric power accounts for 48%, irrigation 34%, and municipal and industrial uses 16%, of all water withdrawn in the U.S. from both fresh and saltwater sources (Hutson et al., 2004). However, irrigation is the largest user of freshwater supplies, especially in the arid west, accounting for about 40% of the total U.S. freshwater withdrawals from both surface and subsurface supplies. The methods of irrigation have changed over time. From early diversions of water from streams by ditches dug by hand, irrigation technology has developed to include

Figure 1.2. Approximate irrigated farm land in the U.S. in 2003 by state (adapted from USDA-NASS, 2004).

10

Chapter 1 Introduction

Table 1.2. Amount of water used from various sources for U.S. farm irrigation in 2003 (adapted from USDA-NASS, 2004). Area Volume Average Irrigated Applied Depth Applied Water Source (hectares) (km3) (cm) Wells 13,094,000 53.6 41 On-farm surface water 2,946,000 14.5 49 Off-farm surface water 5,614,000 39.0 69 U.S. total 21,654,000[a] 107.1 49 [a]

Sum of land irrigated exceeds actual U.S. total (21,300,000 hectares) because some land receives water from more than one source.

massive reservoirs and networks of canals to satisfy gravity irrigation systems, manually and mechanically moved sprinkler systems, and a variety of low-flow systems referred to as microirrigation. Originally, irrigation was accomplished by methods utilizing gravity to distribute and apply water. In the 20th century, sprinkler technology along with low-cost aluminum and later PVC pipe was developed, and currently more land is irrigated by sprinklers than by gravity. Similarly, low-flow microirrigation is based on plastic technologies evolved during the last quarter of the 20th century and is now used on about 5% of the irrigated area in the U.S. (USDA-NASS, 2004), but on less than 1% worldwide. Table 1.3 summarizes the amount of land irrigated by the various irrigation methods. In 2003, 50.5% of the area being irrigated was with sprinklers, 43.4% by gravity methods, 5.6% by microirrigation, and 0.5% by subirrigation. Of the 10.9 million ha irrigated by sprinklers, 79% of this area utilized center pivots, which amounts to about 40% of all the land irrigated in the U.S. To realize energy savings through lowpressure applications, 90% of center pivots reported in the 2002 Census use water pressures below 400 kPa. Furrow irrigation is practiced on 51% of the land using gravity systems, while border or basin methods are utilized on 38% of the surface irrigated lands. The average depth of irrigation water applied per unit land area can also be calculated from the Census results. Nationwide, an average depth of 49 cm of irrigation water is applied annually to irrigated lands in the U.S. (Table 1.2). For land irrigated by pumped groundwater from wells, an average depth of only 41 cm of irrigation water is applied each year. In comparison, an average depth of 69 cm of water is applied when the supply is surface water from off the farm. The USGS reported that more than 195 billion cubic meters of water were diverted from off-farm surface sources annually with an estimated average application depth of about 75 cm in 2000 (Hutson et al., 2004). This is slightly higher than the 2002 Census of Agriculture estimates. The differences in application depth estimates between the USGS and Census of Agriculture reports are probably caused by the USGS analysis being more indicative of gravity systems supplied from off-farm surface water sources. These tend to apply more water than other irrigation methods. The comparisons are also biased by how water source was considered. Center pivots are generally more efficient than gravity methods and are the most prevalent irrigation method in the subhumid regions of the Great Plains and humid areas in the east which use well water to supplement precipitation. In contrast, gravity irrigation is dominant where the source of water is off-farm and where the climate is generally more arid, as in the Western States.

Design and Operation of Farm Irrigation Systems

11

Table 1.3 . Comparison of irrigation methods in the U.S. in 2003 (adapted from 2002 Census of Agriculture, USDA-NASS, 2004). Irrigated Area Irrigation Method (ha) % of Total Gravity systems Furrow 4,746,000 Border/basin 3,582,000 Uncontrolled flooding 930,000 Other 104,000 U.S. total, gravity systems 9,362,000 43.4 Sprinkler systems Center pivot, pressures above 400 kPa 785,000 Center pivot, pressures 200 to 400 kPa 3,910,000 Center pivot, pressures below 200 kPa 3,926,000 Linear move 139,000 Side roll, wheel move, or other mechanized move 739,000 Traveler or big gun 256,000 Hand move 674,000 Solid set and permanent 477,000 U.S. total, sprinkler systems 10,906,000 50.5 Microirrigation systems Surface drip 574,000 Subsurface drip 164,000 Microsprinklers 472,000 U.S. total, microirrigation systems 1,210,000 5.6 Subirrigation 113,000 0.5 Total U.S. irrigation[a] 21,590,000 [a]

The U.S. total irrigated area is larger than the 21.3 million ha quoted previously because more than one irrigation method may be used on some lands.

1.3.3 Trends in U.S. Irrigation The irrigated agricultural area in the U.S. from 1939 to 2003, based on national census data, is given in Figure 1.3. Total irrigated area increased at an average rate of 3.6% during the last six decades of the 20th century. The rate of increase is relatively steady except for the sudden increase in the 1970s followed by a decline in the 1980s. The surge in irrigated area in the 1970s can be attributed to the deployment of center pivot systems and the installation of systems, particularly in the southeast, to augment rainfall. The decline during the 1980s was created by a combination of depressed farm commodity prices, increased energy costs, declining water resources, and some apparent changes in statistical procedures in determining irrigated areas. What is interesting is the return to a positive growth rate of irrigated land during the early 1990s, but a slight decline again during the last 5 years of census data (1997-2002). This slight decline is expected to continue. It is also expected that the conversion from surface irrigation methods to self-propelled (center pivots and linear move) and microirrigation technologies will increase, but will have little effect on total irrigated acres.

12

Chapter 1 Introduction

Millions of Hectares

25 20 15 10 5 0 1935

1954

1974

1992

Years

Figure 1.3. The total agricultural irrigated area in the U.S. from 1939 to 2002 (adapted from U.S. Department of Commerce, 1978, and 2002 Census of Agriculture, USDA-NASS, 2004).

1.4 ISSUES FACING IRRIGATED AGRICULTURE Irrigation is the largest single consumer of water on the planet, accounting for about 80% of the total freshwater consumed and about two thirds of the total diverted for human uses, and it is responsible for more than 40% of all agricultural production. Irrigation has permanently changed the social fabric of many regions around the world. It has provided major economic development of many semiarid and arid areas, stabilizing rural communities, increasing income, and providing new opportunities for economic advancement to many. However, the world’s supply of freshwater is basically constant, and the amounts allocated to irrigated production will undoubtedly decrease substantially because of increased consumption by non-agricultural users. The development of irrigation early in the 20th century in the U.S. was created by an enormous level of governmental involvement, and despite the major contributions to a stable and bountiful food supply, the commitment toward agriculture worldwide, and particularly in the U.S., has diminished in recent years. Many members of society exhibit ambivalent feelings toward irrigated agriculture. This phenomenal development has not been without controversy and not without problems or critics. Opposition to irrigation has two primary themes. First, irrigation is increasingly coming into direct competition with other water uses for scarce water resources. There are escalating demands from other water user sectors, including potable water, recreation, and tourism, that are looking to agriculture to supply the needed water by conservation and other measures (Clemmens and Allen, 2005). Much of this development in the Western U.S. has occurred during a long term, uncharacteristic, wet period causing unrealistic expectations and over-allocations of water supply. Secondly, irrigated agriculture (although not alone) can have negative environmental impacts on water and soil quality over broad areas. Irrigation can degrade water quality, erode soils, reduce groundwater levels, deplete stream flow, and alter hydrologic regimes leading to unhealthy trends in aquatic and riparian ecosystems. The nation’s

Design and Operation of Farm Irrigation Systems

13

commitment to minimize such environmental issues continues to increase and irrigated agriculture will receive more and more pressure to change existing practices to address these concerns. It is expected that the current commodity surpluses will not last in the face of increasing population coupled with the worldwide decrease in ocean fisheries. These factors will place large worldwide demands on agriculture to increase the global production of animal/fish protein, food, fiber, and livestock feed. Agriculture will also be asked to help supplement the world’s ever-increasing energy needs by producing more bio-based fuels, lubricants, and chemical plant feedstocks that are now provided by the petrochemical industry. Irrigated agriculture, including greatly expanded intensive greenhouse culture, is expected to provide more than 70% of this increased future food and fiber demand worldwide in the next 25 to 50 years. Aquaculture will also expand significantly, to offset the worldwide decline in fish catches, but the net amount of fish protein will probably remain about the same. However, most of the potential irrigation development has already occurred, and, in fact, productivity of the irrigated land base around the world is declining due to soil salination, waterlogging, and soil erosion. Despite phenomenal advancements in crop breeding, genetic engineering, and other technologies, it is not known where or how all of this increased productivity, needed to satisfy the increasing population, will occur. Further stresses are being imposed on existing water resources by endangered species regulations, international and interstate agreements on water allocations, energy availability, and a suffering agricultural economy. Declining water tables in many regions due to excessive pumping for irrigation are also a major concern. Temporary and long-term water transfers between users, inter-basin diversions, and emerging economic water markets are also confounding the issues. Water and land resources and their many competing uses must be considered in a regional and international framework. Withdrawals from a river deplete instream flows that may have a large impact on downstream water users (municipalities, industries, natural areas and wildlife, agriculture, livestock operations, recreation, navigation, tourism, and hydropower production). Reservoir releases for power often conflict with other competing demands for water, including irrigation. Withdrawals from aquifers by wells can negatively affect groundwater levels over large areas and reduce recharge to streams and other water bodies. Global climate change may be further exacerbating the problems through changing temperatures and probably by altering annual precipitation amounts and regional rainfall distribution patterns. The Intergovernmental Panel on Climate Change (2001) concluded that a global warming attributable to human activity is now evident in the historic record. Global mean surface temperature relative to 1990 is expected to increase by about 2°C by 2100. If global warming occurs, it is certain to have a major impact on water supplies, and the increased variability in precipitation will provide major challenges for the agricultural sector. A warmer climate would accelerate the hydrologic cycle, increasing both the global rates of precipitation and evapotranspiration (ET). Timing of precipitation as well as runoff from mountain snowmelt may be different from historical norms. There is evidence that some of these changes are already occurring, but regional impacts are uncertain at best. The hydrologic uncertainties are compounded because relatively small changes in precipitation and temperature can have sizable effects on the volume and timing of runoff as well as ET, especially in

14

Chapter 1 Introduction

arid and semiarid areas. In short, the prospect of global warming introduces major new uncertainties and challenges for irrigators and society as a whole. All of these factors have created a water demand crisis that only continues to worsen. Consequently, there is an urgent need to improve the management of irrigation worldwide to conserve water, soil, and energy as well as provide food, fiber and other critical needs. Adjustments will have to be made by all, but the major expectations will likely be focused around agriculture, especially irrigated production in arid areas. Equitable solutions to these issues must include irrigated agriculture as well as all of the other users in determining future programs and directions. Managing and developing infrastructure and policies for water security to equitably satisfy the demands of all users of this limited resource will be a difficult and lengthy task. 1.4.1 Environmental Concerns The major environmental issues relevant to irrigation are those concerned with the protection and management of water resources and water quality. During the past few decades, society has become increasingly conscious of and concerned about the impacts of irrigated agriculture on environmental quality. The relative significance of environmental issues varies among regions of the country, but the types of issues confronting irrigation generally are the same. However, with few exceptions, environmental laws and policies have not addressed irrigation-related concerns. Irrigated agriculture has had profound positive and negative impacts on the environment. Irrigation has contributed to losses and changes of aquatic and riparian habitats and the decline of some native species dependent on those habitats (Wilcove and Bean, 1994). Irrigation runoff is a significant source for potential pollutants in surface waters (National Research Council, 1989). Federal and state responses to environmental concerns regarding irrigated agriculture include efforts to control soil salination and agricultural nonpoint sources of water pollution, policies to protect stream flows and wetlands, and restrictions on the application of pesticides. Environmental issues related to instream flow and wetland ecosystems arise wherever water is withdrawn for irrigation. Dams and diversions for surface water supplies reduce stream flows, altering the natural hydrograph and changing water temperature and flow regimes. These changes may degrade fish spawning and rearing habitats. The draining of wetlands for irrigated agriculture impacts waterfowl and other aquatic species that use these habitats. As an example, 92% of the historic wetland areas in the San Joaquin Valley of California have been converted to irrigated lands (San Joaquin Valley Drainage Program, 1990). Large-scale water resource development projects including irrigation, flood control, recreation, hydropower, and navigation have also altered aquatic and riparian habitat conditions of the Platte, Colorado, Columbia, and Snake rivers. Irrigation development has greatly increased wetland areas in many arid areas. This is generally perceived as a social benefit because of increased land for migrating and non-migratory birds, wildlife, and recreational activities. Streamflows tend to be stabilized due to buffering by increased subsurface return flows, thereby enhancing recreation, fisheries, and waterfowl habitat over pre-irrigation conditions. However, efforts to improve irrigation efficiency may lead to a decrease in these artificially induced habitats because of reduced return flows to the area, thus leading to additional controversy with recreational interests, fishery and wildlife groups, and regulators.

Design and Operation of Farm Irrigation Systems

15

It should also be mentioned that urbanization is creating large pollution loads on freshwater supplies and estuaries. The amount of pollution in the world’s waterways is increasing while total water supply is decreasing due to increasing evaporation due to more intensive use. Urban populations are concentrated in coastal areas where treated and untreated sewage waters are discharged into the seas where the water is no longer available for reuse by agriculture and other uses, thereby contributing disproportionably to water scarcity. Desalination of seawater is an option in these areas but this also raises significant environmental concerns related to salt disposal (Seckler and Amarasinghe, 2000). 1.4.2 Sustainability of Irrigation Sustainability is a broad concept with many different meanings and connotations. Some people define it solely in terms of environmental factors, such as soil and water salinity, soil erosion, agrochemical use, and water pollution, that change the wildlife and riparian ecologies. Others view sustainability only in terms of economics and the continued agronomic production of food and fiber and thus include commodity prices, infrastructure development, equipment costs, pest control, and energy. By either definition, irrigated agriculture is not in fact economically or environmentally sustainable over the long term using existing technologies and policies. Sustainability is an important concept, because irrigation underpins our modern world society and lifestyles by providing at least 40% of our total food and fiber supply. In reality, society cannot afford and will not allow the loss of this tremendous asset. Despite the current problems and negative perceptions in many sectors of society, it is certain that irrigation will continue to be a necessary and important component of the world’s well-being and growth. Agricultural water security is obviously a major part of sustaining irrigated agriculture. It is a term that is used to describe the need to maintain adequate water supplies to sustain the food and fiber needs of the expanding world population. Factors that may impact water security include competition for water, environmental concerns, continued urbanization, government policy, and the globalization of the economy. One of the biggest threats to the sustainability of irrigated agriculture is salinity. Surface and groundwaters contain dissolved salts that are picked up as the water moves through various geologic formations and soils. When plants extract water from the soil for transpiration almost all of the salts in the soil water solution are left behind. Evaporation from the soil surface also deposits salts. (Irrigation without salinity problems occurs in more humid regions such as the irrigated rice culture areas in Asia, where it has been practiced for thousands of years, and also where the timing and amount of rainfall combined with good natural drainage are sufficient to leach salts from the system.) Excess accumulation of salts in the plant root zone can cause large yield reductions or even total crop failure, and inability to remove these salts will make agriculture unsustainable. Soil salination was probably the primary reason for the failure of many ancient societies in irrigated arid areas (see Section 1.2.1 above). Presently, about 30% of the land in the conterminous U.S. has a moderate to severe potential for salinity problems (Tanji, 1990). Many areas in the West, such as the Colorado River Basin, the northern Great Plains, and California’s San Joaquin and Imperial valleys, suffer salinity problems, as do large irrigated areas in India, Pakistan, and elsewhere around the world.

16

Chapter 1 Introduction

Improved irrigation efficiencies in arid areas often require additional water for leaching salts brought in with the irrigation water, which can cause a salinity concentration increase in the groundwater. Irrigation-induced soil salination can be avoided by providing adequate drainage and appropriate water management, but drainage is expensive and can degrade environmental water quality. In closed basins drainage from irrigated areas can render the water terminus biologically uninhabitable, as occurred in the Salton Sea in southern California. The Great Salt Lake in Utah is another drainage catchment terminus that was naturally uninhabitable and is being further degraded by urban and agricultural water users. Subsurface drainage and surface return flows from irrigation are sources for chemical and salt pollution in rivers, streams, lakes, and estuaries. Pollutants mobilized and transported by irrigation return flows and drainage into streams and man-made wetlands include trace elements (e.g. selenium, boron, and molybdenum), nitrogen, and salts, as well as pesticides, herbicides, and other chemicals. In sufficient concentrations, these pollutants may be detrimental to wildlife and birds, as was the case with Kesterson Reservoir in the San Joaquin Valley of California. Field drainage systems are an essential part of controlling salination of agricultural lands due to irrigation activities. Desalination or other costly types of treatment of some drainage waters may be required to overcome these obstacles; however, disposal of these effluents may also be a problem. Human and environmental health implications from reuse of degraded water must also be examined. Thus, the real question facing the world today is how to make irrigation sustainable, both environmentally and economically. The sustainability of irrigation depends on society’s ability to find ways to use this technology so that important benefits continue to be provided, but with less troublesome social, environmental, and economic consequences. Society will need to improve agricultural productivity, change institutional structures, modify water policies, improve delivery and on-farm systems, improve management of degraded soils, enhance water reuse, improve crop water management, and address rising energy prices. Greatly increased investments in irrigation infrastructure and economic incentives will be required throughout the world to enable these required increases in productivity that permit sustainable irrigated production while addressing the changes in the priorities for water use. A recent USDA agricultural water security listening session (Dobrowolski and O’Neill, 2005; O’Neill and Dobrowolski, 2005) developed a list of research and action areas over the next several years for maximizing the efficiency of water use by farmers, ranchers, and communities. These water issues included: ƒ enhancing supplies with new storage facilities; ƒ expanding existing infrastructure; ƒ funding for water reclamation and reuse; ƒ lowering water consumption by rural and urban users with new technologies; ƒ developing new technologies and systems for recycling and reusing degraded water; ƒ providing risk assessment and management for water scarcity; ƒ determining sociological and economic impacts of conservation technologies; ƒ evaluating the impacts of environmental degradation and subsidies for water and crops;

Design and Operation of Farm Irrigation Systems

17

ƒ determining the role of physical and paper water banks and progressive water rate structures and other market-based or incentive mechanisms; ƒ applying biotechnological improvements in water use efficiency; ƒ measuring the consequences of water conservation and reuse; ƒ responding to public health and environmental concerns of reused water and water treatment strategies to improve return flows from agriculture; and ƒ expanding decision making tools for water in agricultural, rural, and urbanizing watersheds including forecasting supply and shortage. 1.4.3 Productivity Challenges The formidable pressures on water resources make it certain that water will be the major natural resource issue of the 21st century (Seckler and Amarasinghe, 2000). There will be large economic and social pressures to reduce irrigation water use. This means that productivity per unit of water consumed must be much higher than current values to meet the high future demands for food, fiber, livestock feed, and biological alternatives to petroleum products. This is a major shift from the current emphasis on maximizing yield per unit area, and it will require a significant re-thinking of how and why irrigation is practiced. There may be small pockets of future growth, but the world’s irrigable land base is essentially developed. Some irrigated areas are actually declining in size and productivity due to waterlogging and soil salination, urban encroachment, soil erosion, and declining water tables. Combining a fully exploited land base with the growing competition for existing freshwater supplies along with the needs for increased agricultural production and the concurrent need for energy conservation will require that irrigators substantially increase efficiency and productivity per unit of water consumed. Postel (1999) estimated that water use efficiencies across the world were only about 40%, indicating that there are large opportunities for improvement. This applies to both agricultural and urban crop production (i.e., turfgrass). The greatest potential may be in developing countries where improved water management strategies and practices have the largest potential to increase production. Productivity of a specific crop is a function of maximizing application efficiencies, whereas improving productivity, in general, implies maximizing both efficiency and cropland net return. Thus, to improve the productivity of crops with any irrigation system we must consider many factors including the crop variety, plant water requirements, quantity and quality of the water supply, soil characteristics, topography, field size and shape, local climate, and a large number of economic concerns, such as labor requirements, available capital, and other resource costs. Many of these factors are interdependent, and it will be necessary to custom design a synergistic mix of strategies that fit the delivery systems, farming culture, crop water use patterns, soils, regional hydrology, and other unique environmental characteristics of an area. Clemmens and Allen (2005) present an excellent discussion of the environmental and economic trade-offs involved in implementing advanced irrigation practices that improve efficiencies. A critical link in improved productivity is the implementation of advanced scientific irrigation scheduling techniques, especially under deficit management. However, most growers lack sufficient economic incentives to implement the techniques. One of the greatest constraints to managing for enhanced productivity as well as protecting water quality is the inability of agricultural producers to control inputs in

18

Chapter 1 Introduction

ways that account for the variability in growing conditions across a field. Infiltration rates can vary between irrigation events as well as with location within the field. Wide variations in soil types, soil chemical properties, subsurface conditions, topography, drainage, insect/weed/disease problems, soil compaction, weather patterns, irrigation system operation and maintenance, and wind distortion of sprinkler patterns, as well as external factors such as herbicide drift, can cause yields and crop water use to vary across a field as well as across adjacent fields. These may also be impacted by tillage practices and crop rotations. Thus, it may be more advantageous to apply water uniformly over smaller portions of a field to minimize production differences and environmental consequences due to the variability of these numerous factors rather than managing the entire field as one management unit. The challenges and opportunities for irrigation equipment manufacturers, designers, researchers, managers, and growers will be immense and ultimately quite profitable. 1.4.4 Water Policy Issues As populations increase, competition for available water supplies will intensify. In some areas of the Western U.S., agricultural water users have some of the most senior water rights. As cities have grown, the consumptive components of water rights have been transferred from agriculture to the municipal sector through the willing sale and purchase of the water rights. Depending on the area and natural rainfall amounts, the farmer selling his water rights reduces the area irrigated or converts the non-irrigated land to dryland farming or pasture. Irrigated land could also be fallowed in alternate years or as part of various irrigated-dryland cropping rotations. Worldwide, water policies involve many entities, each behaving according to its set of rules and incentives. These entities include irrigators, landowners, irrigation districts, water user organizations, state or provincial water agencies, national ministries, development banks and organizations, private voluntary groups, engineering and consulting firms, politicians, voters, and taxpayers. The policies and rules of these entities, in most cases, do not encourage improved irrigation efficiency through design or operation. An estimated $33 billion annually worldwide in government subsidies tend to keep water prices artificially low and discouraging investments to conserve irrigation water supplies (Myers, 1997). In addition, many laws and regulations have been a barrier to marketing or transferring water, leading to inefficient water allocation and use. Similarly, lack of policies to regulate groundwater use has led to over-pumping and depletion of aquifers. There is a worldwide shortage of appropriate institutions to assist growers in managing water more effectively while reducing negative environmental impacts. The following sections provide examples of current policy issues and introduce potential improvements in water policies. For those desiring more information, these issues and more are covered in great detail in numerous publications by private and governmental groups, universities, and agencies that have conducted studies and developed plans to address complex water-related issues in the 21st century. For example, a comprehensive series of reports was prepared for the Western Water Policy Review Commission (established under the Western Water Policy Review Act of 1992 [PL 102-575, Title XXX]). These numerous reports present in-depth analyses of the water resource issues and demographics of every major river system in the West, and are summarized in “Water in the West: Challenge for the Next Century” (Western Water Policy Review Commission, 1998).

Design and Operation of Farm Irrigation Systems

19

1.4.4.1 Irrigation delivery management. More than 25 countries are changing the responsibility for managing irrigation systems from the central government to local groups (Vermillion, 1997). Most of these changes are driven as much by the need to cut government expenditures as by the desire to improve irrigation performance. By reducing government subsidies and oversight, however, the hope is that such management transfers will accomplish both goals. The largest management shift of this type in recent times occurred in Mexico, where the management of 2.8 million ha of publicly irrigated land has been turned over to farmer organizations. With the accompanying large reduction in government subsidies, water fees have risen to cover costs. The irrigation districts are now about 80% financially self-sufficient, up from 37% prior to the transfer. The cost of irrigation water, although higher, is still about 5% of total production costs, a typical value for irrigated agriculture. It has yet to be determined whether local management has promoted greater equity in water distribution and more efficient water use (Johnson, 1997). 1.4.4.2 Water pricing. The inability to agree on the economic and social values of healthy freshwater ecosystems and water quality, as well as flood control, recreation, irrigated agriculture, and tourism, has lead to conflicts in many areas of the world. There is a wide spectrum of options, each with its own range of consequences for various sectors and interests. The price of water can be difficult to alter because of the many diverse viewpoints. In many scenarios, charging the full cost for water without changing other institutional and economic structures would put most irrigators out of business. However, charging little or nothing, which is the situation in many irrigation projects, is a clear invitation to waste water and increases the potential for conflict. One option gaining popularity is a tiered pricing structure, which provides strong incentives to conserve water to avoid higher rates. For example, irrigators might be charged the low rate they are accustomed to paying for up to, say, 60% of their average past water use, a significantly higher price for the next 20%, and the full (high) cost for the last 20%. Any water used above the average use would be at full cost plus a penalty. The Broadview Water District in California introduced a tiered water pricing structure in the 1990s. The district determined the average amount of water used in previous years for each crop. Irrigators were charged the customary rate for up to 90% of the average water need. Any water deliveries above 90% were 2.5 times higher in price. Even though they were still paying much less than the real water cost, the irrigators had incentive to conserve. Depending upon the crop, water savings were from 9% to 31% while crop yields were unaffected or increased (California Department of Water Resources, 1998). Another example is Israel, which currently has a price structure where 65% of full evapotranspiration (ET) is provided at a reduced price, but the price increases exponentially for any additional water. 1.4.4.3 Water marketing. Removing barriers to water marketing can also lead to a more equitable distribution of water and improved water use efficiency. The ability to sell some water gives irrigators an incentive to conserve water so they can profit from the sale of the resulting saved water. Formal water markets work only where farmers have legally enforceable rights to their water, and where those rights can be sold. Australia, Chile, Spain, Mexico, and many states in the Western U.S. now have laws and

20

Chapter 1 Introduction

policies that permit water markets. One important advantage is that this mechanism is voluntary. As with pricing, a variety of marketing options exist. Water rights owners can sell their water on a seasonal basis or they could enter into a multi-year contract. They could also sell their water rights to another user, in which case the legal water entitlement is transferred permanently. 1.4.4.4 Water transfers. Traditionally, there have always been limited water transfers within irrigated agriculture at the farm and project level, normally from field crops to horticultural crops of higher value. Economic decisions have also caused agricultural water to be transferred to municipal and industrial uses in many areas. However, as demands on the world’s scarce freshwater supplies increase, water transfers from agriculture to other sectors will become more and more common and, in fact, some may become mandated by judicial and legislative processes. Artificial groundwater recharge as a form of water banking, as well as the treatment of degraded aquifer and soil waters for later reuse, are all components in water transfer considerations and policies. Mutual agreements have been reached between municipal and agricultural users that allow an irrigation district or farmers to continue irrigating, but require they make improvements that will conserve water, i.e., reduce the consumptive use component of water diverted. The cost of the improvements is borne by the municipal user. Site conditions often will determine the effectiveness of improvements in reducing water consumption. For example, lining a canal whose seepage normally returns to the river from which the water was originally diverted may not result in much, if any, reduction in water consumed in the river basin. An example of a mutual agreement to conserve irrigation water that could be used by a municipality was reached in 1988-1989 between the California Metropolitan Water District (MWD) and the Imperial Irrigation District (IID). The IID is located in the Imperial Valley of Southern California and diverts water by gravity from the Colorado River. Surface and subsurface flows from irrigated lands in the IID do not return to the Colorado River from which water was diverted, but instead flow to the Salton Sea, which lies about 70 meters below sea level. Improvements were agreed upon that would reduce flows to the Salton Sea, but still enable irrigated farming to continue. The improvements were estimated to conserve about 123 million cubic meters of water annually within IID with the saved water available to the MWD for municipal use (MacDonnell and Rice, 1994; National Research Council, 1996). The IID-MWD conservation program involved lining canals within the IID, improved flow-monitoring structures, installation of non-leak gates, prevention or recovery of canal spills, installation of regulating reservoirs and seepage recovery systems, and system automation. Implied in this agreement is that the IID would reduce its net annual diversion of Colorado River water at Imperial Dam by the amount conserved and the MWD would increase its diversion upstream at Parker Dam by a like amount. The MWD benefits because the cost of the conserved water is less than $0.10 per cubic meter, much lower than its best option for a new supply. The IID benefits from the cash payments and an upgraded irrigation system, and no cropland is taken out of production because the water transferred is generated through conservation (Gomez and Steding, 1998). In another transfer agreement with the MWD, however, farmers were required to remove irrigated land from production. In 1992, the MWD entered into an agreement

Design and Operation of Farm Irrigation Systems

21

with the Palo Verde Irrigation District located on the west side of the Colorado River in Southern California. The agreement stipulated that farmers who laid fallow a portion of their cropland for two years would be paid $3,000 for each hectare left unplanted. More than 8000 ha were left fallow by 63 farmers. The equivalent of about 10% of the MWD’s yearly water deliveries were then transferred and stored in federal reservoirs on the Colorado River to be used anytime before 2000 (Loh and Steding, 1996). There are also some associated environmental agreements being developed including a three-state, 50-year agreement covering restoration of 3300 ha of native habitat along the Colorado River from Hoover Dam to the Mexican border. A third transfer agreement, entered into in 2002, provides mechanisms for transferring water from agricultural to urban users, and serves as the basis for California to settle nearly seven decades of disputes among its water agencies. It involves the IID, the MWD, the San Diego County Water Authority, and the Coachella Valley Water District. Implementation of this 75-year landmark agreement also requires California to meet specific benchmarks for receiving Colorado River water (surplus and otherwise) because the state has to reduce its annual Colorado River diversions from 6.4 billion cubic meters to 5.4 billion cubic meters of water over a 15-year period. The agreement combines temporary fallowing of irrigated lands and the transfer of about 1.2 billion cubic meters while the IID implements on-farm and system water conservation measures paid for by the urban water districts. San Diego agreed to a payment schedule for the water transferred plus $20 million to help cover socioeconomic impacts to the local Imperial Valley communities and landowners over the 15-year period as a result of the transfers. This is in addition to the land management, crop rotation, and water supply transfer agreements with the Palo Verde Irrigation District mentioned above. So far, few countries have the necessary institutional and incentive structures to guide water competition. In Chile, where government policies encourage water marketing, negative impacts seem to be minimal, primarily because farmers have only sold small portions of their water rights to cities that have funded upgrading of existing irrigation systems and operational procedures. 1.4.4.5 Economics and incentives. Both rainfed and irrigated agriculture generally provide only a marginal rate of return to growers in today’s world markets. Costs of production are rising while crop prices are remaining static, and many growers need subsidy payments to remain viable despite significant improvements in farming efficiencies. Thus, many producers currently cannot afford to make expensive improvements to enhance productivity. From an economic perspective, many advanced irrigation methodologies have high initial capital costs, which add increased production costs including energy, management, agrochemicals, and land preparation. Researchers and action agencies need to find ways to reduce all inputs, including management time, for growers to remain competitive in a world market. Economists and politicians must find ways to move world crop prices upward to where they truly reflect the increased costs of production, including higher water prices. It is expected that a considerable amount of the cost of any new agricultural infrastructure and field improvements would be privately, rather than publicly, financed. Private development of an intensive, irrigated greenhouse culture will no doubt ex-

22

Chapter 1 Introduction

pand greatly. Individuals and farmer-owned irrigation districts will be asked to fund upgrades to existing facilities and systems, but it should not be a one-way street. Recent years have seen governmental investments in irrigation research and infrastructure reduced significantly. However, it is clear that society will have to change its current attitude of minimizing investments in irrigated agriculture, and instead make substantial additional investments towards improving irrigation infrastructure and management to address increased water demands. Innovative policies must reflect the need and mechanisms for equitably funding these investments, and urban users will need to be much more involved in this process than has generally been the case. This is starting to become apparent in Southern California as they attempt to address complex regional water issues. Economic and social policies that include incentives need to be implemented to motivate farmers to reduce negative externalities and consider opportunity costs when choosing irrigation and drainage strategies. Incentives can be either negative or positive, but should be directed both at inputs in agricultural production as well as effluents such as salt, silt, nutrients, and other constituents in surface runoff and deep percolation. They could include collaborative water markets, payments to irrigators for achieving higher efficiencies, higher costs for water per unit area, service charges for each delivery time, reductions in water rights to more closely match ET, or other ways to encourage profitable deficit irrigation strategies. However, policies and incentives have to be economically and culturally acceptable to growers and be accompanied by realistic programs to upgrade and support improved technologies and management. Positive incentives could include payments to growers who meet or exceed targeted irrigation efficiency levels. Payments could be in the form of funds to improve irrigation systems and monitoring equipment, but must also include ways to offset the increased costs of management and labor associated with the improved efficiency. Increasing the actual prices farmers receive for their products by changing governmental policies would also provide incentives and the means for them to improve production, management, and environmentally beneficial practices in the face of declining water availability and rising energy costs. Another incentive could allow landowners to move water from poorly producing soils (whether due to salinity, erosion, or just natural causes) to more productive areas that may be outside an irrigation district’s boundaries. Some states, such as Washington, are allowing water spreading in certain areas where water saved by converting to low water use crops or deficit irrigation can be moved to previously nonirrigated lands as long as total historical ET is not increased. Some grower and societal incentives to implement improved irrigation technologies at the irrigation district and farm scales include: reduced labor requirements, lower costs for treating water by reducing the volumes treated, lower costs for pumping water, reduced costs for added distribution capacity in an area of growth, less leaching of fertilizers and chemicals and degradation of groundwater, and sustained flows in segments of streams bypassed by irrigation diversions. These savings and potential environmental benefits accrue both to the irrigation manager and, ultimately, the general populace. In the future, both groups will likely perceive these as necessary incentives. Increasing crop productivity while reducing the amount of applied water depends on the ability of a particular type of irrigation system and the managerial skills of the operator to correctly implement the water-saving practices and techniques. Frequently,

Design and Operation of Farm Irrigation Systems

23

the controlling factors are the knowledge base of the grower and the existence of incentives to implement the improved practices. Thus, incentives should also reward higher management levels. A concern is that skilled labor for irrigation is becoming more and more limited and many are unwilling to perform the required activities and technological tasks. In addition, the average farming population age in the U.S. is in the 50s, and this age group is less likely than younger farmers to make major improvements or adopt new technologies without substantial incentives.

1.5 FUTURE DIRECTIONS The discussion above describes standard practices that can be done with current institutional and technological systems. The next step is to look into the future with all of its uncertainty. To aid in that effort, the National Research Council (1996) developed this list of likely future directions for irrigation in the U.S.: ƒ Irrigation will continue to play an important role in the U.S. and the world for the foreseeable future, although there will certainly be changes in its character, methods, and scope. ƒ The total irrigated area will likely decline, but the value of irrigated agriculture will remain about the same because of shifts to crops of higher value. ƒ The amount of water dedicated to irrigated agriculture will decline as societal values change and competition for water increases. ƒ A major factor in the sustainability of U.S. irrigation will be determined by our ability to compete in global markets. ƒ Under-financed irrigation operations or those with less-skilled managers will tend to decline in number. ƒ Previously, irrigation meant irrigation for agriculture. During the past 25 years irrigation has become an important part of the turf industry, and irrigation for urban landscaping and golf courses is growing steadily as urban populations increase. ƒ With time, increasing amounts of water will be removed from agriculture to satisfy environmental goals. In conjunction with this, there will be increasing pressures to reduce environmental degradation associated with irrigation. Substantial research and educational challenges must still be addressed regarding water availability, quantity, and quality, water use, and water institutions (National Research Council, 2001, 2004). Changes in policy and incentives will clearly become necessary. The following sections examine some potential issues and solutions to agricultural water security issues. Urban water users will have a similar set of challenges to reduce and modify water consumption. 1.5.1 Need for Innovation Irrigation has been practiced for more than 6000 years, but more innovation has occurred in this arena in the last 100 years than in all of the preceding centuries. Almost every aspect of irrigation has seen significant innovation: diversion works, pumping, filtration, conveyance, distribution, application methods, drainage, power sources, scheduling, fertigation, chemigation, erosion control, land grading, soil water measurement, and water conservation. Major future improvements in water saving will be realized through innovative design and operation of integrated irrigation systems for both agricultural and urban set-

24

Chapter 1 Introduction

tings. It is obvious that all these technologies will have to continue to be improved and implemented to better manage energy, water, and soil resources. Novel irrigation techniques and management systems will be necessary to increase the costeffectiveness of crop production, improve water quality, improve water reuse capabilities, reduce soil erosion, and reduce energy requirements while enhancing and sustaining crop production and water use efficiency. In addition, innovative water polices and institutional structures must evolve and foster emerging irrigation technologies. Irrigation is a valuable technology, rooted in ancient tradition, and has proven to be dynamic and flexible. However, new and improved strategies and practices are needed to reduce surface and groundwater contamination from agricultural lands, conserve water and energy, and sustain food production for strategic, economic, and social benefits. Systems must be designed and managed to minimize health hazards due to chemical applications of fertilizers and pesticides as well as to minimize insect infestation and parasitic diseases, such as the West Nile virus and malaria. The effects of water conservation and reuse technologies on recreation, tourism, wetlands, and aquatic ecosystems must be assessed and balanced with other societal needs. Future irrigators will often be operating under various managed crop water deficit scenarios. Increasing crop productivity while reducing the amount of applied water implies that producers will often be managing irrigations under severe to moderate soil water deficit conditions during part or all of the growing season. Techniques such as partial root zone drying and regulated deficit irrigation will be more and more common on tree and vine crops as well as many annual crops (Chalmers et al., 1986; Fereres et al, 2003). Techniques such as fallowing of irrigated fields in alternate years to conserve water need to be investigated as to potential water savings and reduced agrochemical use. Water reuse and treatment of impaired waters will be part of agricultural water security. Innovative approaches to groundwater recharge using treated and excess surface waters for later withdrawals by a multitude of users will be an essential part of future water resource programs.. The following brief sections present more details on some of the issues that agriculture will have to implement to address water security issues. These measures will include: modernizing irrigation delivery systems and on-farm systems, improved levels of management, strategies for local water supply enhancement, and biotechnological advances in crop breeding and selection. 1.5.2 Modernizing Delivery and On-Farm Systems Traditional approaches to modernizing irrigation projects have focused on minimizing water loss during delivery and maximizing field application efficiencies. These are necessary first steps, but future water delivery systems and application techniques must be modified to enhance grower flexibility in managing rates, irrigation frequencies, and durations, as well as reduce water evaporation and other losses. Small, distributed internal regulation reservoirs, closed-conduit systems to reduce evaporation and leave unused water in the distribution system, extensive automated water-level controls, accurate automated flow measurement, and improved ways to reduce weed growth on canal and lateral banks to minimize non-beneficial ET are all potential means of improving water delivery efficiencies. Some of these features are just now beginning to be implemented in a few modernized irrigation projects.

Design and Operation of Farm Irrigation Systems

25

To maximize the potential of existing and emerging technologies, irrigators must have the flexibility to manage rate, frequency, and duration of their water supplies. Thus, the delivery system and the farm must be considered as one integrated unit with two parts rather than two independent systems. With the imperative need to implement the agreements and mandates discussed previously, the Imperial Valley in California will be a proving ground for these concepts over the next couple of decades, and there is much to learn. In this case, the delivery system will provide irrigation water to satisfy the specific field condition (i.e., rate, frequency, and duration) that is calibrated for each crop and irrigation condition. This will require extensive canal automation and on-farm monitoring as well as economic incentives for achieving better water use productivity. Positive payments for achieving a certain target efficiency or tailwater levels rather than solely raising water costs are anticipated. The following identifies some of the potential areas where innovation is likely to improve delivery and on-farm systems in the 21st century: ƒ Computers and wireless control systems will play an ever-expanding role. Cell phone and satellite communications and internet technologies will likely play an increasingly major part in management of irrigation systems. Feedback control technologies for automating canal operations, surface and pressurized systems, and drainage systems must be developed, tested, and supported by incentives. ƒ On-farm systems will benefit from advanced technologies, such as precision irrigation, site-specific management, remote sensing, within-field real-time sensor systems, and decision support systems, which collectively have great potential to facilitate reduction of water quantity and quality problems in irrigated agriculture. ƒ The use of real-time irrigation scheduling techniques (sensor-based) and sitespecific precision applications of water through center pivot machines and microirrigation are the next steps in the evolution of those technologies. ƒ It will be necessary to expand modern crop production technologies to less productive rainfed and irrigated lands characterized by poor soils, low and unstable rainfall, steep slopes, and short growing seasons to increase food production and stimulate economic growth. Novel approaches will be needed to address these areas. ƒ Microirrigation with its many variations must be made less expensive before most growers will be able to adopt and utilize these technologies, especially in developing countries. Some localized efforts are ongoing and one company is manufacturing tubing at relatively low cost, but these innovative efforts and technologies need to be extended to other areas. ƒ For developing countries, innovative research and extension education is needed to provide and implement simple but efficient low-cost methods of irrigation (e.g., pitcher irrigation) to make them easy to operate, suitable for the crop, and acceptable to growers. There is also a huge need for low-lift pumps that are inexpensive to buy and operate in these areas. Some of this is already being done on relatively small scales, but there is much room for innovation. ƒ Surface irrigation methods can be made more efficient using surge flow, deadlevel basins, and other techniques for more uniform infiltration along the length of the field. Properly designed and operated level basins eliminate runoff, can be quite efficient and uniform, and are relatively inexpensive to construct and operate. However, considerable investment for delivery system improvements, as

26

Chapter 1 Introduction

well as sensor feedback controls and automation for both the delivery and application systems, is needed to fully realize the potential water savings. ƒ Urban and agricultural irrigators will be the primary users of degraded waters. New approaches and techniques will be required to safely minimize detrimental effects while maintaining production goals. ƒ Farm- and district-level drainage systems will require improved design, evaluation and simulation models defining the physical limits. Automated control systems will assist in providing a more uniform soil water environment for plant growth to improve productivity and minimize the volumes of drainage waters requiring treatment, especially in arid areas. ƒ Water table elevations can be managed to permit subirrigation, if the groundwater is relatively shallow and of suitable quality, by controlling water tables or inducing water tables with irrigation applications. Subirrigation has been practiced successfully in climate regions ranging from humid to arid. Using the effluent from deeper subsurface drainage systems as a source of irrigation water has proven effective in many regions of the world. The biggest concern is the quality (i.e., salinity) of the drainage system effluent. ƒ There is still no reliable, inexpensive electronic soil water sensor that matches or exceeds the accuracy and repeatability of neutron scattering devices. Innovative development of such sensors is essential for water management, particularly under deficit conditions. ƒ Many of the needed and evolving technologies will require stand-alone, spatially distributed electrical power to be feasible. Controllers, monitoring equipment, and communications devices must be low power consumers. Photovoltaic, wind turbine, and storage systems will need to be developed and implemented at low cost at the farm or field level. ƒ Economies of scale have led to large field sizes for irrigated production in many areas, and engineers have been very successful in designing pressurized irrigation systems that apply water quite uniformly over these fields. However, the challenge over the next 50 years for the irrigation industry and designers is to develop highly efficient systems that are also suitable for small-scale farms and provide the necessary extension education to equip the farmers with the skills to run them. ƒ Although widely variable, it is estimated that 1% of the global water storage capacity in reservoirs is lost each year to sedimentation (Palimeri, 1998), decreasing the ability to store water. Innovative methods to reduce erosion at the watershed and basin scale will be needed to increase the life of reservoirs for storage, flood control, and recreation uses. Innovation will be required to enable adoption of more efficient irrigation methods. For example, high-frequency drip irrigation and other microirrigation methods have been shown to increase the yield and quality of fruit and vegetable crops through reduced water and nutrient stresses. Tied to an effective soil water monitoring program, good design, and appropriate management practices, microirrigation can be 95% efficient or better. A modification of center pivot irrigation called “low energy precision application” or LEPA has been found to be 95% efficient as well. However, microirrigation and LEPA irrigation are being used on less than 1% of irrigated lands worldwide.

Design and Operation of Farm Irrigation Systems

27

1.5.3 Management Inherent in the evolution of on-farm management discussed above will be the integration of irrigation, fertilizer, and pest management strategies into systems that optimize total management practices for temporal and spatial variability. This should result in substantial labor, water, and energy savings and minimize losses to the groundwater. Improved irrigation technologies usually result in reduced labor requirements, but require expanded management. However, most producers who irrigate do not have sufficient time to properly manage all their critical inputs. At certain times of the year, a producer’s time is extremely valuable in terms of net returns from the crop, and irrigation water management is often not as high a priority as other concurrent cultural factors. Thus, decision support aids must also be developed that improve the producer’s ability to implement decisions quickly and easily because climatic variations and pest outbreaks require precisely timed water and chemical applications on a daily and seasonal basis. The decision support process must also provide accurate predictions of application efficiencies and uniformities to improve management flexibility. Ensuring the success of irrigated farming enterprises will require the development of reliable and more-timely information on field and plant status to support the decision-making processes. Current plant models capable of predicting the physiological needs of a crop over space and time tend to be complex and impractical for real-time on-farm management. Furthermore, most of these models are point models that are not sensitive enough to adequately predict site-specific plant needs across a field in a timely fashion. Simpler, more appropriate models might be used but will likely need frequent updating via automated, field-based sensor systems to readjust model variables to ensure reasonable tracking and spatial predictions of field conditions. A more focused approach will be required for the development of spatial and temporal management strategies that address site-specific crop water, nutrient, and pest management requirements, and irrigation scheduling in real time. The ultimate goal should be to integrate controls, sensor systems, plant and physical models, and other techniques to provide workable solutions that reduce time requirements for busy decision makers while improving their management capacity. Self-propelled irrigation systems, such as center pivots and linear moves, are particularly amenable to site-specific approaches because of their current level of automation and large area coverage with a single pipe lateral. These technologies hold much promise for spatially varying water and agrochemical applications to match differences in irrigation, nutrient, and pesticide requirements throughout the field. This could potentially increase productivity and minimize adverse water quality impacts. By aligning irrigation water applications with variable water requirements in the field, total water use may be reduced, decreasing deep percolation and surface runoff. This also suggests that water-soluble fertilizers and other agrochemicals can be effectively applied spatially through the irrigation system to match changing conditions across a field. However, there is a need to develop more efficient methods of applying crop amendments (e.g., nutrients, pesticides) with irrigation systems that will reduce agrochemical usage, improve profit margins, and reduce environmental impacts. In some cases, it may be necessary to schedule and irrigate on a plant-by-plant basis. This technology is available today, but is not yet economical or practical. Research is

28

Chapter 1 Introduction

needed to give farmers confidence that the use of these technologies is practical and potentially valuable in improving irrigated crop production. Much effort has gone into ET research in the past several decades, and it is one of the most accurate estimates available to irrigation delivery and on-farm managers as long as plants are at relatively low stress levels. However, much additional information is still required on the yield and crop quality affect at various ET levels at different growth stages of both major and minor crops that are in managed soil water deficit conditions, perhaps in combination with increased soil salinity levels. The long term effect of deficits on perennial crops also needs more research to ensure sustainability. Critical information is lacking on actual ET under irrigation systems that are being used for environmental modification to protect the crop from cold temperatures or excessive heat. In addition, research on irrigation management and water use requirements is lacking for intercropping production systems. More research is also needed on how to truly address spatial variability of ET across large fields. This information will be extremely critical to future determinations of both agricultural and urban water rights, water banking, and governmental allocations. Remote sensing of plant and soil status using integrated satellite, aerial, and field level plant- and soil-based sensor systems is another way of providing information, but it also needs further development to improve spatial-temporal modeling and on-farm management as well as irrigation district operations. Better systems and methods capable of precisely measuring specific plant parameters (e.g., nutrient status, water status, disease, and competing weeds) on a timely basis are needed to improve crop modeling and thus improve within-season management. Real-time, on-the-go irrigation scheduling could be very effective in improving water management when based on distributed networks of farm-level microclimate and soil water sensor stations that feed into a microprocessor control system to manage irrigations based on rules established by the producer. This effort must be supported by expanded agricultural weather networks with a greater spatial density and growerfriendly information delivery systems for scheduling irrigations combined with pest management and marketing information. Input from distributed weather networks must be integrated with other information sources to effectively contribute to on-farm and irrigation district decision support processes. In addition to irrigation, self-propelled irrigation systems also provide an outstanding platform on which to mount sensors for real-time monitoring of plant and soil conditions, which in turn provide input to the control system for optimal environmental benefits. Similar sensor technologies tied to a global positioning system can also be mounted on-farm equipment so that site-specific information is collected each time the grower is in the field. Early detection of diseases, weeds, insects, and even nutrient deficiencies would allow more economical spot treatments of small areas within a field. 1.5.4 Strategies for Local Water Supply Enhancement Techniques to capture more water in either surface or subsurface impoundments and aquifers will only be briefly mentioned because they are beyond the scope of this monograph. However, it is worthwhile to introduce a few practical measures to increase local water availability on the farm and perhaps stimulate additional efforts in this critical area. Unger and Howell (1999) present a much more complete review of

Design and Operation of Farm Irrigation Systems

29

the pros and cons of various dryland and irrigated techniques available for local water conservation. Strategies to be considered as a part of these activities include irrigation system design for higher efficiency, successful treatment and reuse of degraded waters, reducing evaporation losses, site-specific applications, managed-deficit irrigations of tree and vine crops, and techniques to minimize leaching. However, they will not be repeated as many of these have already been introduced and some are discussed in more detail in other chapters. Perhaps the most neglected local water source is the capture, diversion, and storage of precipitation (both rainfall and snowmelt) where it falls in the field. The goal is to capture water falling on croplands and supplement this with water from nearby noncropped areas. This is often referred to as “water harvesting” and is used around the world to supplement irrigation by improving water availability. It can be an effective water conservation tool, especially in arid and semi-arid regions (Unger and Howell, 1999). These practices may also benefit soil fertility by capturing dissolved nutrients, and may also provide broader environmental benefits through reduced soil erosion. Where feasible, small reservoirs or storage tanks that minimize evaporation and seepage losses could be constructed at both regional and farm scales to capture runoff from storms for later use. Where groundwater is a viable water supply, techniques to recharge aquifers, such as recharge wells and percolation ponds, should be considered to increase supplies. Substituting a crop with low consumption for a crop with higher water consumption or switching to crops with higher economic benefit or productivity per unit of water consumed by ET can provide water savings. Reallocation of water from low-value crops to higher-value crops can also increase the economic productivity of water. Conservation tillage practices (e.g., strip till, ridge till) that maintain surface crop residues or mulches to reduce evaporation can increase the share of rainfall that goes to infiltration and ET. Contour plowing, which has been promoted as a soil-preserving technique, can detain and infiltrate a higher share of the precipitation. Techniques such as precision leveling can capture a higher percentage of effective rainfall and improve soil water management. Furrow diking in sprinkler irrigated areas (e.g., LEPA) to hold water where it falls is beneficial. The use of mulches for weed control also reduces non-beneficial ET and soil evaporation. Drip irrigation technologies can conserve water by greatly reducing soil evaporation and maximizing water use efficiencies. These strategies could also incorporate the use of alternative cropping systems including winter crops and deeprooted cultivars that maximize the utilization of soil water and nutrients. Weed control to reduce non-beneficial water use is critical. Another technique employs a series of dikes or terraces placed on the contour to capture precipitation as it runs off and convey it to cropland. This is currently used in certain areas of Australia where some storm water runoff is captured at the low ends of large properties where large pumps convey the collected water into large aboveground tanks for livestock, irrigation, and other uses. 1.5.5 Cultural/Crop/Plant Selection Factors In combination with molecular techniques and processes, plant breeders and molecular biologists will be selecting plants that use less water with a concurrent toler-

30

Chapter 1 Introduction

ance to heat stress (less transpirational cooling) and greater tolerance to salinity, diseases, weeds, and insect pests while producing high-quality crops. The problem is that no single gene or series of genes has been specifically identified with increasing yields (Mann, 1999). Comparatively small advances in disease and pest resistance as well as drought resistance are forecasted. Crop breeding will have greater focus on more efficient water use by selecting for optimal growing season lengths and harvest dates that take maximum advantage of rainfall timing at critical growth stages for each region. These activities must be accompanied by multidisciplinary, collaborative research into soil management, pest control (by tillage, cultural, biological, and chemical means), green manures, and the yield and quality responses of the new cultivars to both full and deficit irrigation amounts, increased salinity levels, degraded reuse water from a variety of sources, and the effects of climate change. Intercropping, which is a cultural practice in some irrigated areas in the world, may be advantageous to help satisfy future production expectations if expanded into rainfed and irrigated areas that are already highly productive. However, because varieties that can tolerate water and nutrient stress must be developed as well as production strategies for interseeding and for the use of short season crops to facilitate production of multiple crops per year in temperate to humid areas, substantial progress in these areas is expected to be a lengthy process.

REFERENCES California Department of Water Resources. 1998. Drainage management in the San Joaquin Valley: A status report. Sacramento, Calif.: San Joaquin Valley Drainage Implementation Program. Chalmers, D. J., G. Burge, P. H. Jerie, and P. D. Mitchell. 1986. The mechanism of regulation of Bartlett pear fruit and vegetative growth by irrigation withholding and regulated deficit irrigation. J. American Soc. Hort. Sci. 111: 904-907. Clemmens, A., and R. Allen. 2005. Impact of agricultural water conservation on water availability. In Proc. of the World Water and Environmental Resources Congress. Reston, Va.: American Society of Civil Engineers. Dobrowolski, J. P., and M. P. O’Neill, eds. 2005. Agricultural water security listening session final report. Washington D.C.: USDA-REE. FAOSTAT. 2005. FAOSTAT statistical database. Rome, Italy: Food and Agriculture Organization of the United Nations. Available at: www.faostat.fao.org. Fereres, E., D. A. Goldhammer, and L. R. Parsons. 2003. Irrigation water management of horticultural crops. HortSci. 38: 1036-1043. Gleick, P. 1998. The World’s Water 1998-1999: The Biennial Report on Freshwater Resources. Washington, D.C.: Island Press. Gollehon, N., 2002. Irrigation and water use: Questions and answers. USDA-Economic Research Service. Available at: www.ers.usda.gov/Briefing/WaterUse/Questions /qa3.htm. Gomez, S., and A. Steding. 1998. California Water Transfers: An Evaluation of the Economic Framework and a Spatial Analysis of the Potential Impacts. Oakland, Calif.: Pacific Institute. Hutson, S. S., N. L. Barber, J. F. Kenny, K. S. Linsey, D. S. Lumia, and M. A. Maupin. 2004. Estimated water use in the United States in 2000. U.S. Geological

Design and Operation of Farm Irrigation Systems

31

Survey Circular 1268. Reston, Va.: U.S. Department of the Interior, Geological Survey. Intergovernmental Panel on Climate Change. 2001. Climate change 2001: Synthesis report. The IPCC 3rd assessment report. Summary for policymakers. New York, N.Y.: Cambridge Univ. Press. Available at: www.ipcc.ch/pub/pub.html. Johnson, S. H. 1997. Irrigation management transfer: Decentralizing public irrigation in Mexico. Water International 22: 159-167. Jones, W. I. 1995. The World Bank and Irrigation. Washington, D.C.: World Bank. Loh, P., and A. Steding. 1996. The Palo Verde Test Land Fallowing Program: A Model for Future California Water Transfers? Oakland, Calif.: Pacific Institute. MacDonnell, L., and T. Rice. 1994. Moving agricultural water to cities: The search for smarter approaches. Hastings West-Northwest J. 2: 27-54. Mann, C. C. 1999. Future food: Crop scientists seek a new revolution. Science 283(5400): 310-314. Milesi, C., S. W. Running, C. D. Elvidge, J. B. Dietz, B. T. Tuttle, R. R. Nemani. 2005. Mapping and modeling the biogeochemical cycling of turf grasses in the United States. Environmental Mgmt. 36(3): 426-438. Myers, N. 1997. Perverse studies: Their nature, scale, and impacts. Chicago, Ill.: Report to MacArthur Foundation. NRC (National Research Council). 1989. Irrigation-Induced Water Quality Problems. Washington, D.C.: National Academy Press. NRC (National Research Council). 1996. A New Era for Irrigation. Washington, D.C.: National Academy Press. NRC (National Research Council). 2001. Envisioning the Agenda for Water Resources Research in the Twenty-First Century. Washington, D.C.: National Academy Press. NRC (National Research Council). 2004. Confronting the Nation’s Water Problems: The Role of Research. Washington, D.C.: National Academy Press. Newson, M. 1992. Land, Water and Development: River Basin Systems and their Sustainable Management. London, UK: Routledge. O’Neill, M. P., and J. P. Dobrowolski. 2005. CRREES Agricultural Water Security White Paper. Washington, D.C.: USDA-REE CSREES. Palmieri, A., 1998. Reservoir sedimentation and the sustainability of dams. 1998 In Proc. of the World Bank Water Week Conference. Annapolis, Md.: World Bank. Postel, S. 1999. Pillar of Sand: Can the Irrigation Miracle Last? New York, N.Y.: Worldwatch Books, W. W. Norton & Co. Postel, S. 2000. Redesigning irrigated agriculture. In Proc. 4th Nat’l Irrigation Symp., 1-12. R. G. Evans, B. L. Benham, and T. P. Trooien, eds. St. Joseph, Mich.: ASAE. San Joaquin Valley Drainage Program. 1990. Fish and Wildlife Resources and Agricultural (0.1) Drainage in the San Joaquin Valley, California. Seckler, D., and A. Amarasinghe. 2000. Chapter 3: Water supply and demand, 1995 to 2025: Water scarcity and major issues. In World Water Vision. Battaramulla, Sri Lanka: International Water Management Institute. Available at: www.iwmi.cgiar. org /pubs/WWVisn/WWSDOpen.htm. Tanji, K. K. 1990. Agricultural salinity problems. In Agricultural Salinity Assessment and Management. K. Tanji, ed. New York, N.Y.: American. Soc. Civil Engineers. Unger, P. W., and T. A. Howell. 1999. Agricultural water conservation—A global perspective. J. Crop Production 2(4): 1-36.

32

Chapter 1 Introduction

USDA-NASS (U.S. Department of Agriculture National Agricultural Statistics Service). 2004. 2002 Census of Agriculture. Vol. 3: Special studies. Part 1: Farm and ranch irrigation survey (2003). Available at: www.nass.usda.gov/census/census02/fris/ fris03.htm. U.S. Department of Commerce. 1978. 1974 Census of Agriculture. Chapter I: Irrigation of Agricultural Lands, I-1 to I-63. Part 9: Irrigation and Drainage on Farms. Washington, D.C.: Bureau of the Census. U.S. Department of Commerce. 1983. 1982 Census of Agriculture. Vol. 1, Part 51: United States Summary and State Data. Washington, D.C.: Bureau of the Census. Vermillion, D. L. 1997. Impacts of Irrigation Management Transfer. Colombo, Sri Lanka: Irrigation Water Management Institute. Western Water Policy Review Commission. 1998. Water in the West: Challenge for the Next Century. Springfield, Va.: NTIS. Wilcove, D. S., and M. J. Bean. 1994. The Big Kill: Declining Biodiversity in America’s Lakes and Rivers. New York, N.Y.: Environmental Defense Fund. World Bank. 1991. India Irrigation Sector Review, Vol. 2. Washington, D.C.: World Bank.

CHAPTER 2

SUSTAINABLE AND PRODUCTIVE IRRIGATED AGRICULTURE Marvin E. Jensen (USDA-ARS, Fort Collins, Colorado) Abstract. A renewable water supply is required for sustained irrigated agriculture. Irrigated agriculture can remain productive indefinitely with assurance of water supplies and infrastructure to deliver water as needed for crop production. Emerging issues facing designers of irrigation systems are competition for limited and declining water resources, climate variability, effects of leasing water during drought years on annualized water costs, environmental concerns, refined irrigation scheduling, and drainage and salinity control issues. Designers of irrigation systems must consider the role of irrigation in food production, and the productivity of water used for irrigation. Keywords. Climate-based irrigation scheduling, Deficit irrigation, Design challenges, Environmental concerns, Sustainable development, Water, Water accounting.

2.1 INTRODUCTION 2.1.1 Scope of the Chapter Since the first edition of this monograph (printed in 1980), numerous papers and books have been published on irrigation systems and practices, trends in food production, climate and water requirements, irrigation experiences in arid, semiarid, subhumid, and humid areas, and various beneficial uses of irrigation water. Some of these subjects are covered in detail in other chapters. This chapter highlights recent trends and emerging issues facing designers and operators of farm irrigation systems. Sustainable irrigated agriculture depends on a renewable source of water and adequate drainage for control of soil salinity. Increasing competition for limited water supplies will require increasing attention to the water productivity, i.e., the production of the marketable component of a crop per unit of water consumed. Water consumed is that fraction of water applied that is evaporated, transpired, or combined within the crop and is no longer available for other uses. The fraction of water applied that is not consumed, i.e., irrigation return flows, and other environmental concerns will need to be addressed by designers of irrigation systems. Maximizing crop yields will no longer be the main objective of irrigation; more attention will be on maximizing net benefits of irrigation.

34

Chapter 2 Sustainable and Productive Irrigated Agriculture

2.1.2 Water, Plant Growth, Irrigation, and Crop Production Water is essential for plant growth. Seeds need water to germinate, most soils must be moist for seedlings to emerge, and many plant growth functions require water. Water provides the transport mechanism within plants for nutrients and products of photosynthesis. However, most of the water consumed by plants is by transpiration, a process that is controlled by climate and leaf area when soil water is not limiting. The combined processes of evaporation from the soil and plant surfaces and transpiration is called evapotranspiration (ET). Procedures for estimating ET are given in Chapter 8. Irrigation supplements precipitation by providing water essential for plant growth, in addition to controlling soil salinity and enhancing the root environment. The impact of irrigation on crop yields is greatest in arid and semiarid climates and in humid and subhumid climates during drought years. Irrigated crop yields are higher and more stable than yields on adjacent dryland areas (often referred to as rainfed areas). National famines, once common, have been avoided because of irrigation. Irrigation was a major factor in the success of the Green Revolution and will become even more important in the future to provide food for the ever-increasing world population. Stability of crop production under irrigation also provides economic stability to independent farmers and to communities. Without irrigation, increases in agricultural yields and outputs that have fed the world’s growing population would not have been possible. Irrigation has also stabilized food production and prices (Rosegrant et al., 2002). Irrigation is also used for many secondary purposes. For example, it is used to control frosts in orchards and citrus groves, and to condition the soil when preparing seedbeds and harvesting root crops such as potatoes, sugar beets, and peanuts (see Section 2.6). More recently, irrigation has become a common and environmentally acceptable method of making economic use of waste effluents for crop production. Previously, most of these effluents were merely disposed of in some manner, such as being discharged to rivers and streams, evaporated in holding ponds, or applied to nonproductive vegetation to accelerate the evaporation process. 2.1.3 Sustainable Irrigated Agriculture During the past decade, many authors have addressed the issue of sustainable development and sustainable agriculture and agricultural systems (Edwards et al., 1990; Hatfield and Karlen, 1993). Their publications address issues of soil fertility, crop and pest management, soil erosion, soil salinity, waterlogging, economics, environment and water quality, bioresources, ecology, policy, sociology, and socioeconomic issues. The quantity of renewable water resources, which is the most vital component of sustainable irrigated agriculture, has been addressed to a limited extent. Sustainable crop production on irrigated lands will require a sustainable and renewable water supply. Seckler et al. (1998) prepared a comprehensive analysis of global water demand and supply from 1990 to 2025. A water balance approach was used to derive estimates of water supply and demand for individual countries and regions and to identify major regions of water scarcity. During an international conference on sustainability of irrigated agriculture, various views of on-farm water management, water quality, and capacity building were discussed along with country and regional concerns in Africa, Asia, and China (Pereira et al., 1996). The Council for Agricultural Science and Technology (CAST, 1992) addressed the impacts of climate change on U.S. agriculture and the limitations of renewable water

Design and Operation of Farm Irrigation Systems

35

resources on adapting to climate change. A committee of the Board of Water Science and Technology of the National Research Council (NRC) conducted a thorough study of the future of U.S. irrigated agriculture (NRC, 1996). The Committee reported that the availability of water has been, and is likely to remain, the principal determinant of the status of irrigation in the western U.S. The availability of water is becoming increasingly important to irrigation in eastern states as well. The cost of water and demands on water resources are increasing. In 1989, the Food and Agriculture Organization (FAO) of the United Nations developed an international action program for the 1990s on water and sustainable agricultural development (FAO, 1989). It emphasized the need for long-term strategies consistent with limited water resources and competing demands for water. The NRC Committee on International Soil and Water Research and Development stated that one of the challenges we face in developing agricultural strategies that is truly sustainable is maintaining the resource base, the soil and water resources that make agriculture possible (NRC, 1993). Globally, the largest threat to sustainable irrigated agriculture has been, and still is, waterlogging and soil salinity. During the three decades from 1960 to 1990, the world witnessed a historic expansion of irrigated land. During this period, irrigated agriculture played a major role in increasing food production. Currently, an estimated 36% to 47% of the world’s food is produced under irrigation (Chapter 1). During the rapid expansion, some principles of irrigation design, operation and maintenance, and water balance appear to have been compromised resulting in waterlogging and soil salinity. Soils become waterlogged if water is imported in addition to precipitation in excess of crop ET and adequate drainage is not provided to remove the excess. Increases in soil salinity usually follow waterlogging because salts present in irrigation water become concentrated near the soil surface as evaporation from the soil removes pure water (see Chapter 7). In some developing countries, the continuing availability of relatively low cost energy may be a serious restraint on converting common surface irrigation systems to modern, more efficient systems that require more energy input. Thus, energy required for irrigation could also become a factor affecting the sustainability of irrigated agriculture. 2.1.4 Productive Irrigated Agriculture Some societies have practiced irrigation for centuries without adverse soil and environmental effects. In general, irrigated agriculture can remain productive indefinitely if a reasonably stable renewable water supply of suitable quality is available and the volume of applied water does not exceed ET and the drainage capacity, so that soils are not waterlogged. Perry (2002), referring to countries that have developed and controlled their water resources for decades and centuries, listed the essential elements of sustained water resources management as: knowledge of resource availability, governing policies, assigned priorities for developed water among users, allocation rules and procedures, defined roles and responsibilities, and infrastructure to deliver the service to each water user. In regions where water is scarce, agriculture may need to temporarily relinquish some water supplies in drought years to provide for domestic and other higher value water uses. This has become common practice in some areas of the western U.S. and will become a more common practice internationally. This practice may not change

36

Chapter 2 Sustainable and Productive Irrigated Agriculture

the productivity of the irrigated land in a given year, but will affect long-term mean production potential from an area because not all lands will be irrigated fully each year. Lands not fully irrigated may be planted to a crop and some crops may not be irrigated for near maximum yields. In addition to water, other inputs are needed to maintain productive irrigated agriculture. When irrigation removes the main constraint affecting crop yields (water), then other factors such as plant nutrition and changes in soil salinity may limit crop production. 2.1.5 Emerging Issues Facing Designers Designers of farm irrigation systems in the 21st century face many new challenges. Competition for limited water supplies will require irrigators to have greater control of the amount and uniformity at which the water is applied over the fields. Typically, this will increase the costs of new or modernized irrigation systems. Future renewable water supplies may not be as dependable as they have been in the past. For example, in parts of the Great Plains farmers are faced with decreasing flow rates from wells as the water table is lowered in the vast Ogallala aquifer. This aquifer has very limited natural recharge over much of its surface area. If a farmer already has an operating well on the farm, the efficiency of the pump will decline as the water table drops and the flow rate decreases. Farmers must consider the potential return on investments required to modernize and replace the pump and existing irrigation systems with more modern and efficient systems. The alternative is to return to dryland farming. The irrigation system designer must take into account the effects of decreasing flow rates and limited water supplies on irrigation system efficiencies. Particularly in regions like these, long-term planning is necessary and designers need to consider long-term returns, not just irrigation to meet the component of ET that is not not provided by precipitation or the consumptive irrigation requirements. The farmer also must consider alternative management practices to cope with the relatively large variation in seasonal precipitation in semiarid areas. Studies in the High Plains of Texas have shown that in most years, limited irrigation of droughttolerant crops can significantly increase yields and the productivity of water applied. In years of above-average rainfall, irrigation may be of limited benefit. For example, a three-year study in Kansas showed that when precipitation was about 30% above normal (during 1989 and 1990), there was no benefit from irrigating grain sorghum [Sorghum bicolor (L.) Moench]. During the following year (1991), precipitation was below normal, and sorghum and wheat yields as well as water use efficiency (WUE) were higher in irrigated treatments than on dryland treatments (Norwood, 1995). WUE is commonly defined as the marketable product produced per unit of water consumed in ET. In 1991 (the year with below normal precipitation) WUE tended to be greatest from the first increment of irrigation water applied. This example illustrates how weather variables can complicate the design of an optimal irrigation system in semiarid areas. Western U.S. water rights have long protected investments in irrigation facilities. But water laws are changing making it easier for farmers to sell the consumptive use (CU) component of their water right to users who are willing to pay more for the water than the farmer’s profit potential from irrigation. If some farmers within a project or district sell their entire CU rights rather than lease it for a season, the cost of operation and maintenance of the district’s main distribution system will be distributed to the

Design and Operation of Farm Irrigation Systems

37

remaining farmers who continue to irrigate. Where this situation occurs or is expected to occur, designers need to give the farmer information about the alternatives of modernizing a system, rather than maintaining the present system, and also about the options of selling the water right and terminating irrigated farming operations. More commonly, farmers are leasing their water during water-short years through various mechanisms (NRC, 1992). Under these conditions, an irrigation system may not be used each year, which will also affect annualized system costs. Distributing fertilizers and pesticides in irrigation water (chemigation) has become commonplace and can be very efficient. Chemigation requires new design considerations compared with traditional systems. In addition, new environmental concerns and associated regulations affecting return flows to streams or to the groundwater are becoming significant factors that the designer must consider. Many states have installed automated weather stations to measure and disseminate estimates of evaporative demand or crop ET. Use of such data by irrigation managers to schedule irrigations can enhance the productivity and efficient use of water resources (Ley, 1995). Designers of modern irrigation systems must consider that, in the future, most farmers will become dependent on real-time estimates of ET and soil water depletion when scheduling their irrigations and determining the amounts that should be applied. The designer needs to consider simulating multiple years of operation based on estimated daily ET using procedures described in Chapter 8. For highvalue crops, many new irrigation systems will be automated with irrigations initiated using soil water sensors, computed depletion of soil water, or remote sensing techniques.

2.2 ROLE OF IRRIGATION IN FOOD AND FIBER PRODUCTION 2.2.1 Population Growth and Renewable Water Supplies World population growth and irrigation are related, i.e., an increasing population requires increased irrigation. World population is increasing about 1.2% per year (Population Reference Bureau, 2005); it was about 5.3 billion in the early 1990s and increasing at a rate of 93 million people per year. The U.N. Population Division suggested that the global population will increase to 9.3 billion by 2050. With the increased population comes increased food needs. Thus, population growth and its effects on water supplies and food requirements is a variable that must be considered by irrigation system designers and operators, particularly in many developing countries. In the U.S., population growth probably will not be significant in the near future. However, hotspots of water scarcity occur in the Colorado, Rio Grande, and Texas Gulf basins (Rosegrant et al., 2002). Irrigation designers must consider planning at a regional scale and be aware of current regional plans and expected new regulations in order to properly design irrigation systems. Rosegrant et al. (2002) provided a useful table of 1995 and projected 2025 populations by country and region using data from several sources. The world annual per capita internal (i.e., within a country) renewable water supply is about 7000 m3, but this varies greatly between regions. For instance, in the Middle East and North Africa, the International Bank for Reconstruction and Development (IBRD) estimates the per capita water supply to be only about 1000 m3 (IBRD, 1992b). Renewable water supplies are finite and can essentially be considered as a constant within a region. The rate of population growth is often very high in regions

38

Chapter 2 Sustainable and Productive Irrigated Agriculture

that already have some of the lowest per capita water supplies. For example, the weighted average rate of increase in population in the Middle East in the late 1980s was 2.7% per year (Jensen, 1990). In that region, the per capita water supply, which is already low, would be expected to decrease to half its present value in just 26 years. The estimated 2005 population in this region was 47% higher than in 1989 (Population Reference Bureau, 2005). The weighted average annual rate of increase in 2005 had decreased slightly to 2.2%. 2.2.2 Increasing Food Demands Relative to Water Supplies As populations increase, competition among agricultural and non-agricultural water uses will affect irrigated agriculture in many semiarid and arid areas. These pressures will become more severe as water supplies are more fully utilized. Conflicts will increase. Solutions to these conflicts will require revisions of water law and allocations. Australian scientists, for example, are proposing a landmark national water trading framework to define water rights for irrigation. This water rights system would be based on banking, share trading, and Torrens Title registration procedures which would allow water to be traded via electronic transfers with licensed brokers and clear trading rules. The system would have three components: an entitlement, an allocation, and a use license (CSIRO, 2002). The authors claim that the advantage of this system is that it would provide flexible control as climate, economic, and technical circumstances vary. In the future, the price of water will be more closely related to its real economic cost. This will affect irrigated agriculture because the current value of the marketable product of some crops per unit volume of water consumed is low. For example, only about 2 kg of maize grain valued at about $0.20 can be produced in arid and semiarid areas per cubic meter of water consumed in ET. For many semiarid and arid areas, it will not be economical to import and use a cubic meter of water (= 1 ton of water) to produce 2 kg of grain even if a renewable source of water was available. These areas will be forced to use local water for domestic purposes and to import more feed and food supplies (Jensen, 1993). Importing grain rather than importing water to produce grain locally is sometimes referred to as importing virtual water. Where agriculture has been the primary source of foreign currency, depending on virtual water may require generating foreign currency from other sources. Few irrigation projects are planned to supplement dryland agriculture in semiarid areas. However, limited irrigation can significantly increase the production per unit of precipitation and stored soil water consumed by ET. For example, in the High Plains of Texas, dryland grain sorghum produces from 0.45 to 0.85 kg/m3 of water with yields from 1000 to 3000 kg/ha. Limited irrigation can increase yields to about 5000 kg/ha and average productivity to about 1.2 kg/m3 (Jensen, 1984). The first 100-mm increment of irrigation water had the most effect by increasing the yield from 2500 to 4000 kg/ha. The productivity of this increment of irrigation water would be 1.5 kg/m3. This is an example of the future role for irrigation in semiarid areas, especially where renewable groundwater supplies are available. Irrigation development must be integrated with dryland production rather than planned, developed, and managed as a competitive enterprise (Jensen, 1993). 2.2.3 Downstream Environmental Effects Downstream environmental effects may include reductions in wetlands and lakes as crops consume water that formerly flowed to lakes and wetlands. One dramatic exam-

Design and Operation of Farm Irrigation Systems

39

ple is in the former Soviet Union. In 1918, the Soviet Union decided to become selfsufficient in cotton. To irrigate cotton fields, they began diverting 55 km3 of water per year from two rivers that feed the Aral Sea (Sun, 1988). Over the years, little water reached the Aral Sea and it has shrunk drastically in size. Huge salt storms have carried salt from the shores of the shrinking basin great distances, damaging agricultural lands (Ellis, 1990; Micklin, 1988). The ICID organized a special session on the Aral Sea Basin in 1994 (ICID, 1994), a NATO-sponsored workshop was held on this issue in Wageningen, The Netherlands, in 1995, and the World Bank has established an independent advisory panel to provide guidance to the Bank on managerial and technical issues of this problem. However, the Aral Sea problem still exists. The river system has been encroached upon by bridges, etc., so that even if irrigation diversions ceased, the increased flow could not reach the Sea. In hindsight, a better approach would have been to balance the water requirements for irrigation against the inevitable reduction in the flow to the Aral Sea and the resulting environmental consequences. Heaven et al. (2002), using a real-time mass balance river/reservoir model for Syr Dara (one of the rivers), concluded that water is either being used extremely inefficiently, or river offtake data or cropped area data are unreliable, or both. Water is still being diverted for irrigation. The river system has been encroached upon by bridges etc. so that even if diversions ceased, the increased flow could not reach the Sea. A great misconception still exists that by increasing irrigation efficiency, there would be more water for the Sea. "Irrigation efficiency" is the fraction of water consumed. On a much smaller scale in the U.S., the diversion of water from the Truckee River in 1902 lowered Pyramid Lake in Nevada about 24 m during the following 80 years. Although reallocation of water supplies appears to have stabilized the lake, the large, deep freshwater lake, which used to overflow periodically into Winnemucca Lake, has become slightly more saline (Jensen, 1993). 2.2.4 Water Quality Aspects A major external effect of irrigation is the dissolved solids in return flows. ET concentrates the salt that is in applied water. This decrease in water quality is inevitable with removal of pure water by any consumptive water use, not just agriculture. This is an environmental cost that represents a trade-off between allowing water to flow in streams to sustain instream environmental benefits versus offstream beneficial consumptive uses. In the future, more regulations or compromises mutually agreed upon by users and regulators can be expected. Some fertilizer, mainly nitrogen, is added to the return flow downstream from irrigated areas by leaching. Better nutrient and irrigation management, however, can limit the impact of this problem. For example, Ferguson et al. (1990) showed that both irrigation and fertilizer management influenced the amount of nitrate-nitrogen leached. They showed that the potential reduction in fertilizer costs by using irrigation scheduling, soil testing, and fertilizer management services could be profitable. Studies in other areas have shown similar results. Application of excess plant nutrients and ineffective water management will continue until the real cost of water pollution is paid for by the polluters or regulations are implemented. Pesticide residues have also been found in return flows; these can be likewise prevented or reduced to minimal and safe levels by improved management practices.

40

Chapter 2 Sustainable and Productive Irrigated Agriculture

Irrigation in some arid areas with complex geology can affect the quality of water flowing to downstream wildlife areas. For example, Lico (1992) reported drainage to wildlife areas in west-central Nevada showed high concentrations of potentially toxic arsenic, boron, lithium, molybdenum, un-ionized ammonia, uranium, and vanadium. Several such investigations by the U.S. Geologic Survey have been completed. A committee commissioned by the Water Science and Technology Board of the U.S. National Research Council analyzed irrigation-induced selenium contamination and irrigation-related drainage problems associated with the Kesterson Reservoir in California (National Research Council, 1989). In that case, the reservoir captured drainage water that originally was intended to drain to the San Francisco Bay area. Another potential harmful effect of poorly managed irrigation is spreading human diseases: schistosomiasis (bilharzia), dengue and dengue hemorrhagic fever, liver fluke, filariasis, onchocerciasis (river blindness), malaria, Japanese encephalitis, and diarrhea. Schistosomiasis is most closely linked to irrigation. Available guidelines for policy makers, planners and managers forecast vector-borne diseases and help incorporate safeguards into irrigation projects (Birley, 1989, and Tiffen, 1989). 2.2.5 Irrigation Status and Trends The area of land irrigated in the U.S. increased from the mid-1940s until the 1980s (Figure 2.1), and is continuing to increase in some regions. In the Great Plains, depletion of groundwater and some governmental programs caused the decline in irrigated land area during the early 1980s, mainly in the High Plains of Texas. Increased irrigation mainly in Nebraska and in Texas increased the total in 1997 and 2002 to about that in 1978. (Some apparent changes in the 1980s are due to changed statistical procedures.) Irrigation expanded rapidly in the Texas High Plains during the drought years of the 1950s. Excellent soils, topography well-suited for surface irrigation without any land leveling, and a good quality, easily accessible water source enabled rapid irrigation expansion. The well-known Ogallala aquifer was the main source of water. However, its natural recharge rate in the area was very low and groundwater levels dropped rapidly during the next 20 years. As a result, well flow rates decreased drastically and pumping costs increased due to larger pumping lifts and higher fuel costs. Some lands put under irrigation in the 1950s and 1960s have been returned to dryland farming. Recently, most surface irrigation systems have been replaced by center pivot (CP) sprinkler systems because center pivots can apply water uniformly with smaller flow rates. Also, less water needs to be pumped because with CP systems there is usually little or no runoff and less than full irrigation (deficit irrigation) can be achieved. In Florida, farmers irrigate a high percentage of cropped land because of very sandy soils and the need for quality control of high-value crops even though annual rainfall is about 1270 mm. Florida growers have converted many sprinkler systems to microirrigation systems. In citrus groves, microsprinklers are designed to cover the citrus trunks for freeze control (Parsons and Wheaton, 1995). Irrigated lands have also increased steadily in other eastern states since 1978 (Figure 2.2). These include the north-central states of Michigan, Minnesota, and Wisconsin where sandy soil has been a major factor influencing expansion of irrigated lands. In these states, groundwater has been the main source of water. Without a convenient and renewable source of water, farmers do not irrigate, even though potential increases in crop yields would offset

Design and Operation of Farm Irrigation Systems

41

8.0

Irrigated Land, hectares, millions

7.0 6.0 5.0 4.0 3.0 2.0 1.0 0.0 1940

1945

1950

1955

1960

1965 SW

1970 NW

1975

1980

MTN

1985

1990

1995

2000

2005

GP

Figure 2.1. Expansion of irrigated land from 1944 to 2002 in the western U.S. SW = Southwest (Arizona, California, New Mexico, Nevada); NW = Northwest (Idaho, Oregon, Washington); MTN = Mountain States (Colorado, Utah, Wyoming); and GP = Great Plains (Kansas, Nebraska, Montana, North Dakota, Oklahoma, South Dakota, Texas). Source: USDA 1997 and 2002 Censuses of Agriculture (USDA-NASS 1999, 2004).

3.5

Irrigated Land, hectares, millions

3.0

2.5

2.0

1.5

1.0

0.5

0.0 1940

1945

1950

1955

1960

1965

FL

SE

1970 L-Miss

1975

1980 MW

1985 NC

1990

1995

2000

NE

Figure 2.2. Expansion of irrigated lands from 1978 to 2002 in the eastern U.S. FL = Florida; SE = Southeast (Alabama, Georgia, Kentucky, North Carolina, South Carolina, Tennessee, Virginia); L-Miss = Lower Mississippi (Arkansas, Louisiana, Mississippi, Missouri); NC = North-Central (Michigan, Minnesota, Wisconsin); and NE = 12 remaining northeast states. Source: USDA 1997 and 2002 Censuses of Agriculture (USDA-NASS 1999, 2004).

2005

42

Chapter 2 Sustainable and Productive Irrigated Agriculture

capital and operating irrigation costs, especially in drought years. Irrigated land in the lower Mississippi states of Arkansas, Louisiana, Mississippi, and Missouri has increased markedly since 1978. Over half of the irrigated land in the lower Mississippi region is in Arkansas. Conflicts over water have developed in the southeast because of limited groundwater supplies and lack of surface water storage facilities. The impact of sustainable water supply on the expansion or decrease in irrigated area is readily apparent when considering its recent history in the Great Plains. The area of irrigated land in Texas expanded rapidly until about 1980, and then decreased as groundwater supplies decreased. In contrast, the irrigated area in Nebraska also increased rapidly, especially with the development of the center-pivot sprinkler system in the 1960s, and has continued to increase to 2002 (Figure 2.3). Nebraska’s water supply is from both surface and groundwater supplies, which have been relatively stable. Global statistics on irrigated land area are not as robust as that in a single country because uniform statistical methods are not always used. For example, in some countries cropping intensity is considered, so land that is double cropped within one year may be counted twice. Also, sometimes the area commanded by a canal may be reported instead of the actual land irrigated. The irrigated land reported by FAO appears to be the most consistent and is the basis of the values reported in this section. The rate of increase in irrigated land reached a peak of about 2.2% per year in the mid-1970s (Jensen, 1993). The total arable land irrigated at the beginning of the 21st century was about 275 million hectares. The annual rate of increase has decreased since the mid1990s to less than 0.5% per year at the beginning of the 21st century (Figure 2.4). 3.5

Irrigated Land, hectares, millions

3.0

2.5

2.0

1.5

1.0

0.5

0.0 1940

1945

1950

1955

1960

1965 TX

1970 NE

1975 KS

1980

1985

1990

1995

2000

MT+ND+OK+SD

Figure 2.3. Expansion of irrigated land in the U.S. Great Plains regions. TX = Texas; NE = Nebraska; KS = Kansas; and MT = Montana, ND = North Dakota, OK = Oklahoma, SD = South Dakota. Source: USDA 1997 Census of Agriculture (USDA-NASS, 1999).

2005

Design and Operation of Farm Irrigation Systems

43

300 y = 3.4292x - 6583.3 Irrigated Land, hectares, millions

250

200

150 y = 2.2046x - 4232.6373 100

50

0 1970

1975

1980

1985

1990

1995

2000

Year World

Asia

N&C Amer

Africa

S Amer

Linear (World)

Linear (Asia)

Figure 2.4. Expansion of irrigated land from 1972 to 2002 for the World, Asia, North and Central America, Africa, and South America. Source: FAO Yearbooks and FAOSTAT at www.fao.org.

2.3 CROP PRODUCTION AND IRRIGATION WATER REQUIREMENTS 2.3.1 Crop Production-Evapotranspiration Relationships When soil water is not adequate for plants to transpire at the climate-based evaporative demand rate, plants experience water stress and plant stomates begin closing to limit transpiration. Since plant stomates also provide for uptake of carbon dioxide, the rate of photosynthesis decreases. Numerous studies have shown that plant growth and crop yields decrease linearly with decreased transpiration. Since evaporation is only a part of total ET, plant water stress distributed over the growing season also decreases dry matter and marketable yields as ET decreases (Howell, 1990; Howell et al., 1990). Howell (1990) reviewed the results of previous studies and presented a detailed analysis of yield-transpiration and yield-ET relationships for corn, grain sorghum, and winter wheat. The general linear statistical relationship between total ET and yield of cereal grains in a given climate zone is: (2.1) Yg = b ET + a where Yg = the yield or marketable product, Mg/ha ET = growing season evapotranspiration, mm b = the slope of the yield-ET line, Mg/(ha mm) a = a constant, Mg/ha. Because some ET occurs before the first increment of marketable product is produced, the value of a is negative. For example, Stewart et al. (1977) reported that for corn yields from several non-saline treatments at Davis, California, in 1974 could be estimated as Yg = 0.019 ET – 1.17 Mg/ha, i.e., the increase in corn grain per unit increase

44

Chapter 2 Sustainable and Productive Irrigated Agriculture

in ET will be 1.9 kg per cubic meter of water consumed in ET. Values of b and a will vary with crop cultivars and climate. As new cultivars are developed with higher harvest indices and improved cultural practices, yields may increase with little increase in ET, thereby changing the constants. Climate also will affect the equation constants. With frequent rains when annual plants are small, the evaporation component will be higher than in arid areas. The slope (b) may be less in humid areas. Similar relationships can be developed for dry matter or for forage production. Forage yield-ET relationships from carefully controlled field experiments tend to vary less than grain yields. For example, when using regression data from dryland studies in North Dakota (Bauder et al., 1978), sprinkler-irrigated alfalfa in New Mexico in 1979 (Sammis, 1981), and regression data from studies in Israel (Kipness et al., 1989), the resulting yield of alfalfa at zero moisture could be expressed as Yalf = 0.015 ET – 0.50 Mg/ha, that is, the increase in alfalfa yield per unit of ET will be 1.5 kg/m3. This value obtained from widely differing climates agrees closely with the seven-year average of well-watered, lysimeter-based yield reported by Wright (1988). Wright reduced the lysimeter yields 5% to agree with adjacent field yields of 1.76 kg/m3 at 12% moisture, or 1.6 kg/m3 at zero moisture. Howell et al. (1990, 2000) also presented detailed analyses of concepts of water use efficiency (WUE), which is the the economic or marketable production per unit of water consumed in ET (WUE = Y/ET). Bos (1980), working with an ICID group, defined a related term of irrigation water use efficiency (WUEi), which considers the increase in production per unit increase in ET due to irrigation, or WUEi = (Yi – Yd)/(ETi – ETd). If there is no dryland yield or ET, the values are the same. Both of these terms have served useful purposes as indices of the effectiveness of various agronomic practices. I have used the WUE terms since the 1950s. However, due to confusion with irrigation efficiency that has often resulted in the literature, I suggested that these terms be considered as indices of the productivity of water used in crop production (Jensen, 1993). Either “total water consumed” or “irrigation water consumed” in ET is needed to enable uniform comparisons of practices or between areas. A comprehensive series of articles on the response of crops to carbon dioxide, temperature, and water quality, and on water shortage, is presented in a publication edited by Kirkham (1999). Chapters on water use by several crops such as cotton, olives, rice, sorghum, sunflowers, turfgrass, and wheat are presented along with a chapter on risk assessment of irrigation requirements of field crops. Stewart and Nielsen (1990) compiled detailed information on irrigation of various crops in a comprehensive monograph. Relationships between yield and water applied are much more variable than those between yield and ET, because the rainfall during the growing season and the amount of water percolating below the root zone are not the same each year, or from place to place, and will vary with the irrigation method. Even surface runoff may not be taken into account in data reported on water applied and yield. Yield-water application relationships tend to be curvilinear, as shown by Howell et al. (1995) and Letey (1993). Such relationships also tend to be applicable to specific conditions and are not as transferable or as valuable in comparing production practices and systems as WUE parameters. Unger and Howell (1999) presented a global perspective summarizing basic principles and practices for achieving agricultural water conservation, both under dryland (rainfed) and irrigated conditions.

Design and Operation of Farm Irrigation Systems

45

2.3.2 Deficit Irrigation and WUE A relatively new practice of not replenishing the soil water profile at each irrigation, where rainfall during the growing season is significant, is called deficit irrigation. This practice provides soil water storage space for capturing rainfall that otherwise may percolate beyond the root zone. For example, Howell et al. (1995) have shown that when replacing only two-thirds of the depleted soil water at irrigation, the irrigation water use efficiency for irrigation water (WUEi) for corn increased from 1.8 to 2.3 kg/m3. WUEi is defined in Section 2.3.1. The net effect is that a higher proportion of rainfall may be effective with deficit irrigation than with full irrigations. Another form of deficit irrigation is managed plant water stress. For example, when irrigating some small grain cultivars that are subject to lodging, mild plant water stress prior to the boot stage can shorten the stems and increase the resistance to lodging. Terminating irrigation at critical times such as after the soft dough stage of cereals can reduce ET and irrigation water required without a decrease in economic yield. Similarly, on sugar beets, more leaf area may be produced than necessary. Under these conditions, allowing increased plant water stress by limiting irrigations near the end of the season can reduce the mass of beets produced, but increase sugar content. As a result, the total sugar yield may not be affected. A light irrigation before harvest may be required to harvest the beets. A side benefit is that less tonnage needs to be hauled to the sugar factory. Studies on cotton in California showed that irrigations that are too early not only lower final yield of cotton lint, but also result in lower production efficiency (Grimes and El-Zik, 1990). These are examples of how operators of irrigation systems can manage irrigations to reduce water requirements and costs without reducing economic yields. Local experimental data should be used as a guide to manage deficit irrigation. FAO published a Water Report summarizing experiences of deficit irrigation in a number of countries (FAO, 2002). The potential effects of deficit irrigation practices also will need to be considered when designing farm irrigation systems. 2.3.3 Irrigation Requirements for Evapotranspiration and Salinity Control For design purposes, in most arid areas and especially during the dry part of the season, precipitation has little effect on estimates of irrigation water requirements for ET and salinity control. Methodology for estimating crop ET and irrigation water requirements is presented in Chapter 8. Procedures for estimates the amount of water required for leaching are presented in Chapter 7. Leaching for control of soil salinity does not always need to be provided at each irrigation. Pre-plant irrigation may adequately leach the surface soil for new seedlings. Depending on the sensitivity of the crops to be grown and the rainfall during the winter season, planning for leaching during the low ET-demand rainy period may make more effective use of rainfall. Estimating the design capacity in subhumid and humid areas can be more complicated than in arid areas. The economic benefit of irrigation in subhumid and humid areas is derived mainly during extended drought periods. If the irrigation system is not able to replenish water during these periods, the potential economic benefits of irrigation may not be achieved. These issues are discussed in detail in Chapter 8. 2.3.4 Drainage Requirements for Water and Salinity Control In the 1990s, the World Bank (IBRD) estimated that a quarter of all irrigated land was affected by salinity caused by poor irrigation practices (IBRD, 1992b). On newly irrigated land, some 10 to 50 years may be required for excess water to fill the underlying unsaturated zone. If excess water input is ignored during this period, the rise in

46

Chapter 2 Sustainable and Productive Irrigated Agriculture

the water table inevitably results in waterlogging unless the natural drainage capacity increases with the rise in the water table. On a project basis, well-designed farm irrigation systems and efficient management of excess water from the start of a project can delay the need for increased drainage capacity for many decades, or in some cases prevent the need to install additional drainage facilities (Jensen, 1993). The benefits of controlled drainage are beginning to receive more attention. This means that drainage is allowed only when the water table needs to be lowered to prevent crop damage or when encouraging leaching of salts. Abbott (2003) reported a study that showed that controlled drainage can give significant benefits to both the farmer as increased yields and to the wider community in terms of reduced water requirements. Areas of potential application for controlled drainage were identified in North Africa (Algeria, Egypt); the Middle East (Israel, Syria, Iraq, Bahrain); India (Punjab, Haryana, Rajasthan); and Asia (Pakistan, Northern China, Central Asian States). The need for increased drainage capacity is strongly linked to the types of farm irrigation systems used and how they are managed. When most new irrigation projects were dominated by surface irrigation, increased drainage capacity was nearly always an essential requirement for sustained agricultural productivity. With improved irrigation systems that apply water more uniformly and with more efficient water management practices including deficit irrigation, the capacity of large-scale drainage above the existing natural capacity may be decreased. However, because of nonuniform subsoils, some increased localized drainage capacity will usually be required in most new irrigation projects. Seepage from unlined canals and ditches can also cause high water table levels and increase drainage requirements. Details on salinity control and drainage requirements and design can be found in Chapters 7 and 9. Reuse of waste effluents has become a common practice where renewable water supplies are limited. Use of irrigated agriculture to dispose of waste effluents has also become an accepted and preferred practice over adding this burden to sewage treatment plants. Special management practices may be required and there usually are limitations on the types of crops on which waste effluents can be used. Details on use of waste effluents in irrigation are presented in Chapter 20.

2.4 SYSTEM DESIGN AND INCREASING COMPETITION FOR RENEWABLE WATER SUPPLIES Increasing competition among water uses will increase pressures on irrigated agriculture because agriculture consumes most of the renewable water supplies in most semiarid and arid areas. These pressures will become more severe as water supplies are more fully utilized. Conflicts will increase. Solutions to these conflicts may result in revisions of water laws and allocations. In the future, the price of water will be more closely related to its real economic cost or value. This will affect irrigated agriculture because the marketable product of many crops per unit volume of water consumed is low. The designer working with individual farmers or on farm irrigation systems for a new project or a project that is being modernized must consider the available water supplies, particularly renewable water supplies. These supplies are not only based on hydrological aspects and the physical facilities available to store and deliver the water, but on legal and social aspects. In the U.S., an individual farmer’s water right to water

Design and Operation of Farm Irrigation Systems

47

from a common surface water source is the most common assurance of a renewable water supply. In some cases, the water right may be granted to an quasi-public irrigation district rather than to individuals. It is generally understood that in the western U.S., when a water right is sold or transferred, only that portion of the right that is used consumptively can be transferred. If groundwater is the only source of water, then the expected recharge rate relative to the expected annual withdrawal rate must be considered to determine if the water supply can be sustained indefinitely, or at least until the investment in the irrigation system can be recovered. In areas where groundwater is essentially being mined, as in parts of the Ogallala aquifer of the Great Plains, state water laws regulating pumping must be considered. In addition, if groundwater is being mined, the rates of withdrawal and declining well-yields as the saturated layer decreases must be factored into the planning and design process. Renewable water supply issues are not covered in depth in this monograph. Details addressing hydrologic issues and water storage and conveyance systems must be obtained from other sources. The U.S. has agreements with Canada and Mexico on sharing common water supplies. However, in many regions, no international agreement governs sharing waters of rivers that flow through several countries. More than half the world's largest rivers flow through two or more countries. The potential for conflict over shared water is increasing greatly. The implications for food security and managing irrigation projects are enormous (Jensen, 1993). International law for sharing water will become more important as competition increases. Guidelines for international agreements are available (Kimball, 1992). 2.4.1 Environmental Aspects, New Dimensions The International Conference on Water and the Environment, held in Dublin in January, 1992, focused on sustainable development, management, and utilization of water resources in harmony with environmental conservation. It identified environment and development priorities to be agreed upon at the June, 1992, United Nations conference in Rio de Janeiro. Agenda 21, adopted at the Rio conference, gave high priority to win-win policies of poverty reduction, economic efficiency, and economic development. Also about this time, the policy of the World Bank relative to new projects changed. The Bank staff now required investment projects to be classified based upon their potential environmental impact. Projects in Category A required a full environmental assessment. These include dams and reservoirs and large-scale irrigation projects (IBRD, 1992a). Policies for sustained development include “removing subsidies that encourage excessive use of fossil fuels, irrigation water, and pesticides and excessive logging” (IBRD, 1992b, p. 2). The possible trade-offs between water for food production and water for nature became one of the most contentious issues in discussions of the Second World Water Forum in The Hague in 2000 (Rosegrant et al., 2002). They concluded that this conflict might be one of the most critical problems to be tackled in the early 21st century. Gleick (2002) indicated that in the early 20th century in the U.S. West and throughout the world, water left in the stream was considered “wasted.” The language of water development and reclamation indicated a perception that water was a resource to be extracted and put to use. The meaning of beneficial use of water is beginning to expand from a narrow economic interpretation to a broader view recognizing the growing public awareness of the value of environmental resources. South Africa

48

Chapter 2 Sustainable and Productive Irrigated Agriculture

has formally recognized the right to water for natural ecosystems in their new water laws and constitution (Gleick, 2002). In the U.S., competition for water for endangered species and wildlife, such as for the whooping crane in Nebraska and salmon in northwest rivers, will affect water supplies available for irrigated agriculture The increasing demand for domestic and other water uses may affect the order of water reuse. Foe example, Falkenmark and Widstrand (1992) suggested that because domestic use requires the cleanest water, the ideal order of reuse would be household first, then industry, then agriculture. Agricultural use would be last because most irrigation water is consumed, i.e., it returns to the atmosphere through evapotranspiration and is not available for reuse. This concept is becoming very real in some areas as clean water is being provided first to municipalities. Agriculture, which used to have first access to the water, is being required to use municipal return flow. 2.4.2 Water Supply and Consumption In semiarid areas, the area of irrigated land alone does not indicate the volume of irrigation water that is consumed because of variable rainfall and because some crops are often not fully irrigated. Likewise, the traditional concept of irrigation efficiency is not very meaningful when considering water supplies for irrigation and water consumed by evaporation and transpiration. A better procedure for accounting for water use, based on a water balance approach, was developed by Molden (1997). Water accounting is a procedure for analyzing the uses, depletion, and productivity of water in a water basin context. A key term is water depletion, which is the use or removal of water from a water basin such that it is permanently unavailable for further use as a liquid within the basin. He described process and non-process depletions. Process depletion is where water is depleted to produce an intended good. In agriculture, process depletion is transpiration plus that incorporated into plant tissues⎯the product. Non-process depletion includes evaporation from soil and water surfaces and any nonevaporated component that does not return to the freshwater resource. It would also include depletion by weeds and other non-economic vegetation. The depleted fraction is that part of inflow that is depleted by both process and non-process uses of water. The productivity of water is a performance parameter that can be related to the physical mass production or economic value per unit volume of water. Molden suggested that the productivity of water can be measured against gross or net inflow, depleted water, process-depleted water, or available water in contrast to the production per unit of water consumed in ET (Viets, 1962). Water accounting can be done at various levels such as the field, irrigation service, basin, or sub-basin levels. Molden et al. (1998) presented a detailed example of water accounting at the basin level using data from Egypt’s Nile River, where some detailed information on water use and productivity was available. Molden and Sakthivadival (1999) presented another example for a district in Sri Lanka. Numerous conferences have been held resulting in papers concerning the quantification of real water, which is water that may be available for transfer or reallocation to higher value uses. Irrigation efficiency concepts enter in when discussing possible real water amounts that can be saved and made available for transfer by making changes on irrigated lands. Roos (2001), in discussing agricultural water conservation in relation to water available for transfer to another basin, stated: “Unless evapotranspiration can be reduced, such measures do not add to the regional supply. Depletion remains essentially the same even with greater application efficiency....” His com-

Design and Operation of Farm Irrigation Systems

49

ments related to the Sacramento and San Joaquin valleys, which function like closed basins. He also stated: “But the general rule is still valid: unless depletion changes, there is no real water savings available for transfer.”

2.5 IRRIGATION WATER MANAGEMENT DURING DROUGHTS Changes in water rights of renewable water supplies from agriculture to higher value uses have been underway in the western U.S. for decades. Today, new methods and practices are used to provide temporary exchanges in water supplies. For example, when water supplies are plentiful, farmers who irrigate continue to use the water available under their water rights, but when annual supplies are below normal, temporary exchanges in water uses have become more common. NRC (1992) summarized some of the mechanisms being used to provide flexibility in the use of renewable water supplies. The NRC Committee discussed various water market mechanisms, third party impacts, water transfer opportunities, and the role of water law in the transfer process. Case studies of water transfers in several areas of the West were presented along with several examples of water banking practices.

2.6 OTHER AGRICULTURAL PURPOSES AND BENEFITS OF IRRIGATION Irrigation is used mainly to provide water used in ET and for control of soil salinity. However, irrigation is also used for other purposes that are associated with irrigated agriculture. These secondary uses may be important to the success of an irrigated enterprise. Some of the special uses that need to be considered by the designer are summarized below. The designer should consider local practices and special crop needs when planning and designing an irrigation system. 2.6.1 Soil Conditioning In arid areas, where crops are planted to meet specific marketing opportunities, irrigation is essential to germinate and establish plants because rainfall may not occur at planting time or is not dependable. In some areas, solid set sprinklers are set up solely for this purpose and then removed after the new crop has been established. After establishing the new crop, irrigations are continued using the existing irrigation system, which is usually a surface system. 2.6.2 Crop Cooling Crop cooling is one of several special uses of irrigation. It is a minor component of irrigated agriculture, but is used on diverse crops for various purposes. For example, when planting lettuce in the Imperial Valley of California about mid-September, daily 3-hour sprinkler irrigations are applied to cool the germinating lettuce seedlings until they are established. Vineyards have been irrigated for 2 to 3 minutes at 15-minute intervals to cool the plants down to 8° to 14°C. Other crops that have been cooled by sprinkler irrigation are almonds, apples, beans, cherries, cotton, cranberries, cucumbers, flowers, potatoes, prunes, strawberries, sugar beets, tomatoes, and walnuts (Sneed, 1972; Gray, 1970; Kidder, 1970; Carolus, 1971; Unrath 1972a,b). Air conditioning of crops for higher yield and better quality has been studied in California, Michigan, and Washington. Summaries of crop cooling principles have been presented by Chessnes et al. (1979) and Merva and Vandenbrick (1979). Evaporative cooling of apples in North Carolina has increased their red color and soluble solids (Unrath and

50

Chapter 2 Sustainable and Productive Irrigated Agriculture

Sneed, 1974). Sprinkler irrigation has been used to delay fruit bud development in Utah to avoid damage due to low spring temperatures (Alfaro et al., 1974; Anderson, 1974). Irrigation designers should consider recent technological developments, local practices, and current economics in planning crop cooling by irrigation. 2.6.3 Frost and Freeze Protection Freeze protection is another special use of irrigation. By coating plants with water as air temperatures drop below freezing, the heat of fusion released as the water freezes maintains the plant temperature at 0°C. The ice coating on plants must continually be in contact with unfrozen water until the ice melts. Early studies have shown that plants could be protected against freezing against temperature as low as –9°C with no wind and applications of 6.5 mm/h. Early studies of frost and freeze protection for small fruit, flowers, potatoes, and grapes were described in the first edition (Braud and Horthorne, 1965; Harrison and Gerber, 1965; Loscascio et al., 1967; Lamade, 1968; Kidder, 1970; Sneed, 1970; Braud and Esphahani, 1971; Harrison et al., 1974). Severe freezes in Florida in the 1980s killed about 82,000 ha of citrus. Since then, microsprinklers have largely replaced heaters and wind machines in Florida (Parsons and Wheaton, 1995). Various techniques for frost and freeze protection were discussed in an ASAE monograph, Modification of the Aerial Environment of Crops, by Barfield and Gerber (1974). The designer should consider recent technological development, local practices, and current economics in planning for frost and freeze protection by irrigation. 2.6.4 Soil Conditioning In arid areas, the surface soil may become very dry following the harvest of the previous crop. In these areas, it is often impossible to prepare a seed bed for smallseed crops without first moistening the surface soil. Thus, irrigation becomes a necessity for mechanical purposes rather than providing water for plant growth. Similarly, irrigation may be necessary to permit harvesting root and tuber crops such as sugar beets, potatoes, and peanuts. Much of the water applied for this purpose may be stored as soil water for use by the next crop, stored for the next cropping season, or used to leach salts. 2.6.5 Applications of Chemicals in Irrigation Water Application of fertilizers, pesticides, herbicides, desiccants, and defoliants— chemigation—is well established and documented. Because this practice has become so widespread a new chapter has been added on this subject (see Chapter 19).

2.7 SYSTEM DESIGN AND OPERATION CHALLENGES Recent and emerging design issues concern environmental aspects affected when using surface runoff from surface irrigation systems. These include any pesticides and plant nutrients that may be included in the runoff, erosion and sediment deposition in the collecting flow channels and receiving water bodies. Likewise, when converting from surface irrigation systems to drip or sprinkler systems, wetlands that had developed from previous irrigation practices may gradually dry up. The impact of both surface and subsurface water quality on instream and downstream water uses will become significant factors in the future if not already a factor in the area under consideration. Designers of irrigation must also consider the various legal, environmental and physical constraints that may affect the design and operation of an irrigation system.

Design and Operation of Farm Irrigation Systems

51

Today, designers of irrigation systems must take into account many other aspects in addition to those considered essential several decades ago. Economic aspects will be more challenging if stable renewal water supplies are not assured and prices received for farm produce fluctuate widely. English et al. (2002) indicate that irrigated agriculture will need to adopt a management paradigm that is based on maximizing net benefits rather than maximizing yields. Irrigation to maximize benefits will be more complex and challenging than the conventional practice of maximizing crop yields per unit of land. Two important concerns are sustainability and risk. The increased complexity of analysis will require more sophisticated analytical tools. One of the basic concepts of surface irrigation in the past is that there must be surface runoff if the lower ends of the borders or furrows are to be irrigated so as to replace most of the available water. In many areas, because of declining water tables and limited water supplies, surface runoff is being restricted or prohibited. The general approach in coping with this restriction was to install recirculating or pump back systems or to physically prevent surface runoff. Recirculating systems are required in some areas. The challenge to designers who are restructuring surface systems is to consider variable grade systems that can increase the head to tail end uniformity of water application with little or no surface runoff. The concept is not new (Powell et al., 1972). However, such a design can be very complex when considering varying infiltration rates with or without furrows, the variation in vegetative growth in border strips and some row crops during the year, and the special requirements of some crops. Again, there will be need for more sophisticated analytical tools to address these complex relationships. Chapters 13 and 14 address these issues and provide the concepts and tools necessary for meeting these design challenges. Chapters 15-17 address the design of sprinkler and microirrigation systems. In developing countries, the stability of a community organization is essential for sustained and productive irrigated agriculture. Rapidly increasing populations and relatively stable renewable water supplies often involve complex social, food security, and natural resources issues. Addressing these issues will be a major challenge to designers of new or improved irrigation systems to consider, especially in developing countries.

2.8 SUMMARY The factors to be considered in the design and operation of farm irrigation systems have changed greatly since the first edition of this book was published in 1980. In this chapter, I have attempted to highlight key factors that will influence the design and operation of new irrigation systems and the upgrading of old systems in order to make irrigated agriculture more productive and more sustainable. Some of these factors are renewable water supplies, coping with droughts, adequacy of drainage, control of salinity, the productivity of water and water use efficiency, new management concepts such as deficit irrigation, and environmental concerns about surface runoff. I have cited references where readers can find additional details on the problems and issues that will need to be addressed by irrigation system designers, trainers providing irrigation management guidance, and operators of irrigation systems. Sustainable and productive irrigated agriculture will require adopting new technology and management concepts and increased emphasis on the productivity of irrigation water.

52

Chapter 2 Sustainable and Productive Irrigated Agriculture

REFERENCES Abbott, C. 2003. Integrated irrigation and drainage to save water—“controlled drainage.” Water, Dept. of Int’l. Devel. 16: 6. Alfaro, J. F., E. E. Griffin, J. Keller, G. R. Hanson, J. L. Anderson, G. L. Aschroft, and E. A. Richardson. 1974. Preventive freeze protection by preseason sprinkling to delay bud development. Trans. ASAE 17: 1025-1028. Anderson, J. L. M. 1974. Effects of over tree sprinkling on fruit bud development and fruit quality. Mimo. Report. Logan, Utah: Plant Sci. Department, Utah State Univ. Barfield, B. J., and J. F. Gerber, eds. 1979. Modification of the Aerial Environment of Crops. St. Joseph, Mich.: ASAE. Bauder, J. W., A. Bauer, J. M. Ramirez, and D. K. Cassel. 1978. Alfalfa water use and production on dryland and irrigated sandy loam. Agron. J. 70: 95-99. Birley, M. H. 1989. Guidelines for forecasting the vector-borne disease implications of water resources development. PEEM Guidelines Series 2, VBC/89.6. Geneva, Switzerland: World Health Organization. Bos, M. G. 1980. Irrigation efficiencies at the crop level. ICID Bulletin, Int’l. Comm. Irrig. and Drain. 29(2): 18-25, 60. Braud, H. J. Jr., and P. L. Hawthorne. 1965. Cold protection for Louisiana strawberries. Bulletin No. 591, March. Baton Rouge, La.: Louisiana State Univ. Braud, H. J., and M. Esphahani. 1971. Direct water spray for citrus freeze protection. ASAE Paper No. 71-234. St. Joseph, Mich.: ASAE. Carolus, R. L. 1971. Evaporative cooling techniques for regulating plant water stress. Hort. Sci. 6: 23-25. CAST. 1992. Preparing U.S. Agriculture for Climatic Change. Ames, Iowa: Council for Agricultural Science and Technology. Chesness, J. L., L. A. Harper, and T. A. Howell. 1979. Sprinkling for heat stress reduction. In Modification of the Aerial Environment of Crops, 388-399. B. J. Barfield, and J. F. Gerber, eds. St. Joseph, Mich.: ASAE. CSIRO. 2002. New framework for water. Media Release—Ref 2002/200. Clayton South VIC, Australia: CSIRO. Edwards, C. A., R. Lal, P. Madden, R. H. Miller, and G. House, eds. 1990. Sustainable Agricultural Systems. Ankeny, Iowa: Soil and Water Conservation Society. Ellis, W. S. 1990. The Aral: A Soviet sea lies dying. National Geographic 177(2): 7392. English, M. J., K. H. Solomon, and G. J. Hoffman. 2002. A paradigm shift in irrigation management. J. Irrig. Drain. Eng. 128: 267-277. Falkenmark, M., and C. Widstrand. 1992. Population and Water Resources: A Delicate Balance Pop. Bulletin 47(3). Washington, D.C.: Population Reference Bureau, Inc. FAO (Food and Agriculture Organization). 1989. Sustainable Development and Natural Resources. Conference, C 89/2 – Sup. 2, Nov., 54 pp. FAO. FAO (Food and Agriculture Organization). 2002. Deficit irrigation practices. FAO water reports No. 22. Rome, Italy: FAO. Ferguson, R. B., D. E. Eisenhauer, T. L. Bockstadter, D. H. Krull, and G. Buttermore. 1990. Water and nitrogen management in central Platte Valley of Nebraska. J. Irrig. Drain. Eng. 116(4): 557-565.

Design and Operation of Farm Irrigation Systems

53

Gleick, P. 2002. The World’s Water, 2002-2003: Biennial Report on Freshwater Resources. Washington, D.C.: Island Press, with contributions from the staff of the Pacific Institute. Gray, A. S. 1970. Environmental control using sprinkler systems. In Nat’l. Irrigation Symp., RR1-RR10. Grimes, D. W., and K. M. El-Zik. 1990 Cotton. In Irrigation of Agricultural Crops, 741-773. Agron. Series 30. B.A. Stewart, and D. R. Nielsen, eds. Madison, Wis.: Am. Soc. Agron. Harrison, D. S., and J. F. Gerber. 1965. Research with sprinkler irrigation for cold protection. Trans. ASAE 7(4): 464-468. Harrison, D. S., J. F. Gerber, and R. E. Choate. 1974. Sprinkler irrigation for cold protection. Cir. 348 Tech. Gainesville, Fla.: Florida Coop. Ext. Serv., Univ. of Florida. Hatfield, J. L., and D. L. Karlen, eds. 1993. Sustainable Agricultural Systems. Boca Raton, Fla.: Lewis Publishers. Heaven, S., G. B. Koloskov, A. C. Lock, and T. W. Tanton. 2002. Water resource management in the Aral Basin: A river basin model for the Syr Darya. Irrig. and Drain. 51: 109-118. Hoffman, G. J., T. A. Howell, and K. H. Solomon, eds. 1990. Management of Farm Irrigation Systems. St. Joseph, Mich.: ASAE. Howell, T. A. 1990. Relationships between crop production and transpiration, evapotranspiration, and irrigation. In Irrigation of Agricultural Crops, 391-434. B.A. Stewart and D.R. Nielsen, eds. Agron. Series 30. Madison, Wis.: American Soc. Agron. Howell, T. A. Irrigation’s role in enhancing water use efficiency. 2000. In Nat’l. Irrigation Symp., 66-80. R. G. Evans, B. L. Benham, and T. P Trooien, eds. St. Joseph, Mich.: ASAE. Howell, T. A., R. H. Cuenca, and K. H. Solomon. 1990. Crop yield response. In Management of Farm Irrigation Systems, 93-122. G. T. Hoffman, T. A. Howell, and K. H. Solomon, eds. St. Joseph, Mich.: ASAE. Howell, T. A., A. D. Schneider, and B. A. Stewart. 1995. Subsurface and surface microirrigation of corn—Southern High Plains. In Microirigation for a Changing World, 375-381. F. R. Lamm, ed. St. Joseph, Mich.: ASAE. IBRD. 1992a. The World Bank and the Environment, Fiscal 1992. 3rd Annual Report, Washington, D.C.: The World Bank. IBRD. 1992b. World Development Report 1992: Development and the Environment. New York, N.Y.: The World Bank, Oxford Univ. Press. ICID. 1994. Aral Sea Basin. In Special Session, Report, Int’l. Comm. on Irrig. and Drain. New Delhi, India: ICID. Jensen, M. E. 1984. Water resource technology and management. In Future Agricultural Technology and Resource Conservation, 142-146, B. C. English et al., eds. Ames, Iowa: Iowa State Univ. Press. Jensen, M. E. 1990. Arid lands, impending water-population crisis. In Hydraulics/Hydrology of Arid Lands, Proc. Int’l. Symp., 14-19. R. H. French, ed. New York, N.Y.: American Soc. Civil Eng.

54

Chapter 2 Sustainable and Productive Irrigated Agriculture

Jensen, M. E. 1993. The Impacts of Irrigation and Drainage on the Environment. 5th Gulhati Memorial Lecture, Int’l. Comm. on Irrig. and Drain. The Hague, The Netherlands: ICID. Kidder, E. H. 1970. Climate modification with sprinklers. In Natl. Irrig. Symp., V1V6. Kimball, L. A. 1992. Forging International Agreement: Strengthening Intergovernmental Institutions for Environment and Development. Washington, D.C.: World Resources Institute. Kipness, T., I. Vaisman, and I. Granoth. 1989. Drought stress and alfalfa production in a Mediteranean environment. Irrig. Sci. 10: 113-125. Kirkham, M. B., ed. 1999. Water Use in Crop Production. Binghamton, N.Y.: Food Products Press. Co-published as J. Crop Production, Vol. 2, No. 2(#4). Lamade, W. 1968. Frost control on potatoes. Veg. Crop Mgmt. 4(4): 27-41. Letey, J. 1993. Relationship between salinity and efficient water use. Irrig. Sci. 14: 7584. Ley, T. W. 1995. Weather stations C Satellites to scientific irrigation scheduling. Irrigation III(4): 20-23. Lico, M. S. 1992. Detailed study of irrigation drainage in and near wildlife management areas, West-Central Nevada, 1987-90. U.S. Geological Survey Water Resources Investigations Report 92-4024A, Part A. Carson City, Nev.: USGS. Locascio, S. J., D. S. Harrison, and V. P. Nettles. 1967. Sprinkler irrigation of strawberries for freeze protection. Proc. Florida State Hort. Soc. 80: 208-211. Merva, G. E., and C. Vandenbrick. 1979. Physical principles involved in alleviating heat stress. In Modification of the Aerial Environment of Crops, 373-387. B. J. Barfield, and J. F. Gerber, eds. St. Joseph, Mich.: ASAE. Micklin, P. P. 1988. Desiccation of the Aral Sea: A water management disaster in the Soviet Union. Science 241: 1170-1176. Molden, D. 1997. Accounting for water use and productivity. IWMI/SWIM Paper No. 1. Colombo, Sri Lanka: Int’l. Water Mgmt. Inst. Molden, D. J., M. el Kady, and Z. Zhu. 1998. Use and productivity of Egypt’s Nile water. In Contemporary Challenges for Irrigation and Drainage, USCID 14th Tech. Conf., 99-116. J. I. Burns, and S. S Anderson, eds. Denver, Colo.: USCID. Molden, D., and R. Sakthivadivel. 1999. Water accounting to assess use and productivity of water. Int’l. J. Water Resources Development 15(1): 55-71. Norwood, C. A. 1995. Comparison of limited irrigated vs. dryland cropping systems in the U.S. Great Plains. Agron. J. 87: 737-743. NRC (National Research Council). 1989. Irrigation-Induced Water Quality Problems. Washington, D.C.: Committee on Irrigation-Induced Water Quality Problems, Water Science and Technology Board, National Research Council. Washingtion, D.C.: National Academy Press. NRC (National Research Council). 1992. Water Transfers in the West: Equity, Efficiency, and the Environment. Washington, D.C.: Committee on Western Water Management, Water Science and Technology Board, National Research Council. Washingtion, D.C.: National Academy Press. NRC (National Research Council). 1993. Toward Sustainability, Soil and Water Research Priorities for Developing Countries. Washingtion, D.C.: National Academy Press.

Design and Operation of Farm Irrigation Systems

55

NRC (National Research Council). 1996. A New Era for Irrigation. Washington, D.C.: Committee on the Future of Irrigation in the Face of Competing Demands, National Research Council. Washingtion, D.C.: National Academy Press. Parsons, L. R., and T. A. Wheaton. 1995. Freeze protection of Florida citrus with microsprinkler irrigation. In Microirrigation for a Changing World, 25-30. F. R. Lamm, ed. St. Joseph, Mich.: ASAE. Pereira, L. S., J. R. Gilley, R. A. Feddes, and B. Lesaffre, eds. 1996. Sustainability of Irrigated Agriculture. Dordrecht, The Netherlands: Kluver Academic Publishers. Perry, C. 2002. Keynote address. Wallingford, UK. Personal communication. Population Reference Bureau. 2005. 2005 World population data sheet-Aug. Available at: www.prb.org. Powell, G. M., M. E. Jensen, and L. G. King. 1972. Optimizing surface irrigation uniformity by nonuniform slopes. ASAE Paper No. 72-721. St. Joseph, Mich.: ASAE. Roos, M. 2001. How do we determine the real amount of water available for transfer from one basin to another. In Transbasin Water Transfers, Proc. USCID Water Mgmt. Conf., 245-255. J. Schaack, ed. Denver, Colo.: USCID. Rosegrant, M. W., X. Cai, and S. A. Cline. 2002. World Water and Food to 2025: Dealing with Scarcity. Washington, D.C.: Int’l. Food Policy Res. Sammis, T. W. 1981. Yield of alfalfa and cotton as influenced by irrigation. Agron. J. 73: 323-329. Seckler, D., U. Amarasinghe, D. Molden, R. de Silva, and R. Barker. 1998. Water demand and supply 1990 to 2025: Scenarios and issues. Colombo, Sri Lanka: Res. Report 19. Int’l. Water Mgmt. Inst. Sneed, R. E. 1970. Using sprinkler irrigation for crop cooling. Sprinkler Irrig. Assoc. May: 5-7. Sneed, R. E. 1972. Frost control. Sprinkler Irrig. Assoc. Proc., Annual Tech. Conf., 41-53. Church Falls, Va.: The Irrigation Assoc. Stewart, J. I., R. M. Hagan, R. E. Danielson, R. J. Hanks, and E. B. Jackson et al. 1977. Optimizing Crop Production Through Control of Water and Salinity Levels in the Soil. Logan, Utah: Utah Water Res. Lab., Utah State Univ. Stewart, B. A., and D. R. Nielsen, eds. 1990. Irrigation of Agricultural Crops. Agron. Series 30. Madison, Wis.: American Soc. Agron., Madison. Sun, M. 1988. Environmental awakening in the Soviet Union. Science 241: 10331035. Tiffen, M. 1989. Guidelines for the incorporation of health safeguards into irrigation projects through intersectoral cooperation. PEEM Guidelines Series 1, VBC/89.5. Geneva, Switzerland: World Health Organization. Unger, P. W., and T. A. Howell. 1999. Agricultural water conservation: A global perspective. In Water Use in Crop Production, 1-36. M. B. Kirkham, ed. Binghamton, N.Y.: The Haworth Press, Inc. Unrath, C. R. 1972a. The evaporative cooling effects of over tree sprinkler irrigation on ‘Red Delicious’ apples. J. Am. Soc. Hort. Sci. 97(1): 55-58. Unrath, C. R. 1972b. The quality of ‘Red Delicious’ apples as affected by over tree sprinkler irrigation. J. Am. Soc. Hort. Sci. 97(1): 58-61.

56

Chapter 2 Sustainable and Productive Irrigated Agriculture

Unrath, C. R., and R. E. Sneed. 1974. Evaporative cooling of ‘Delious Apples’: The economic feasibility of reducing environmental heat stress. J. Am. Hort. Soc. 99(4): 372-375. USDA-NASS (U.S. Department of Agriculture National Agricultural Statistics Service). 1999. 1997 Census of Agriculture. Vol. 3: Special Studies. 1998 Farm and Ranch Irrigation Survey. Available at: www.nass.usda.gov/census/census97/fris/fris.htm. USDA-NASS (U.S. Department of Agriculture National Agricultural Statistics Service). 2004. 2002 Census of Agriculture. Vol. 3: Special Studies Part 1: Farm and Ranch Irrigation Survey (2003). Available at: www.nass.usda.gov/census/census02 /fris /fris03.htm. Viets, F. G. 1962. Fertilizer and efficient use of water. Adv. in Agron. 14: 223-264. Wright, J. L. 1988. Daily and seasonal evapotranspiration and yield of irrigated alfalfa in southern Idaho. Agron. J. 80: 662-669.

CHAPTER 3

PLANNING AND SYSTEM SELECTION Kenneth H. Solomon (California Polytechnic State University, San Luis Obispo, California) A. M. El-Gindy (Ain-Shams University, Cairo, Egypt) Sagit R. Ibatullin (Research Institute of Water Economy,Taraz City, Kazakhstan) Abstract. This chapter reviews the process of planning for and selecting an irrigation system. Decision making involves the identification and evaluation of alternatives, and making and implementing the decision. The initial problem situation is improved through the cause-and-effect feedback loop, and a learning feedback loop influences future decision making. Irrigation system selection involves both objective and subjective components of problems, goals, and values, and available augmentations to the problem situation. An irrigation system in the broadest sense includes not only hardware, such as equipment, land improvements, water conveyance, and control structures, but technology and support infrastructure as well. These may include guides for equipment installation, operation, and maintenance, and plans for water management and irrigation-related crop husbandry, training, and extension. Planning for an irrigation system must consider a wide range of physical, economic, human, and social factors. Irrigation system selection must consider the capabilities, costs, and limitations of potential irrigation methods. This chapter presents a brief review of such factors for the most common surface, sprinkler, and microirrigation systems. Keywords. Decision making, Irrigation systems, Microirrigation, Planning, Selection, Sprinkler irrigation, Surface irrigation.

3.1 INTRODUCTION The process of planning for, and selection of, an irrigation system should be based upon an understanding of the goals and objectives to be met, a study of the resources available, a creative determination of the possible alternatives, and an evaluation of these alternatives according to economic and other criteria. This chapter presents factors to be addressed during these activities. Conceptual models for decision making and for irrigation system selection provide a framework for the review of individual factors. Irrigation system components beyond just hardware must be considered. Water supply and system capacity must be sufficient to meet requirements. Economic

58

Chapter 3 Planning and System Selection

evaluation of irrigation options depends on the perspective of the evaluator and the goals and values of all those involved with the irrigation development. Capabilities, limitations, and evaluation factors for the most common types of irrigation systems are briefly discussed. 3.1.1 A Conceptual Model for Decision Making The decision-making process is illustrated in Figure 3.1. The process starts with a problem situation to be improved or solved. Therefore, a decision must be made. The situation at hand contains not only the actuality of the problem, but the potentiality of a number of alternative courses of action to improve the situation or solve the problem. These action alternatives exist whether or not we are clever enough to recognize them. Experience, judgment, and creativity increase our ability to identify the alternatives that exist. Part of this process is to recognize that certain constraints are acting on the situation with which we must deal. Notice that in the example of Figure 3.1, only alternatives B, C and D were identified. Alternatives A and E are in fact possibilities, but they were not thought of. Next, value system(s) are applied to evaluate those alternatives that have been identified. Though it may be helpful to quantify these evaluations as much as possible, value systems contain non-quantifiable elements. So, the final decision must take into account not only the quantitative score for each alternative, but some qualitative judgments as well. In the diagrammed example, for instance, alternative B was selected over C, even though C has a higher quantitative score: apparently the intangibles involved outweigh the six-point difference in score. Like many human endeavors, decision making is a cyclical process. In this case, two main feedback loops are identified. The primary feedback loop is the cause and effect loop. This is the way in which the chosen action will alter the current situation. Other aspects of the situation may be altered as well. The results of our decision will influence the list of alternatives for further action, either with regard to this problem or some future one. Situation

Problem

Identify Alternatives

Action Alternatives A B C D E

Evaluate Alternatives Alt. Score B C D

B C D

Make Decision Select Reject Reject

80 86 62

Value System(s)

Experience Judgment Creativity Constraints

B C D

Monitor Results of Decision Learning Process

Cause and Effect

Figure 3.1. A conceptual model of the decision process.

Design and Operation of Farm Irrigation Systems

59

An additional feedback loop is the learning process. By monitoring the results of a decision, we learn things. This learning increases our experience, judgment and perhaps creativity. It may alter the constraints (or the way we perceive the constraints) of the situation. Learning may allow us to identify action alternatives that we did not see before—either for the problem just considered, or for future problems. The learning process will also have an effect on our value systems. Learning, living, and growing are inseparable and go hand in hand with the evolution of our value systems. Thus the learning process will affect the way in which we evaluate alternatives and make future decisions. 3.1.2 A Conceptual Model for Irrigation System Selection Decision making within an irrigation context, based on our understanding of the decision process in general, is illustrated in Figure 3.2. The selection process is drawn onto a conceptual blackboard, divided into four quadrants: right to left into Objective vs. Subjective; and top to bottom into the Existing Situation vs. things Brought to the Situation. (Note that the items listed under each general heading are for illustration only. The lists are not intended to be exhaustive, nor will all items listed necessarily apply to any given situation.) The objective part of the existing situation contains resources and problems. The resource category is subdivided into physical and people-related resources. Problems may be associated with resources (amount, timing, place, quality) or with constraints of various types which prevent some otherwise desirable action. The subjective part of the existing situation contains the goals and priorities we are trying to meet with this irrigation decision, and the values by which the alternatives and results will be judged. To the existing situation we bring items that we hope will solve the problems observed; the existing situation is augmented. This augmentation always involves several components: physical things such as irrigation equipment; people-related things such Objective Resources

Subjective

Problems

Goals & Priorities

Physical

People Resources: Not Enough Irrigators/Others Soil/Water Too Much Skills Climate Bad Timing Experience Cultivars Poor Quality Education Energy Constraints Infrastructure Motivation Markets Financial

Augmentation Irrigation

Other

Equipment Management Plans Training Programs New Problems

Equipment Management Plans Training Programs New Problems

Results Augmented Situation Allows New Matchup of Problems & Capacities for Solution

Food Nutrition Jobs Exports Income

Values Profit Health Employment Independence Technology Tradition

New Net Benefits + Advantages – Disadvantages

Figure 3.2. A conceptual model for irrigation system selection. The items listed under each general heading are for illustration only. The lists are not intended to be exhaustive, nor will all items listed necessarily apply to any given situation.

60

Chapter 3 Planning and System Selection

as management plans and training; and inevitably, we bring along some new problems to be dealt with. For example, irrigation adds a drainage requirement, or advanced equipment adds training and maintenance requirements. The results of our actions are formed out of the mix of the elements in the existing situation and the items brought to the situation in the augmentation. The augmented situation includes new capacities for solving the original problem, as well as new problems. However, the results box in Figure 3.2 is not the end. The results should be viewed in light of the goals and priorities, and the values of the decision makers to determine the net benefits. The net benefits box spans both the objective and subjective halves of the conceptual blackboard, reminding us (as in the model for decisions) that what we seek has both quantitative and qualitative aspects. This is the framework in which irrigation system selection should take place. These models do not constitute a selection process itself, but are rather a reminder of all the things beyond the engineering details that must be considered. 3.1.3 An Irrigation System—More Than Just Hardware An irrigation system is commonly thought of in terms of its hardware aspects: infield equipment, water conveyance and control structures, and land improvements such as grading. It is important to realize, though, that a complete irrigation system includes additional aspects of technology and support infrastructure. A given technology includes not only relevant hardware, but the collection of knowledge necessary to accomplish the desired function. Technology includes facts, relationships, and rules for making decisions. In order to accomplish its function, an irrigation system must include equipment and a variety of related “software.” Specifically, the irrigation technology package must include guides for design, installation, operation, and maintenance of the equipment; plans for irrigation scheduling and water management; and plans for the crop husbandry practices impacted by irrigation. It must also include plans for irrigation-related training, research, development, and extension. Just as equipment requires operational knowledge and plans, it requires a support infrastructure as well. This infrastructure is partly physical and partly institutional. Physical elements include dams, canals, offices, and facilities for research and demonstration. Institutional elements include water rights establishment and enforcement bodies, irrigation districts, farmer cooperatives, repair and maintenance services, and extension organizations. The infrastructure is required to make available the water, other agronomic inputs, services, and technological information necessary for successful irrigation. In developed irrigation markets, the widespread existence and availability of technical irrigation information and infrastructure may lead to the misconception that irrigation system selection consists solely of equipment selection. Consider, for example, the introduction of microirrigation to a developing irrigation market. In this case, it is obvious that a choice of microirrigation equipment without a matching choice to fund and implement a suitable package of microirrigation information, plans, and the necessary support infrastructure is doomed to failure.

Design and Operation of Farm Irrigation Systems

61

3.2 PLANNING FOR IRRIGATION Planning for irrigation consists of identifying and collecting information on relevant factors, followed by formulating and evaluating realistic options for irrigation systems. “System” here is taken in the broad sense, including elements of equipment, technical information, and infrastructure. As indicated in Figure 3.2, the initial planning phase certainly includes inventories of resources, problems, and the goals, priorities, and values of the decision makers. Table 3.1 lists a number of considerations that may have a bearing on irrigation decisions. A few of the key points listed are discussed further in the following sections. 3.2.1 Water Requirements and Water Supply A key point to consider in any irrigation plan is whether the water available is sufficient to adequately meet the water requirements for crop production. This comparison must be made in terms of both the total volume of water required and available during the year, and the peak rate at which water is used and supplied. The water supply must meet or exceed the minimum requirement less any effective precipitation. Effective precipitation is that portion of total precipitation which becomes available to support plant growth (ASAE, 1998a). The timing and rate of precipitation can influence its availability for plant growth. Rainfall during the growing season is ineffective if it runs off or percolates below the root zone. Precipitation falling outside the growing season is effective only to the extent that it is stored in the eventual crop root zone. Hence, precipitation resulting in runoff, deep percolation, or evaporation prior to the season is not effective. Precipitation and deep percolation occurring prior to the season may be effective in meeting part of the leaching requirement (see Section 3.2.2). The total volume of water required to produce a crop includes the water consumed during crop growth, water used to maintain a favorable salt balance in the root zone, and water used for certain husbandry practices such as germination, climate control, and beneficial vegetation such as windbreaks or cover crops. If the water supply does not meet the water volume requirement (reduced by any effective precipitation), irrigation options are limited. Either the supply must be augmented or the production plan altered, for example by using shorter-season varieties, eliminating water-consuming husbandry practices, or using a deficit irrigation schedule. The highest rate of crop water use depends both on climatic factors and on the timing of crop growth stages. Because weather varies from year to year, the peak-use rate will vary as well. It is expected that the peak-use rate determined from average values of weather variables will be exceeded one year in two. It may be prudent to study historical weather and base plans on a more conservative figure, for example the peak-use rate exceeded only one year in five, or one year in ten. The costs of both water supply and infield irrigation systems are sensitive to the peak flow rate, so it may not be economical to design for abnormally high peak-use rates that are expected to occur only rarely. The water supply and in-field irrigation systems should have the capacity to meet the design peak-use rate, adjusted to compensate for water application efficiency, and possibly adjusted again according to the strategies discussed in the following paragraphs. Note that application efficiency may vary throughout the year. For example, with surface irrigation systems, it may be difficult to uniformly apply the small amounts required by crops during early growth stages, thus reducing application efficiency. The application efficiency anticipated during the peak-use period should be used in this adjustment.

62

Chapter 3 Planning and System Selection

Table 3.1. Factors for consideration in irrigation decision making. Physical Factors Economic Factors Crops, crop rotation Crop value (price) Crop yields (especially if quality or quantity Investment capital differs for different irrigation systems) Foreign exchange, hard currency Cultural practices Credit Soils Source & terms Texture, depth, uniformity Interest rate Intake rate, water holding capacity For long term debt Erosion potential For operating loans Salinity Cash flow requirements Internal drainage Equipment life Topography Costs & inflation Water supply Equipment Water rights Water, energy Source, delivery schedule Other agricultural inputs Quantity available, reliability Services Current uses of water Labor, various skill levels Water quality Supervision Salinity & other chemical constituents Management Suspended solids Operation & maintenance Climate (wind, heat, frost) Repairs & replacement Land value, availability Warranties Flood hazard Incentives and subsidies Catastrophic weather events Taxes Water table Insurance Energy availability, reliability, and form Opportunity costs for inefficiencies Pests Markets, export/import Infrastructure available Impact of irrigation on related field and Water supply & conveyance factory operations Agronomic inputs Potential for upgrades as conditions change Equipment repair service Uncertainty Technological support Social Factors General values Weather data Specific goals and priorities Existing standards Legal constraints Equipment performance Political issues Design, installation Local cooperation and support Practices Local and governmental expectations Common good practice Environmental issues Human Factors Labor Quality standards Availability Wildlife habitat Skills, experience Mitigation Education Health issues Potential for vandalism or theft History of irrigation experience Level of automatic control desired Biases and taboos

Design and Operation of Farm Irrigation Systems

63

Certain management strategies may allow additional adjustments in the design capacity of water supply and irrigation systems. Probably the most common of these is an increase in the design capacity to create an allowance for downtime, that is, time when the system is not operated so that it may be serviced, repaired, or maintained. For example, ASAE (1998b) recommended that the design capacity for microirrigation systems be sufficient to meet the peak-use rate in about 90% of the time available, or when operating no more than 22 hours per day. In some cases the schedule of electric power rates will impact downtime decisions. Another strategy is to plan for the use of stored soil water to meet the crop water needs during the peak-use period. Successful use of this strategy requires that the irrigation system be operated so that the available soil water reservoir is full (or nearly full) prior to entering the peak-use period. Peak-use water needs can be met by deliveries from the irrigation system and by planned depletion from the root zone. The amount that the design flow rate for the irrigation system can be decreased by this strategy depends on the amount of available water that can be stored in the crop root zone, and the duration of the peak-use period. This strategy carries the risk that an abnormally long peak-use period or unexpected system failures during or just prior to the peak-use period can result in unplanned stress to the crop due to insufficient water. In those areas where rainfall may be expected during the peak-use period, agronomic practices to increase the effectiveness of precipitation may reduce the required flow rate for irrigation and supply systems. Tillage methods that increase temporary surface storage of rainfall can increase effective precipitation by reducing runoff. Irrigation scheduling methods that don’t completely refill the crop root zone can increase effective precipitation by reducing deep percolation. In cases where the water supply cannot match the peak-use rate, water storage in a pond or reservoir may solve the problem as long as the water supply volume is sufficient to meet the total water volume required by the crop. During the off-peak periods, water in excess of crop needs is diverted to storage. During the peak-use period, the normal water supply is augmented from stored water. Any action that increases the attainable application efficiency, such as the use of return-flow systems to capture and re-use surface irrigation runoff water, can help to meet design peak-use needs. For a given gross application rate, increasing the application efficiency will increase the system’s net application rate. The advisability of these various water management strategies is frequently affected by the value of the crop and the sensitivity of the crop yield to water. 3.2.2 Need for Drainage Irrigation water inevitably contains salt, which is left behind when plants take water from the root zone for evapotranspiration. Thus, salts tend to accumulate unless leached away by excess water passing through the root zone. In humid regions the excess may be provided by rainfall, but in arid regions additional irrigation water (in excess of the crop’s consumptive use) is required for the purpose of leaching. Unless this excess or leaching water is removed, high water tables and waterlogged soils can result. High water tables can lead to soil salination due to upward movement of water and salt. Drainage, natural or artificial, is necessary to carry away excess or leaching water applied to maintain a favorable salt balance in the root zone. Irrigation planning should include an investigation of the potential need for artificial drainage. Equipment, technology, and infrastructure for the drainage system, if

64

Chapter 3 Planning and System Selection

needed, should be included in the economics and evaluation of irrigation options. 3.2.3 Irrigation Economics An important part of irrigation planning and system selection involves economic evaluation of the alternatives. It is tempting to consider only the most obvious, initial costs, but this is too simplistic an approach for sound decision making. Life cycle costing, which treats operation and maintenance costs as well as initial costs, is the preferred approach. The basic price of the equipment or improvement (such as land grading) is not the only initial cost. To this may be added sales tax, import duties or tariffs, and initial loan fees if the purchase is to be financed with a loan. Freight and delivery charges may also apply, and can be particularly important when considering systems utilizing imported equipment. Freight charges are usually based on the weight of the material to be shipped and the volume it occupies. Equipment assembly and installation may involve additional initial expenses. Important initial costs that are sometimes overlooked are expenses for a stock of spare parts and for initial training. Particularly in remote areas or in developing markets without a fully mature maintenance and repair infrastructure, initial equipment purchases should include an inventory of spare parts. An initial effort to train irrigators and equipment operators may be necessary if the anticipated benefits of the chosen irrigation system are to be achieved. The energy provider may assess an initial charge to bring service (particularly electric power lines) to the farm site. Annual costs to operate and maintain an irrigation system include charges for labor, water, and energy. Allowances for preventative maintenance and repairs, repair parts, taxes, insurance, and ongoing training should also be included in the annual operating budget. Different irrigation options often involve trade-offs between initial and annual operating costs. For example, larger pipe sizes have higher initial costs, but reduce friction losses and hence may lower operating costs for pumping energy. Less equipmentintensive systems may have reduced initial costs, but higher operating costs for irrigation labor. In order to compare the full, life cycle cost of various irrigation options, some method is needed to place initial and annual irrigation costs on the same basis. A common approach is to convert the initial costs to an equivalent annual cost. This is done by multiplying the initial cost by a factor (often called the capital recovery factor) that depends on the interest rate and the expected economic life for the initial cost item. The capital recovery factor may be calculated according to Equation 3.1: CRF =

i( 1 + i )n ( 1 + i )n − 1

(3.1)

where CRF = capital recovery factor, dimensionless i = annual interest rate, decimal n = expected economic life, years Table 3.2 lists economic lives for different irrigation items and conditions. Increasing the interest rate or reducing the economic life will increase the capital recovery factor and the annualized cost. The total annualized cost for an irrigation option is the sum of its annualized initial cost and its annual operating cost.

Design and Operation of Farm Irrigation Systems

65

Table 3.2. Expected economic lives for irrigation equipment (years). Adapted from McCulloch et al. (1967). Item Maximum[a] Conservative[b] Difficult[c] Irrigation systems Side roll 20 15 12 Center pivot, linear move, LEPA 25 20 15 Surface irrigation systems 30 20 15 Individual items Farm reservoirs (heavy silting) 25 20 15 Farm reservoirs (light silting) 50 35 25 Well 40 25 20 Irrigation pump 20 15 10 Pump power units Diesel 15 12 8 Gas, propane 13 9 7 Gasoline 8 6 5 Electric motor 30 25 15 Water conveyance Open ditches (unlined) 8 6 4 Concrete structures 25 20 15 Concrete pipe systems 20 15 10 Wood flumes 10 8 6 Large aluminum pipe (lightweight) 10 8 6 Coated steel pipe (underground use) 20 15 10 Coated Steel pipe (surface use) 12 10 8 Galvanized steel pipe (surface use) 25 20 15 PVC pipe (underground use) 40 20 18 Sprinkler equipment Aluminum sprinkler laterals 25 15 10 Sprinklers (medium size) 15 12 10 Sprinklers (giant, large-volume guns) 13 10 8 Microirrigation equipment Filters 20 15 10 Emitters, hose (permanent) 15 10 8 Thin-walled tape 3 1 1 Surface irrigation Land grading 40 25 20 [a] [b] [c]

Assumptions regarding maintenance, operating and environmental conditions: [a] excellent maintenance, ideal operating conditions, and favorable environmental circumstances; [b] typical or ordinary conditions; [c] inadequate or infrequent maintenance, harsh operating or environmental conditions.

The current price for an initial or annual cost item may not always be the most appropriate figure to use in an economic analysis. Subsidies, incentives, taxes, or governmental policies may result in prices that do not reflect an item’s true cost or value. The correct number to use in a comparative analysis can often be determined only after considering who is paying the costs, receiving the benefits, or making the decisions.

66

Chapter 3 Planning and System Selection

For example, consider the price and value of water. If, according to local policy, the government pays all costs of water supply development and conveyance, the price of water to the farmer may be zero and private irrigation decisions may be based on that premise. For public irrigation projects, those sponsored or funded by the government or financed by international organizations (for example, a World Bank loan), the use of a zero water cost is questionable. In this case, a more appropriate water cost might include allocations for at least the operation, and perhaps the facilities, of the water delivery system. Further, perhaps credits for the recreational and environmental benefits of water development, and charges for environmental degradation or the cost of mitigation measures to avoid it, should be included as well. Costs for other inputs to irrigated agriculture are subject to similar considerations. While private decision making may be based on local prices, full cost, world market cost, or even opportunity cost (net return possible by diverting that input to alternative uses), values should be considered for public project decision making. Note also that financial or resource constraints may prevent the selection of the “most economical” irrigation option. Limitations on available capital, or credit limits, may preclude some capital-intensive options, even if these would otherwise be the best option economically. Constraints on available resources, such as land or water, may have a similar effect on the decision process. In these instances, it may be helpful to consider indicators such as the maximum net return per unit of constrained resource (for example, maximum net return per unit land area, per unit water volume, or per unit of capital invested) instead of or in addition to the usual economic indicators.

3.3 IRRIGATION SYSTEM SELECTION To do a proper job of system selection, one must give careful consideration to the capabilities and limitations of potential irrigation methods. The remainder of this chapter will summarize such considerations for the most common types of irrigation systems. Labor, management, energy, and economic factors relevant to each system type are briefly addressed. This section is adapted from Solomon (1988), which was based on an unpublished study by Anderson et al. (1987). U.S. costs cited are representative of U.S. irrigation practice as of 1996, and are based on Solomon et al. (1992), Westlands Water District (1988), as updated by the study of Jorgensen (1996). Egyptian cost figures are based on studies conducted by the second author (El-Gindy) in 1995, based on a field size of 20 ha. 3.3.1 Surface Irrigation 3.3.1.1 Types of surface irrigation. Basin, border strip, and furrow irrigation are common types of surface irrigation. In basin irrigation, water is applied to a completely level (sometimes called “dead level”) area enclosed by dikes or borders. This method of irrigation is used successfully for both field and row crops. The floor of the basin may be flat, ridged, or shaped into beds, depending on crop and cultural practices. Basins need not be rectangular or straight sided, and the border dikes may or may not be permanent. Basin irrigation is also called by a variety of other names: check flooding, level borders, check irrigation, check-basin irrigation, dead-level irrigation, and level-basin irrigation. Factors affecting basin geometry are available water stream size, topography, soil properties, and degree of leveling required. Basins may be quite small or as large as 15 ha or so. Level basins simplify water management, since the irrigator need only

Design and Operation of Farm Irrigation Systems

67

supply a specified volume of water to the field. With adequate, non-erosive stream size, the water will spread quickly over the field, minimizing non-uniformities in inundation time. Basin irrigation is most effective on uniform soils, precisely leveled, when large stream sizes (relative to basin area) are available. High efficiencies are possible with low labor requirements. Border strip irrigation uses land formed into graded strips, level across the narrow dimension but sloping along the long dimension, and bounded by ridges or borders. Water is turned into the upper end of the border strip and advances down the strip. After a time, the water is turned off, and a recession front, where standing water has soaked into the soil, moves down the strip. High irrigation efficiencies are possible with this method of irrigation, but may be difficult to obtain in practice, because the balancing of advance and recession phases during water application is difficult. Border strip irrigation is one of the most complicated of all irrigation methods. The primary design factors are border length and slope, stream size per unit width of border, planned soil water deficiency at the time of irrigation, soil intake rate, and degree of flow retardance by the crop as the water flows down the strip. However, because of the large variations in field conditions that occur during the season, the irrigator can have as great an effect on irrigation efficiency as the system designer. Furrow irrigation uses furrows, which are small, sloping channels formed in the soil. Infiltration occurs laterally and vertically through the wetted perimeter of the furrow. Systems may be designed with a variety of shapes and spacings. Optimal furrow lengths are primarily determined by intake rates and stream size. Even when soils are uniform, the intake rates in furrows may be quite variable due to cultural practices. The intake rate of a newly formed furrow will be greater than a furrow that has been irrigated. Wheel-row furrows can have greatly reduced infiltration rates due to compaction. Because of the many design and management-controllable parameters, furrow irrigation systems can be utilized in many situations, within the limits of soil uniformity and topography. With runoff-return flow systems, furrow irrigation can be a very uniform and efficient method of applying water. However, the uniformity and efficiency are highly dependent on the timing of the irrigation and the skill of the irrigator. Mismanagement can severely degrade system performance. 3.3.1.2 Capabilities and limitations. Some form of surface irrigation is adaptable to most any crop. Basin and border strip irrigation have been successfully used on a wide variety of crops, but furrow irrigation is less well adapted to field crops if cultural practices require equipment travel across the furrows. Basin and border strip irrigations flood the soil surface and will cause some soils to form a crust, which may inhibit the emergence of seedlings. Surface irrigation systems perform better when soils are uniform, since the soil controls the intake of water. For basin irrigation, basin size should be appropriate for soil texture and infiltration rate. Basin lengths should be limited to 100 m on very coarse-textured soils, but may reach 400 m on other soils. Furrow irrigation is possible with all types of soils, but soils with extremely high or low intake rates require excessive labor or capital cost adjustments that are seldom economical. Uniform, mild slopes are best adapted to surface irrigation. Undulating topography and shallow soils do not respond well to grading to a plane. Steep slopes and irregular topography increase the cost of land leveling and reduce basin or border size. Deep cuts may expose areas of nonproductive soils, requiring special fertility management. Erosion control

68

Chapter 3 Planning and System Selection

measures may be required if large stream sizes are used. In areas of high-intensity rainfall and low intake rate soils, surface drainage should be considered with basin irrigation, to reduce damage due to untimely inundation. It is important that irrigation stream size be properly matched to basin or border size and furrow lengths for uniform irrigation. Since intake rates may vary during the season, it is helpful if the water supply rate for border and furrow systems can be varied from one irrigation to the next. Border and furrow systems are not suitable for leaching of salts for soil reclamation, since the water cannot be held on the soil for any length of time. The basin method, however, is ideal for this purpose. Under normal operating conditions, leaching adequate for salinity control can be maintained with basin, border, or furrow irrigation. High application efficiencies are possible with all surface irrigation methods, but it is much easier to obtain these potential efficiencies with the basin method. Design application efficiencies for basin systems should be high, perhaps 80% to 90%, for all but very high intake rate soils. Reasonable efficiencies for border strip irrigation are from 70% to 85%, and from 65% to 75% for furrow irrigation. With either the border of furrow methods, runoff-return flow systems may be needed to achieve high application efficiencies. The system designer and operator can control many of the factors affecting irrigation efficiency, but the potential uniformity of water application with surface irrigation is limited by the variability of soil properties (primarily infiltration rate) throughout the field. Field studies indicate that even for relatively uniform soils, the uniformity of infiltration rates may be only about 80%, as calculated in a manner analogous to distribution uniformity for irrigation application amounts. Surface irrigation uniformity estimates based on infiltration time differences need to be decreased somewhat to account for soil variability. Basin irrigation involves the least labor of the three surface methods, particularly if the system is automated. Border and furrow systems may also be automated to some degree to reduce labor requirements. The complicated art of border irrigation (and to a lesser extent furrow irrigation) requires skilled irrigators to obtain high efficiencies. The labor skill needed for setting border or furrow flows can be decreased with costly equipment (gated pipe or automated valves/gates). The setting of siphons or slide openings to obtain the desired flow rate is a required skill, but one that can be learned. With surface irrigation, little or no energy is required to distribute the water throughout the field, but some energy may be expended in bringing the water to the field, especially when water is pumped from groundwater. In some instances these energy costs can be substantial, particularly with low application efficiencies. Some labor and energy will be necessary for land grading and preparation. 3.3.1.3 Economics. A major cost in surface irrigation is that of land grading or leveling. The cost is directly related to the volume of earth that must be moved, the area to be graded, and the length and size of farm ditches. Typical earth moving volumes are on the order of 800 m3/ha, but have on occasion exceeded 2500 m3/ha. Volumes greater than 1500 m3/ha are generally considered excessive, suggesting a design review may be needed. For the U.S., typical earth moving costs are $0.50/m3. For basin irrigation, final finishing with laser-controlled drag scrapers after primary earth moving will cost about $110/ha. Touch-up leveling (at about $50/ha) may be required every two to three years, although some farmers choose to touch-up each year. In Egypt, land leveling costs for surface irrigation are about $170/ha.

Design and Operation of Farm Irrigation Systems

69

Ditch construction can cost from $3/m for earthen ditches to $50/m for large, concrete-lined ditches. Buried, low-pressure plastic or concrete pipelines for low flows can cost about double the cost of concrete-lined ditches, and may cost 5 to 10 times as much for higher flows. They are generally uneconomical on flat terrain where pumping is not required. They may be desirable on steeper slopes (over 1%) or on undulating terrain. Lined canals for a 20-ha Egyptian farm can cost about $1300/ha. U.S. estimates for furrow irrigation labor requirements range from 1.9 to 2.5 h/ha per irrigation or 10 to 13 h/ha per year (based on about 5 irrigations per season). In Egypt, labor costs for surface irrigation range from $80/ha to $170/ha per year. In the U.S., siphon tubes may cost about $30/ha, while gated pipe may cost $370/ha to $735/ha, depending on pipe size, material and length of furrow run. In Egypt, gated pipe for a 20-ha field costs about $340/ha. An operational storage reservoir (i.e., one for short-term storage of water) may be advisable to permit use of a large stream size accumulated from a smaller steady flow. A medium-sized, compacted-earth reservoir capable of storing 24 h water volume would cost about $250/ha for a small (to 16 ha) farm in the U.S. For larger U.S. farms, the cost can drop to about $125/ha. A lined reservoir may cost two to five times as much. In California’s Westlands Water District (1988), a typical runoff recovery and reuse system (3700 m3 sump capacity, serving four fields of 65 ha each) costs about $170/ha served. 3.3.2 Sprinkler Irrigation 3.3.2.1 Types of sprinkler irrigation. In sprinkler irrigation, water is delivered through a pressurized pipe network to sprinklers, nozzles, or jets, which spray the water into the air, to fall to the soil in an artificial “rain.” The basic components of any sprinkler system are a water source, a pump to pressurize the water, a pipe network to distribute the water throughout the field, sprinklers to spray the water over the land, and valves to control the flow of water. The sprinklers, when properly spaced, give a relatively uniform application of water over the irrigated area. Sprinkler systems are usually designed to apply water at a lower rate than the soil infiltration rate, so that the amount of water infiltrated at any point depends upon the application rate and time of application, but not the soil infiltration rate. Hand move or portable sprinkler systems employ a lateral pipeline with sprinklers installed at regular intervals. The lateral pipe is often made of aluminum, with 6-, 9-, or 12-m sections, and special quick-coupling connections at each pipe joint. The sprinkler is installed on a pipe riser so that it may operate above the crop being grown (in orchards, the riser may be short, so that the sprinklers operate under the tree canopy). The risers are connected to the lateral at the pipe coupling, with the length of pipe section chosen to correspond to the desired sprinkler spacing. The sprinkler lateral is placed in one location and operated until the desired water application has been made. Then the lateral line is disassembled and moved to the next position to be irrigated. This type of sprinkler system has a relatively low initial cost, but a high labor requirement. It can be used on most crops, though with some, such as corn or sugar cane, the laterals become difficult to move as the crop reaches maturity. On bare “sticky” soils, moving the lateral lines is very difficult, and an extra line (a “dry” line) is used to give the soil under the “wet” line some time to dry out before that particular line is moved. The side roll sprinkler system is a variation on the hand move lateral sprinkler line. The lateral line is mounted on wheels, with the pipe forming the axle (specially

70

Chapter 3 Planning and System Selection

strengthened pipe and couplers are used). The wheel diameter is selected so that the axle clears the crop as the lateral is moved. A drive unit, usually powered by an aircooled gasoline-powered engine located near the center of the lateral, is used to move the system from one irrigation position to another by rolling the wheels. Gun type sprinkler systems utilize a high-capacity, high-pressure sprinkler (the “gun”) mounted on a trailer, with water being supplied through a flexible hose or from an open ditch along which the trailer passes. The gun may be operated in a stationary position for the desired time, and then moved to the next location. However, the most common use is as a continuous move system, where the gun sprinkles as it moves, in a traveling gun system. The trailer may be moved through the field by a winch and cable, or it may be pulled along as the hose is wound up on a reel at the edge of the field. The gun used is usually a part-circle sprinkler, operating through 80% to 90% of the circle for best uniformity, and allowing the trailer to move ahead on dry ground. These systems can be used on most crops, though due to the large droplets and high application rates produced, they are best suited to coarse soils having high intake rates and to crops providing good ground cover. The center pivot system consists of a single sprinkler lateral supported by a series of towers. The towers are self-propelled so that the lateral rotates around a pivot point in the center of the irrigated area. The time for the system to revolve through one complete circle can range from a half a day to many days. To keep the lateral aligned as it moves around the circle, the outer sections of the lateral must travel faster, and cover larger areas per rotation, than the inner sections. Thus, the instantaneous water application rate must increase with distance from the pivot to deliver an even application amount. The high application rate at the outer end of the system may cause runoff on some soils. A variety of sprinkler products have been developed specifically for use on these machines to better match water requirements, water application rates, and soil characteristics. Since the center pivot irrigates a circle, it leaves the corners of the field unirrigated (unless additions of special equipment are made to the system). Center pivots are capable of irrigating most field crops, and have on occasion been used on tree and vine crops. Linear move (or lateral move) systems are similar to center pivot systems in construction, except that neither end of the lateral pipeline is fixed. The entire line moves down the field in a direction perpendicular to the lateral. Water delivery to the continuously moving lateral is by flexible hose or open ditch pickup. The system is designed to irrigate rectangular fields free of tall obstructions. Both the center pivot and the linear move systems are capable of very highly controlled and efficient water applications. They require moderately high capital investments, but have low irrigation labor requirements. Low energy precision application (LEPA) systems are similar to center pivot and linear move irrigation systems, but are different enough to deserve separate mention. The lateral line is equipped with drop tubes and very low-pressure orifice emission devices discharging water just above the ground surface into furrows. This distribution system is often combined with micro-basin land preparation for improved runoff control (and to retain rainfall which might fall during the season). High efficiency irrigation is possible, but requires either very high soil intake rates or adequate surface storage in the furrow micro-basins to prevent runoff or non-uniformity along a furrow.

Design and Operation of Farm Irrigation Systems

71

Solid set systems are similar in concept to the hand move lateral sprinkler system, except that enough laterals are placed in the field that it is not necessary to move pipes during the season. The laterals are controlled by valves that direct the water into the laterals irrigating at any particular moment. The pipe laterals for the solid set system are moved into the field at the beginning of the season (after planting and perhaps the first cultivation), and are not removed until the end of the irrigation season (prior to harvest). The solid set system utilizes labor available at the beginning and ends of the irrigation season, but minimizes labor needs during the irrigation season. A permanent system is a solid set system where the main supply lines and the sprinkler laterals are buried and left in place permanently (this is usually done with PVC plastic pipe). 3.3.2.2 Capabilities and limitations. Nearly all crops can be irrigated with some type of sprinkler system, though the characteristics of the crop, especially the height, must be considered in system selection. Sprinklers are sometimes used to germinate seed and establish ground cover for crops like lettuce, alfalfa, and sod. The light frequent applications that are desirable for this purpose are easily achieved with some sprinkler systems. Most soils can be irrigated with the sprinkler method, although soils with an intake rate below 5 mm/h may require special measures. Sprinklers can be used on soils that are too shallow to permit surface grading or too variable for efficient surface irrigation. In general, sprinklers can be used on any topography that can be farmed. Land leveling is not normally required. Leaching salts from the soil for reclamation can be done with sprinklers using much less water than is required by flooding methods, although a longer time is required to accomplish the reclamation. This can be particularly important in areas with a high water table. A disadvantage of sprinkler irrigation is that many crops (citrus, for example) are sensitive to foliar damage when sprinkled with saline waters. In contrast, other crops (apples, peas) react favorably to sprinkler irrigation at harvest, actually producing higher-quality fruit. Attainable irrigation efficiencies for different sprinkler systems are given in Table 3.3. Labor requirements vary depending on the degree of automation and mechanization of the equipment used. Hand move systems require the least degree of skill, but the greatest amount of labor. At the other extreme, center pivot, linear move, and LEPA systems require considerable skill in operation and maintenance, but the overall amount of labor needed is low. Energy consumption relates to operating pressure requirements (at the inlet to the system), which vary considerably among sprinkler systems. At the extremes, the LEPA systems may require only 100 kPa or so, while the traveling gun system may require 700 kPa or more. Other systems may require 200 to 400 kPa, depending on the design of the sprinklers and nozzles chosen, spacing, crop, climate, and other factors. Table 3.3. Attainable application efficiencies with sprinkler irrigation. System Type Efficiency (%) Hand move or portable 65-75 Side roll 65-75 Traveling gun 60-70 Center pivot 75-90 Linear move 75-90 LEPA 80-95 Solid set or permanent 70-80

72

Chapter 3 Planning and System Selection

3.3.2.3 Economics. Table 3.4 summarizes cost factors for sprinkler irrigation systems based on U.S. prices and conditions. Capital costs depend on the type of system and size of the irrigated area. These capital costs assume that water is available at ground level at the side of the field, and include mainline and pumping plant. Table 3.5 summarizes costs for pressurized irrigation systems in Egypt. Energy costs are highly variable from place to place. The energy requirements listed in Table 3.4 may be used to estimate costs by applying the locally appropriate unit energy cost. A pumping plant efficiency of 75% has been assumed. The energy figures cited are in terms of kWh per 1000 m3 (gross) of water applied. Table 3.4. Sprinkler irrigation system costs, U.S. irrigation practice.

System Type Hand move or portable Side roll Traveling gun Center pivot: without corner system with corner system Linear move (ditch fed) Linear move (hose fed) Solid set aluminum Permanent [a]

Labor Maintenance Typical Energy Required Cost Field Use Capital Factor[a] Size (kWh per (hours per Cost (ha) ($/ha) 1000 m3) 1000 m3) (%) 65 500-750 85-215 1.65 2 65 1300-1500 85-215 1.17 2 32 960-1200 350-490 0.68 6 55-80 60 130 130 65 65

800-1250 1000-1450 1375-2500 1625-2750 3250-4000 2500-3750

85-235 100-245 85-235 125-265 85-215 85-215

0.10 0.10 0.19 0.19 0.97 0.10

3 6 8 7 2 1

Annual maintenance costs are expressed as a percentage of the system capital cost.

Table 3.5. Irrigation costs, Egyptian irrigation practice, assuming a 20-ha field size. Source Initial Energy Maintenance of Capital & & Spare Water Cost Labor Consumables Parts Supply ($/ha) ($/ha) ($/ha) ($/ha) Solid set sprinkler Canal 3570 30 160 115 Well 4860 30 170 135 Hand move sprinkler Canal 2285 60 160 85 Well 3570 60 170 130 Center pivot, Canal 2235 35 35 70 low pressure type Well 3930 35 50 117 Minisprinkler Canal 3570 60 85 100 Well 5000 60 100 160 Bubbler Canal 3570 60 85 100 Well 5000 60 100 160 Drip on orchards Canal 2235 60 85 60 Well 3570 60 100 100 Drip on vegetables Canal 3430 70 130 100 or row crops Well 4860 70 145 150

Design and Operation of Farm Irrigation Systems

73

Operating labor costs vary by system type and local costs for labor. Table 3.4 gives typical values for labor hours required per 1000 m3 (gross) of irrigation water applied. Maintenance costs are difficult to predict, but the data in Table 3.4 may be used as an approximate guide. The annual maintenance cost is estimated by multiplying the initial capital cost of the system by the tabulated percentage factor. 3.3.3 Microirrigation 3.3.3.1 Types of microirrigation. Microirrigation is a general term, and includes several more specific methods (ASAE, 1998b). Drip or trickle irrigation (the terms are often considered equivalent) applies the water through small emitters to the soil surface, usually at or near the plant to be irrigated. Subsurface drip irrigation is the application of water below the soil surface. Emitter discharge rates for drip and subsurface drip irrigation are generally less than 8 L/h for single-outlet emitters, 12 L/h per meter for line-source emission devices. Bubbler irrigation is the application of a small stream of water to the soil surface. The applicator discharge rate (up to 250 L/h) exceeds the soil’s infiltration rate, so the water ponds on the soil surface. A small basin is used to control the distribution of water. Microspray irrigation applies water to the soil surface by a small spray or mist. Discharge rates are usually less than 175 L/h.. 3.3.3.2 Capabilities and limitations. Microirrigation irrigation is best suited for tree, vine, and row crops. The main limitation is the cost of the system, which can be quite high for closely spaced crops. Complete cover crops such as grains or pasture cannot be economically irrigated with microirrigation systems. Microirrigation is suitable for most soils, with only the extremes causing any special concern. On very finetextured soils, localized high application rates may cause ponding, with potential runoff, erosion, and aeration problems. On very coarse-textured soils, lateral movement of water will be limited. This can require several emission outlets per plant to wet the desired root volume. With proper design, using pressure-compensating emitters and pressure regulators if required, microirrigation can be adapted to virtually any topography. In some areas, microirrigation is successfully practiced on such steep slopes that cultivation becomes the limiting factor. Microirrigation uses a lower rate of water application over longer periods of time than other irrigation methods. The most economical design would have water flowing into the farm area throughout most of the day, every day, during peak-use periods. If water is not available on a continuous basis, on-farm water storage may be necessary. Microirrigation can be used successfully with waters of some salinity, although some special cautions are needed. Salts will tend to concentrate at the perimeter of the wetted soil volume. If too much time passes between irrigations, the movement of soil water may reverse itself, bringing salts back into the root zone. Salts concentrating on the surface around the edge of the surface wetted area can be a hazard should a light rain occur. Such a rain can move the salts down into the root zone, without applying enough water to leach the salts through and below the root zone. When rain falls after a period of salt accumulation, irrigation should continue as normal until about 50 mm of rain have fallen, to prevent salt damage. In arid regions where annual rainfall is insufficient to leach the salts (less than 300 to 400 mm), artificial leaching may be necessary from time to time, requiring the use of a supplemental sprinkler or surface irrigation system. Though a form of pressurized irrigation, microirrigation is considered a lowpressure, low flow-rate method. These conditions require small openings in the flow

74

Chapter 3 Planning and System Selection

channel of the emission devices, which can be prone to plugging. The sensitivity of emitters to plugging varies with design, but virtually all emitters will require some degree of water treatment in agricultural situations. Cyclonic separators and screen filters are used to remove inorganic particles from the irrigation water, and media filters are used to remove organic contaminants. Chemical treatment of the water may also be required to control biological activity in the water, to adjust pH, or to prevent chemical precipitation which could plug emitters. Proper design and care of the water treatment system is vital to the successful use of microirrigation. Properly designed and maintained microirrigation systems are capable of high efficiencies. Microirrigation systems should be designed for efficiencies on the order of 90% to 95%. With reasonable care and maintenance, field efficiencies in the range of 80% to 90% may be expected. Where plugging is a problem, or emitter performance is highly variable (high variations in emitter flow rate due to either pressure differences in the system or inconsistent manufacturing processes), field efficiencies may be as low as 60%. A large field study in California (Hansen et al., 1995) found that fieldmeasured microirrigation system efficiencies averaged 80%. Due to their low flow characteristics, microirrigation systems usually have few subunits and are designed for long irrigation times. The systems are easily operated manually, but can also be fully automated. Thus, the major labor requirement is for system maintenance and inspection. The amount of maintenance labor required is related to the sensitivity of the emitters to plugging and the quality of the irrigation water. In a vineyard situation, one irrigator can inspect and maintain about 20 ha per day. Another estimate for microirrigation labor is 0.4 h per 1000 m3 (gross) of irrigation water applied. Microirrigation systems may use less energy than other forms of pressurized irrigation. The emission devices usually operate at pressures ranging from 35 to 175 kPa. Additional pressure is required to compensate for pressure losses through the control head (filters and control valves) and the pipe network. System pressures range from about 135 kPa (small systems on flat terrain) to 400 kPa (larger systems on undulating terrain). 3.3.3.3 Economics. Microirrigation system costs can vary greatly, depending on crop (plant, and therefore emitter and hose, spacings) and type of hose employed (semi-rigid or “disposable” thin-walled tubing called “drip tape”). Microirrigation costs will be the lowest for widely-spaced orchard crops, and then increasingly costly for closer spaced vines, and for very closely spaced vegetables and row crops. Microirrigation system costs are summarized in Table 3.6. These cost figures are for high-quality systems and include pumps, filters, controls, mainlines, manifolds, and emitters. In situations where more basic pump, filtration, and control equipment will suffice, costs may be 20% to 25% lower than the figures given. Typical operation and maintenance costs for microirrigation systems vary greatly depending on local circumstances and irrigation efficiencies achieved. One approach is to estimate operation and maintenance costs (per hectare per year) as a percentage of the initial capital cost, as shown in Table 3.7. Other approaches to figuring operating costs are based on estimates of energy and labor requirements. An energy requirement of 70 to 140 kWh per 1000 m3 (gross) of irrigation water applied may be used for microirrigation systems. A corresponding estimate for labor required is 0.4 h per 1000 m3 (gross) of irrigation water applied.

Design and Operation of Farm Irrigation Systems

75

Table 3.6. Microirrigation system costs, U.S. irrigation practice. Microirrigation System Type Capital Cost ($/ha) Drip 2250-2500 Minisprinkler 2700-3200 Microsprayer 2500-3000 Vines Drip 2000-3000 Sprinkler/drip combination[a] 5000-6000 Row crops Drip, retrievable laterals 3000-5000 Drip, disposable laterals[b] 1900-3000 Crop Trees

[a] [b]

Combination permanent set sprinkler and drip system for frost protection and irrigation on vines. Disposable laterals cost $500-600/ha annually.

Table 3.7. Annual operating costs for microirrigation systems. Annual Cost of Operation and Maintenance Expense Category as a Percentage of Initial Capital Cost Labor 1.5 Power[a] 3-7 Maintenance 3 Taxes and insurance 2 [a]

Depends on system efficiency.

REFERENCES Anderson, C. L., A. W. Blair, R. D. Bliesner, A. J. Clemmens, G. L. Dobbs, M. J. English, A. D. Halderman, D. R. Hay, J. D. Hedlund, J. L. Merriam, J. A. Replogle, L. J. Ring, M. L. Soffran, K. H. Solomon, R. E. Walker, I. A. Walter, J. E. Welton and M. Yitayew. 1987. Selection of Irrigation Methods for Agriculture. June 22, 1987, Draft Version. Unpublished manuscript. Reston, Va.: On-Farm Irrigation Committee, Irrigation and Drainage Division, ASCE. ASAE. 1998a. S526.1: Soil and water engineering terminology. St. Joseph, Mich.: ASAE. ASAE. 1998b. EP405.1: Design, installation and performance of trickle irrigation systems. St. Joseph, Mich.: ASAE. Hanson, B., B. Bowers, B. Davidoff, D. Kasapligil, A. Carvajal, and W. Bendixen. 1995. Field performance of microirrigation systems. In Proc. Fifth International Micorirrigation Congress, Microirrigation for a Changing World: Conserving Resources/Preserving the Environment, 769-774. Jorgensen, G. 1996. Survey of current irrigation costs. Unpublished notes. Fresno, Calif.: Center for Irrigation Technology, California State Univ. McCulloch, A. W., J. Keller, R. M. Sherman, and R. C. Mueller. 1967. Ames Irrigation Handbook for Irrigation Engineers. W. R. Ames Company, Publisher. Solomon, K. H. 1988. Irrigation system selection. Irrigation Notes, July 1988. CATI Publication No. 880702. Fresno, Calif.: Center for Irrigation Technology, California State Univ. Solomon, K. H., D. L. Nef, G. Jorgensen, and D. F. Zoldoske. 1992. Higher agricultural electricity rates: A San Joaquin Valley perspective. Technical Report for the California Energy Commission. Fresno, Calif.: Center for Irrigation Technology, California State Univ. Westlands Water District. 1988. Economic guidelines for on-farm irrigation system modifications. Fresno, Calif.: Westlands Water District.

CHAPTER 4

ENVIRONMENTAL CONSIDERATIONS Thomas J. Trout (USDA-ARS, Fort Collins, Colorado) Dean D. Steele (North Dakota State University, Fargo, North Dakota) Keith O. Eggleston (U.S. Bureau of Reclamation, Denver, Colorado) Abstact. Irrigation, through storing, diverting, spreading, consuming, and returning water, and enabling more intensive agriculture, affects the environment in both positive and negative ways. In this chapter, we provide information that will assist in the identification and evaluation of environmental impacts of irrigation, and tools to reduce negative impacts. Three processes that can impact the environment are discussed in detail: (1) water storage, diversion, and consumption; (2) chemical leaching to groundwater, and (3) soil erosion and surface water runoff. Included as possible impacts are threats to agricultural sustainability and diminished human aesthetics, as well as effects on water quality, plant and animal habitat, and human health. Keywords. Deep percolation, Environmental impacts, Erosion, Irrigation, Leaching, Runoff.

4.1 INTRODUCTION

Each year, 3100 km3 of water—about 5 times the annual flow of the Mississippi— is removed from the Earth’s rivers, streams, and underground aquifers to water crops. Irrigation accounts for 70% of the global water diversions (FAO, 1996). In the U.S. about 189 km3 or 65% of offstream freshwater withdrawals, excluding thermoelectric power, is for irrigation. In the semi-arid western states, the portion withdrawn for irrigation exceeds 80% (Hutson et al., 2004). In the process of storing, diverting, transporting, spreading, consuming, and returning water, and enabling more intensive agricultural activities, the land, water, plant and animal, and human resources are changed. Some changes are positive and others are negative. The purpose and primary benefit of irrigated agriculture is the production of food and fiber for human use and feed for livestock consumption. Irrigated agriculture produces 40% of the global harvest on 20% of the cropland (FAO, 1996). Several populous countries in Asia depend on irrigated land for a majority of their food production (FAO, 2002). In the U.S., about 50% of crop value is produced on 18% of the harvested crop area that is irrigated (NASS, 2004), including most of the fresh fruits and vegetable crops. Without irrigation, increases in agricultural yields and outputs that

Design and Operation of Farm Irrigation Systems

77

have fed the world’s growing population would not have been possible (Rosegrant et al., 2002). Without irrigation, more land would need to be brought under cultivation and production would be more weather dependent, and thus less dependable. Irrigation improves the quality of life for the local agricultural community by increasing and stabilizing incomes and creating jobs. Irrigation water storage facilities may also provide recreation for humans, habitat for animals, and relief from flooding; and irrigation conveyance and drainage systems often provide habitat for plants and animals. Irrigation water often provides habitat oases in the desert. Irrigation may also negatively impact the environment. Erosion and salination of cropland reduce its productivity. Excessive groundwater pumping results in depletion of this valuable resource for current and future users. Diversion of river flows may degrade or destroy wildlife habitat and recreation opportunities. Water shortage in arid areas may limit urban and industrial expansion. Surface and subsurface drainage water from irrigation is usually of lower quality than the water supply and may result in pollution of both groundwater and surface waters. Proper irrigation system design and management can prevent, or at least minimize, many of the potential negative impacts. Unavoidable negative impacts of irrigation should be recognized and evaluated. Only through recognition and evaluation of the various environmental effects can it be assured that the positive impacts exceed the negative for the system owner, the community, and society. The range of irrigation systems and environmental situations varies too widely to discuss all possible environmental impacts. Tables 1 and 2 list many of the impacts of irrigation on the environment. The lists are not comprehensive, but many of the processes and potential impacts are common to most systems. One common environmental hazard, soil salination, is discussed in detail in Chapter 7 and will not be discussed here. The authors’ experiences are primarily in semi-arid environments in the western U.S., and the discussion will be oriented toward problems experienced in that setting. Mock and Bolton (1993) give a more comprehensive listing of potential environmental impacts of water projects. In this chapter, we provide information that will assist in the identification and evaluation of environmental impacts of irrigation, and tools to reduce negative impacts. Three processes that can impact the environment are discussed in detail: (1) water storage, diversion, and consumption; (2) chemical leaching to groundwater; and (3) soil erosion and surface water runoff. Included as possible impacts are threats to agricultural sustainability and diminished human aesthetics, as well as effects on water quality, plant and animal habitat, and human health. Irrigated agriculture requires large amounts of water. Most irrigated agriculture is located in areas where normal water supplies during the growing season are low. Therefore, water is collected from large areas, stored in reservoirs until required, and diverted to irrigated lands where much of it is consumed. The surface water supply is often augmented by pumping from groundwater aquifers that have accumulated and reached equilibrium conditions over hundreds of years. The diverted or pumped water that is not consumed return flows into drains, streams, lakes, and marshes or percolates to groundwater aquifers. These processes disrupt the natural hydrology, and thus change the environment associated with that hydrology. Although these processes are often external to the on-farm irrigation process, they are driven by irrigation water needs and are a consequence of irrigated agriculture.

78

Chapter 4 Environmental Considerations

Table 4.1. Possible negative environmental impacts of irrigation. Impact Agricul- Natural tural Pro- Envi- Human Human Process Effect Possible Consequences ductivity ronment Health Aesthetics Degraded fish and X X wildlife habitat Decreased transport and/or dilution of sediX X X ments and contaminants Reduced or altered Increased river water river flows temperatures and deX creased dissolved oxygen Decreased water supplies for environmental, municX X X ipal and industrial needs Surface water Increased habitat for storage, disease vectors/hosts X X diversion, (mosquitoes, snails, etc.) and Physical hazards of consumption Creation of drowning (humans and X X artificial animals) surface water Entrapment and concenbodies tration of substances (storage X X X (nutrients, salts, metals, reservoirs and pesticides) conveyance Flooding due to failure of channels) X X X an irrigation structure Submersion of land and free flowing river secX X X tions Increased pumping costs X Increased Saltwater intrusion in pumping depths X X coastal areas Depletion of domestic and irrigation water supX X Groundwater plies depletion Aquifer consolidation X Groundwater Land subsidence X X pumping Reduced wetlands X X and Degraded fish and consumption X X wildlife habitat Reduced Decreased dilution of groundwater sediment and contamireturn flows to X X X nant concentrations in surface water surface water Increased river/lake water X temperatures

Design and Operation of Farm Irrigation Systems

Table 4.1 continued. Excess water application Water table with rise inadequate drainage Topsoil loss

79

Waterlogging of agricultural land

X

X

Salination of agricultural or native lands

X

X

X

X

Loss of soil productivity River/lake/wetland sedimentation Sediment in Irrigation facility irrigation runsedimentation off water Degraded fish and wildlife habitat Nutrients in River/lake/estuary irrigation runexcessive plant growth Soil off (primarily and eutrophication erosion phosphorus) Crop pests in Spread of undesirable irrigation run- pests (weeds, diseases, off (organic nematodes) residues, weed Degraded fish and seeds, nemawildlife habitat todes, diseases) Pesticides River/lake water in irrigation contamination runoff Decreased soil fertility Leaching of Groundwater nutrients contamination (primarily Surface water nutrient nitrogen) enrichment Increased groundwater Water Leaching of and surface water percolation salts salinity through the Leaching of soil to Contamination of potentially groundwater harmful groundwater and surface elements from water soil profile Pesticide contamination Leaching of of groundwater and pesticides surface water

X

X X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

X X

X

X

X

X

X

X

X

X

80

Chapter 4 Environmental Considerations

Table 4.2. Possible positive impacts of irrigation. Impact Natural Agricultural Envi- Human Human Process Effect Possible Consequences Productivity ronment Health Aesthetics Increased recreation X Increased plant and Creation of X X Water animal habitat reservoirs storage and Improved flood control X X and water Hydropower generation X X X conveyance channels More dependable X X X water supply Increased quantity, quality, and dependability of food X X X and fiber production Crop Irrigation production Increased employment and economic X X development

4.2 WATER STORAGE, DIVERSION, AND CONSUMPTION Irrigated agriculture requires large amounts of water. Most irrigated agriculture is located in areas where normal water supplies during the growing season are low. Therefore, water is collected from large areas, stored in reservoirs until required, and diverted to irrigated lands where much of it is consumed. The surface water supply is often augmented by pumping from groundwater aquifers that have accumulated and reached equilibrium conditions over hundreds of years. The diverted or pumped water that is not consumed return flows into drains, streams, lakes, and marshes or percolates to groundwater aquifers. These processes disrupt the natural hydrology, and thus change the environment associated with that hydrology. Although these processes are often external to the on-farm irrigation process, they are driven by irrigation water needs and are a consequence of irrigated agriculture. 4.2.1 Impacts of Water Storage on Rivers and Streams Irrigation water supply development from surface water resources frequently requires placement of structures in the stream (Figure 4.1). The purpose, size, and use of the structure as well as the natural hydrology and biology of the stream determine the environmental impacts that will occur both downstream and upstream of the structure. On-stream reservoirs create large pools of water that react with the environment differently than flowing streams. The changes can be predicted with adequate water quality, meteorological, and other environmental data. Depending on one’s point of view, changes can be interpreted as beneficial or adverse. One thing is certain: changes will occur. It is important to describe the expected changes and what can be done to prevent or minimize the effects, so the project planners and the interested public can determine the significance of a change and possible mediation of the effects. Under normal operation, a reservoir on a stream stores water during periods of high or excess flows and releases it during periods of low or deficit flows. This process changes the high and low downstream flows that existed in the past. The reduced high

Design and Operation of Farm Irrigation Systems

81

Figure 4.1. Morrow Point Dam on the Gunnison River in Colorado.

flows affect riparian habitat that depends on over-bank flooding. The flows may not be sufficient to keep the channel gravel clean enough for spawning or to control aquatic growth. The reduced flow variation can impact channel size and result in reduced channel capacity downstream of the reservoir. In extreme cases (large reservoirs on small drainages), the downstream channel may essentially disappear. Dams trap much of the sediment that is carried under natural flow conditions. This can change the geomorphic stability of the downstream channel and cause increased erosion and channel instability in areas with geologic materials that are easily eroded. Reduced sediment in the reservoir discharge will degrade the stream channel if there is a significant amount of erodible material in the channel. This can impact natural habitat, recreation, and other stream uses downstream of the dam. Likewise, reservoir water diverted into irrigation canals can scour sediment out of canal cross-sections resulting in increased seepage losses. Surface irrigators with field lengths designed for infiltration rates reduced by sediment-laden water may find that they can no longer get the water to the end of the row or field. As sediment accumulates in the reservoir, nutrients and contaminants that are associated with the sediment may become available to aquatic life in the reservoir. Contaminants can accumulate and concentrate in the aquatic food chain and become harmful or even toxic. Actions may be needed to prevent accumulation of harmful contaminants. Reservoirs also impact water temperature and dissolved oxygen. The changes depend on depth of water, organic loading, and eutrophic condition. In temperate, seasonal climates such as the western U.S., impounded water will stratify during summer months if the depth is greater than 5 m. Rising spring and summer air temperatures cause the reservoir surface water temperatures to increase faster than heat can be transmitted to the reservoir depths or consumed in evaporation, and a temperature profile is created. The stratification depends on several factors such as wind mixing, reservoir detention time, solar radiation, and temperature variation. The temperature stratification prevents the warm, less dense, surface water layer from mixing with the deeper, colder layer. In some instances this can also cause oxygen deficiencies at depth. Usually, oxygen deficiency in western U.S. irrigation reservoirs has not been a problem.

82

Chapter 4 Environmental Considerations

Reservoir outlets that draw water from near the bottom of the dam release colder water in the summer and sometimes warmer water during the cold winter months than would naturally flow in the stream. This temperature change may adversely impact native biota and, in extreme cases, result in population decline or even extinction of sensitive species in downstream reaches. Temperature change may also favor some species, including trout. In some areas below dams where the stream conditions are changed, excellent trout fisheries have developed; for example, the Colorado River below Glen Canyon dam and the San Juan River below Navajo reservoir. Where water temperatures are too cold for optimum fish production, selective withdrawal outlets can be provided to blend water from strata of various temperatures to provide the best downstream temperature for river management goals, such as maintaining the native fishery or providing for maximum trout production. This was done at Flaming Gorge reservoir on the Green River. Dams and associated reservoirs submerge free-flowing streams and convert them to pools of water. For recreationists who prefer rafting or fishing in free-flowing streams, this is a significant negative impact. However, for those who prefer lake-like conditions, the reservoir pool is a benefit. Some reservoirs become good fisheries and provide additional recreational benefits over the prior stream opportunities. Net recreational use in an impounded river reach has usually increased after reservoir construction. Dams and reservoirs create fish migration barriers and adversely impact native fisheries that are blocked from their spawning areas. Much of the decline of anadromous fish (salmon) in the northwest U.S. is the result of dams blocking native spawning areas and reservoir pools slowing smolt migration to the sea. Fish ladders or other facilities to assist fish passage can reduce impacts. Water diversions from stream or reservoirs into canals or other conveyances may remove fish from the stream along with the water. Fish screens or migration barriers should be constructed at points of diversion to minimize this problem. Water diversions into large, unlined irrigation channels sometimes create fisheries in the channels, as is the case for the Henry’s Fork of the Snake River near St. Anthony, Idaho. 4.2.2 Impacts of Surface Water Diversions About 110 km3 of surface water was diverted for irrigation in the U.S. in 2000 (Hutson et al., 2004). Irrigation diversions change the stream environment. These changes usually include reduced flow and increased water temperature; they sometimes increase pollutant concentrations, and they often adversely impact the existing aquatic habitat and stream ecosystem. The level of impact is a function of stream size and portion of flow diverted. Most irrigation diversions occur only during the crop growing season, usually late April to early October in much of the western U.S. The flow is left in the channel or stored in reservoirs during the non-growing period. The larger the percent of flow removed from the stream, the greater the impact. Flow removal reduces aquatic habitat and can make the stream less productive. The ultimate impact occurs when the diversion removes all flow from the stream, either on a permanent basis or for short periods of time, which usually ends most aquatic production in the dry reach. This can severely reduce fishing and other water-based recreation. The FAO estimates that 2350 km3 of global annual runoff (6% of natural runoff) is needed for instream flow to maintain healthy ecosystems (FAO, 1996). This is roughly equivalent to global water currently consumed by irrigated agriculture.

Design and Operation of Farm Irrigation Systems

83

Another potential impact of diversions is increase in stream temperature below the diversion. This occurs during summer months when the remaining flow spreads in a channel that was developed by the full flow. Usually the heat exchange opportunities are increased by exposing more rocks and channel bottom, which absorb radiation and then conduct the heat to the remaining flow. Frequently there is less shading of the remaining flow, due to not filling the channel, which allows the remaining flow to absorb proportionally more radiant energy. If stream temperatures are sufficiently elevated, the stream may not support the original biota. It may change from a cold-water fishery to a warm-water one. Increased temperatures during critical periods can also block migration of anadromous fish in their spawning journey. Streams dilute and have a capacity to assimilate pollutants discharged into the flow from point and nonpoint sources. The ability to flush pollutants out of stream sections is reduced when water is removed by diversions. This increases the adverse effects of pollutants that enter the stream downstream of the diversion. Many of the world’s urban areas are located at the mouths of major river systems. These cities have depended on river flows both for their water supply and to flush pollutants they generate out into the ocean. When irrigation diversions reduce river flows, their water supply may be degraded and their pollutants accumulate, degrading the aquatic ecosystem, increasing health risks, and degrading the recreational opportunities and aesthetics of the river. In some cases, the reduced flows may allow ocean saltwater to intrude further into river estuaries. These impacts on urban populations often result in major conflicts between urban and agricultural interests. 4.2.3 Impacts of Irrigation Pumping on Groundwater Aquifers About 79 km3 of water was pumped from groundwater aquifers in the U.S. in 2000 for irrigation (68% of the total pumped) (Hutson et al., 2004). Irrigation pumping can have several impacts on the aquifer sources including water table decline, groundwater depletion, return flow reductions to surface streams, land consolidation, and saltwater intrusion. When groundwater pumping exceeds recharge, the water table level, or piezometric head, declines. Although water table levels fluctuate with seasonal and annual variations in pumping and recharge, the decline will be continual if the pumping exceeds long-term aquifer recharge. Excessive pumping is common in arid and semi-arid areas unless the number and discharge of pumps are regulated to prevent groundwater depletion. More than 4 million hectares of U.S. land are irrigated from aquifers in which pumping exceeds recharge and water tables are declining (Postel, 1993). Water table decline increases pumping lift and therefore pumping costs. A one meter increase in the pumping depth in the U.S. requires over 1 billion MJ (300 million kW h) of extra energy each year for pumping irrigation water. If the decline is significant, wells may have to be deepened or replaced. With continued excess pumping, portions of the aquifer can be completely dewatered and the groundwater resource lost. In many agricultural areas, the aquifers supply not only irrigation but also domestic and municipal water needs. Pumping amounts should be designed and regulated to sustain the yield of the aquifer at a reasonable water table level, although sustainable aquifer yield is often difficult to determine. Groundwater pumping can reduce return flows from aquifers to local streams, springs, and seeps. Water table decline can dry marsh areas and areas that receive subirrigation from the water table. This can have serious environmental effects, espe-

84

Chapter 4 Environmental Considerations

cially when limited natural habitat, such as wetlands in arid areas, is destroyed. Reduced groundwater contributions to local streams can be critical because groundwater sources often provide base flows during periods of low surface runoff. Groundwater usually has a cool, constant temperature, so reduced aquifer contributions in summer result in higher stream temperatures. Proper groundwater investigations are required to identify expected impacts and problems and make informed decisions on the development potential. Aquifer depletion can cause land subsidence. Subsidence is often a greater problem in confined than in unconfined aquifers. Non-uniform subsidence can seriously damage surface structures. Subsidence of confined aquifers also damages the groundwater resource by decreasing the potential storage capacity if the aquifer re-saturates. Water table decline along coastal areas may allow saltwater intrusion into groundwater aquifers. When water table levels are reduced sufficiently that the normal gradient toward the ocean reverses, saltwater moves inland making the groundwater unfit for most uses. 4.2.4 Return Flow from Irrigated Agriculture Irrigation is not 100% efficient, so return flow is produced. The return flow can be surface runoff or percolation to groundwater. In some instances drains are constructed to intercept shallow groundwater and return it to surface flows. These drainage and surface return flows are routed through constructed drains to natural channels. Canal operational spillage is also usually diverted to natural channels. Return flows can have positive or negative effects on the environment. Drainage water from irrigated lands can create marshes and wetlands along drainage flow paths. These wetlands can create critical habitat for wildlife in otherwise arid environments. However, these created wetlands can also create problems for the local residents. If wetlands are desired with the irrigation development, they should be planned and engineered to get maximum benefits with minimum adverse effects. The wetlands can be designed for wildlife habitat, and to remove nutrients and sediments while minimizing water consumption. When return or spillage flows are put in natural channels, the additional flows can change the hydraulic equilibrium and may result in channel erosion. Sediment in surface return flows can change the stream biota and have serious effects on downstream water users. An example of return flows causing problems is the Greenfields Division of the Sun River Project in Montana. Much of the project drainage enters Muddy Creek, a stream that flows south along the east end of the Greenfields irrigated area, and discharges into the Sun River. Prior to irrigation development, Muddy Creek had minimal flows during the summer and late fall. The additional flows contributed from the irrigation project drainage increased Muddy Creek flows and caused serious channel erosion. Sediment from Muddy Creek has changed the Sun River fishery from fair above the confluence to poor below. Irrigation surface return flows may increase temperatures in receiving streams, which can cause problems for aquatic biota and water users. The combined impacts of reduced cool groundwater return flows (see above) and warm surface return flows can change the stream from a cold- to a warm-water environment. A benefit of irrigation water that percolates to groundwater can be an increase of base flows from groundwater to streams. These return flows increase stream flows during drought and low flow periods. The cool groundwater also reduces stream water

Design and Operation of Farm Irrigation Systems

85

temperatures during summer low flow periods. Both the increased flows and reduced summer temperatures can improve the environment for fish and other biota. The South Platte River in northeastern Colorado is a stream and riparian area that has benefited from flow augmentation during late summer and early fall from irrigation water imports, storage, and return flows. 4.2.5 Identifying and Quantifying Impacts Prior to water diversion, storage, conveyance, or drainage project formation and design, all management goals and adverse impacts need to be identified and discussed with the appropriate state and federal agencies and the public. Both the potential benefits and negative impacts need to be carefully considered and evaluated. The National Environmental Protection Act (NEPA) environmental impact analysis required of all such U.S. developments formalizes that process. Potential impacts of irrigation project development can be identified by evaluating the impacts of projects with similar conditions. Since conditions are seldom identical, computer models are often used to predict and evaluate expected changes. The simplest approach would be to develop a spreadsheet mass balance model. More complex models include CALSIM (CDWR, 2000), MODSIM (Labadie et al., 2000), and CEQualW2 (Cole and Wells, 2002). The latter offers water quality options and twodimensional reservoir representation. Accurate model use requires thorough understanding of the physical and biological systems associated with the water resource to be modeled. Proper use of models also requires sufficient input data and an understanding of model’s limitations. More complex models usually require more detailed data. The models must describe both surface and groundwater resources from a quantity and a quality standpoint. Results then must be related to the biological and environmental resources to determine probable project impacts. When an irrigation project is built, a monitoring plan should be implemented to assess impacts. The monitoring plan should be designed to identify and quantify changes that indicate potential problems or problems in areas where the project investigations were inconclusive or did not provide a high level of confidence. This will allow preventive measures or mitigation to be implemented at the earliest possible date. 4.2.6 Strategies for Reduction or Control of Negative Impacts Irrigation has impacts. During new project development, these impacts must be evaluated and weighed against project benefits in the decision-making process. Actions can be implemented to reduce, partially control, or mitigate the impacts caused by irrigation project development. Impacts that must be prevented or minimized need to be brought into the project design process. If fish passage needs to be provided, then instream structures must have features that will provide sufficient passage. If water quality will be negatively impacted, then steps must be taken to correct, prevent, or minimize the impacts. A thorough investigation of adverse impacts and ways to prevent or minimize them must be included in the planning and design of irrigation projects and facilities. The National Environmental Protection Act (NEPA) process (see above) incorporates reduction and mitigation of negative impacts.

4.3 GROUNDWATER QUALITY 4.3.1 Impacts of Irrigation on Groundwater Quality When irrigation water application exceeds soil water storage capacity, excess infiltrated water percolates below the root zone and eventually reaches the groundwater.

86

Chapter 4 Environmental Considerations

Over-application of irrigation water, non-uniform irrigation water distribution, nonuniform soils, salt leaching requirements, and rainfall can all result in water application in excess of the amount that can be stored in the crop root zone and thus produce deep percolation events. Runoff from areas of a field can collect in depressions and cause deep percolation (Derby and Knighton, 2001; Spalding et al., 2003). To support the high yields that irrigation can provide, agricultural chemical use (nutrients and pesticides) is often high. Some of these chemicals, and other substances that may occur naturally in arid soils, can move with downward percolating water to the groundwater. Among U.S. states, tribes, and territories reporting groundwater data to the U.S. Environmental Protection Agency in 2000, fertilizer applications and irrigation practices were listed among the top ten sources of groundwater contamination by 23 and 6 of the states, respectively (USEPA, 2002). Reviews of the impacts of irrigated agriculture on groundwater in the U.S. have been compiled by various authors, e.g., Bouwer (1987), Helweg (1989), NRC (1989), and Ritter (1989). A review of the problems in the arid western states was compiled for California (Schmidt and Sherman, 1987); Arizona and New Mexico (Sabol et al., 1987); Oklahoma and Texas (Law, 1987); and Idaho, Montana, Nevada, Oregon, Utah, and Washington (Sonnen et al., 1987). In the humid regions of the U.S., the impacts of irrigation on groundwater were reviewed for the Corn Belt and Lake States (Mossbarger and Yost, 1989); South (Shirmohammadi and Knisel, 1989); and East (Ritter et al., 1989b; Krider, 1986). The most common groundwater contamination problems associated with irrigation are nitrate-nitrogen (NO3-N) and pesticides (Hunter, 1986), and these affect people in various direct and indirect ways. Many people, especially those in rural areas, rely on groundwater for their source of drinking water, and pollution of this resource directly affects human health. The U.S. maximum contaminant level (MCL) or drinking water standard for NO3-N is 10 mg L-1. Elevated levels of NO3-N can cause methemoglobinemia in babies (Ritter, 1989b; Dahab and Sirigina, 1994; Skipton and Hay, 1998). Where groundwater reaches surface waters, pollutants in the groundwater may affect the surface water’s usefulness for recreation, fishing, wildlife habitat, and associated tourism industries. Where groundwater is used for aquaculture, pollutants may decrease productivity and even render the water unfit for fish production. Nitrate-nitrogen is the crop nutrient most commonly leached from the root zone. Although NO3-N leaching occurs widely in rainfed agriculture, some of the most severe cases of NO3-N leaching to groundwater, both in the arid western states and in more humid regions of the U.S., have been associated with irrigated agriculture (Hallberg, 1986; Hamilton and Helsel, 1995). One difficulty in controlling NO3-N leaching is that plant N uptake is often inefficient compared with applied fertilizer amounts. For example, less than half of fertilizer N applied to corn is typically removed in the grain, and one-third to one-half of the fertilizer N may leave the profile via deep percolation (Hallberg, 1986). On the other hand, leaching problems may be reduced in many situations with irrigation because of water control and management possible with irrigation (Hallberg, 1989). Irrigation may reduce leaching by promoting increased N uptake by the crop, thereby reducing the amount of N available for leaching during subsequent off-season rainfall events (Juergens-Gschwind, 1989). Other chemicals used as fertilizers that may be leached to groundwater include chloride and potassium from potash, and calcium and magnesium from lime (Hamilton and Helsel, 1995). These elements generally do not cause water quality problems.

Design and Operation of Farm Irrigation Systems

87

Pesticides include herbicides, insecticides, fungicides, soil fumigants, and nematicides. Hallberg (1986) reported that generally less than 5% of applied agricultural pesticides reach groundwater. Flury (1996) reported that, depending on chemical properties, as much as 5% of the applied pesticide may leach beyond the root zone in worstcase rainfall events (i.e., heavy rainfall immediately following pesticide application). When heavy rainfall does not immediately follow application, Flury reported that annual losses are typically less than 0.1% to 1% of the applied pesticide, but may reach as high as 4%. Pesticides that have been found below the root zone or in groundwater underneath irrigated areas include alachlor, atrazine and its metabolites desethylatrazine and desisopropylatrazine, carbofuran, metolachlor, metribuzin (Burgard et al., 1993; Goesselin et al., 1997; Ma and Spalding, 1997b; Ellerbroek et al., 1998; Kraft et al., 1999; Batista et al., 2002; Spalding et al., 2003), benomyl (Merwin et al., 1996), bromacil, 2,4-DP, diazinon, dibromochloropropane (DBCP), 1,2-dibromoethane, dicamba, 1.2dichloropropane, diuron, prometon, prometryn, propazine, simazine (Domagalski and Dubrovsky, 1992; Walker and Ross, 1999; CDPR, 2000), 3,4-dichloroaniline, dimethoate, alpha- and beta-endosulfan, lindane, molinate, prometryn (Batista et al., 2002), cyanazine (Ritter et al., 1989a), and propalchlor (Spalding et al., 1989). The importance of pesticide problems in groundwater may not yet be fully apparent because some pesticides that have leached below crop root zones are still in transport though the unsaturated zone between the root zone and the groundwater. In an irrigated farming area in eastern Washington, Jones and Roberts (1999) compared pesticide occurrences in monitoring wells, designed to sample only the groundwater nearest the surface, and domestic wells. They attributed the variations in groundwater quality to well depths, groundwater age, and greater effects of sorption, dispersion, dilution, and degradation corresponding to longer residence times, and found that over 60% of the pesticides appearing only in domestic wells were compounds no longer used. In arid and semiarid areas where irrigation is common, lack of rainfall over long time periods allow soluble substances such as salts and several trace elements to accumulate and remain in the soil profile. The elements may have been present in the original material, deposited by some past geologic event such as deposition on ancient lake beds, or brought in with the irrigation water. In areas of high rainfall, these elements would have gradually leached out of the profile, drained to rivers, and eventually reached the ocean. Some soils or the geologic formations underlying irrigated areas contain trace elements such as boron, selenium, cadmium, molybdenum, and arsenic that can cause water quality problems. When subjected to water percolation from irrigation, these elements gradually dissolve into the water and may move to shallow groundwater and drain to lakes and marshes, where they accumulate and concentrate through evaporation and can become toxic to plants or animals. Examples include the leaching of selenium from shale formations underlying irrigated areas in Wyoming (Smith, 1994) and boron and selenium from the west side of the San Joaquin valley in California (Presser and Ohlendorf, 1987; San Joaquin Valley Drainage Program, 1990; Frankenberger and Engberg, 1998). Many arid and semi-arid soils and all irrigation water contains dissolved salts. When irrigation water containing salts or water from a shallow water table or aquifer

88

Chapter 4 Environmental Considerations

is transpired by plants or evaporated from the soil surface, salts are left behind in the soil. High concentrations of dissolved salts are harmful to plants and must be removed by leaching to maintain crop production. Salt tolerance of various crops and salinity management techniques have been extensively documented (e.g., Rhoades and Loveday, 1990; Chapter 7 of this monograph). Leaching moves the salts and any other soluble elements or chemicals to the groundwater. High salt concentrations in the groundwater make it unsuitable for domestic use and limit its use for irrigation. Although groundwater in some semi-arid areas was salty before agricultural development, irrigation often compounds the problem by importing water containing salts and by leaching salts previously stored in the soil profile into the groundwater. In many areas, the greatest impact of irrigation is to increase deep percolation and raise water table levels so that return flows of salty groundwater to surface waters are increased. Where groundwater is shallow, drains are sometimes installed to lower the water table and prevent accumulation of salts. This drainwater often has elevated salt concentrations. 4.3.2 Percolation and Leaching Processes Percolation of water through the soil has the potential to leach soluble chemicals from surface and sub-surface soil to the groundwater. Percolation of water through and beyond the root zone of crops occurs by flow through the porous matrix of the soil and through preferential flow pathways. Flow through the porous matrix is described by Darcy’s law, which states that flow is proportional to the hydraulic conductivity of the soil and the hydraulic gradient. Water and chemical leaching potential generally increase directly with soil hydraulic conductivity and inversely with water holding capacity. High hydraulic conductivity permits rapid downward movement and provides little time for chemical uptake or transformation in the unsaturated zone. Soils with low water holding capacity provide less room for error in irrigation scheduling and for storage of unexpected rainfall events, and thus increase the risk of deep percolation loss. Soils with low water holding capacity also often have high hydraulic conductivity. Flow through preferential pathways includes water movement through soil cracks, root channels, earthworm holes, and channels formed by other biological activities. The resulting macropore flow of water through soil resulting from these factors provides a short-circuiting effect by which water and chemicals at the soil surface bypass much of the soil matrix and reach groundwater faster than by Darcian flow through the soil matrix. Saturated or near-saturated surface conditions are necessary for water to move through macropores. Nieber (2001) presented an overview of preferential flow mechanisms—including macropore flow, gravity-driven unstable flow, heterogeneitydriven flow, oscillatory flow, and depression-focused recharge—and the reader is referred to his paper for a review of research on these preferential flow mechanisms and their impact on water quality. Surface irrigation is not the only type of irrigation that can cause preferential flow, because saturated soil surfaces may occur with sprinkler or even drip irrigation when water application rate exceeds infiltration. Ponding may occur in topographic depressions that collect runoff from adjacent areas. This concentrated recharge may significantly influence chemical transport to groundwater. In addition to preferential flow, description of water flow and chemical transport in the root zone of irrigated agricultural systems is complicated by nonuniformity in space and time of soil physical and chemical properties, as well as sources and sinks within the root zone.

Design and Operation of Farm Irrigation Systems

89

Chemical, physical, and biological processes affect chemical fate and transport in porous media (Toride et al., 1993). Processes affecting the state of chemical occurrence in soil include adsorption, dispersion, dilution, and degradation (Jones and Roberts, 1999), volatilization, amounts and timing of chemical applications, soil texture, soil organic carbon or organic matter content, irrigation method, groundwater depth (Hallberg, 1986; Domagalski and Dubrovsky, 1992), chemical solubility (Elliott et al., 2000), composition of ground cover, and photochemical degradation (Wilson et al., 1995). Allen et al. (1985) developed a numerical index (termed DRASTIC) of groundwater vulnerability to pollution based on the following seven factors: “Depth to water, net Recharge, Aquifer media, Soil media, Topography (slope), Impact on the unsaturated zone, and hydraulic Conductivity of the aquifer” (Knox and Moody, 1991). Degradation rates may be quite slow for some chemicals, e.g., DBCP (Walker and Ross, 1999), causing groundwater quality problems to persist for long time periods after chemical use has been discontinued or banned. Biological factors influencing chemical fate and transport in the soil include crop and pest uptake, immobilization on the organic matter in the soil, and microbiological transformations. Models can be used to study the impacts of irrigation on groundwater (Knisel and Leonard, 1989; Watts et al., 199). Various authors (e.g., van Genuchten and Alves, 1982; Leij et al., 1991; Leij et al., 1993) have formulated mathematical models that include one or more of the above transformation and transport processes under different boundary and initial conditions. Many numerical models have been developed to describe water flow and chemical fate and transport in the unsaturated zone, including RZWQM (Ahuja et al., 2000), UNSATCHEM (Suarez and Simunek, 1997), HYDRUS (Kool and van Genuchten, 1992), DRAINMOD-N (Breve et al., 1992), TETrans (Corwin et al., 1991; Corwin and Waggoner, 1991), NLEAP (Shaffer et al., 1991), LEACHM (Wagenet and Hutson, 1987), GLEAMS (Leonard et al., 1987), MOUSE (Steenhuis et al., 1987), PRZM (Carsel et al., 1985), and BAM (Jury et al., 1983). The persistence and fate of chemicals leached to groundwater depends on the nature and use of the aquifer. Chemicals are generally diluted as they reach groundwater. Chemical and biological processes may degrade or transform the chemicals in the groundwater. Natural and pumping-induced groundwater gradients mix and transport chemicals within an aquifer. Subsurface formations with zero or low hydraulic conductivity may effectively isolate contaminated aquifers from deeper aquifers that may be used as drinking water sources. Groundwater carrying pesticides, salts, and metals may return to surface waters such as streams and rivers. This return flow to surface waters may result from natural hydraulic gradients in the groundwater system or it may be due to discharge from manmade drainage systems installed to improve soil aeration and to control salinity. 4.3.3 Evaluating Groundwater Problems Chemical movement from the root zone to groundwater can be estimated from water percolation rate and chemical concentration in the water. Quantification of downward chemical movement is difficult because of chemical transformations, as well as spatial and temporal variability in water movement and chemical concentrations. Methods to measure or estimate water percolation to groundwater may be classified as flux methods or water balance methods (Wagenet, 1986). Flux methods use Darcy’s law to measure or estimate deep percolation (Hillel, 1982). Water balance

90

Chapter 4 Environmental Considerations

methods estimate leaching by applying the principle of conservation of mass to a control volume in the unsaturated zone. The sum of inflows and outflows must equal the change in storage in the control volume. In this case, inflows include precipitation, irrigation, and surface and subsurface inflows (e.g., due to surface runoff reentering the profile, tile drainage, or lateral subsurface flow under sloping topography); outflows include evapotranspiration (ET), surface and subsurface outflows, and leaching; and change in storage may be determined by measuring changes in soil water content. The percolation quantity can be calculated when other variables in the water balance are known or can be assumed negligible. Percolating water can be sampled for chemical constituents by extracting soil water samples with vacuum sampling techniques (Rhodes and Oster, 1986), or by collecting soil samples and extracting the water. Monitoring or other wells are commonly used to sample groundwater for chemical constituents. Sampling may be done in situ using sensors to assess parameters such as electrical conductivity, or by removing a water sample for subsequent laboratory analysis for constituents such as nitrogen, pesticides, and metals. Because it is difficult to estimate mixing and dilution rates in groundwater, samples at several depths may be required. In situations where groundwater returns to surface waters, the quality may be determined using in situ or grab sampling techniques. Biological indicators in surface waters affected by irrigation return flows, such as plant and animal populations, algae growth, or metal accumulations in fish and wildlife, may support the overall assessment of return flow impacts. Hornsby (1990) reviewed public health and pollution problems associated with irrigated agriculture. For highly soluble and mobile nutrients such as nitrate-nitrogen, leaching may in some cases be inferred from signs of visible crop deficiency such as yellowing of leaves or from plant nutrient tests. Nitrogen leaching may also be estimated from soil sampling and accounting for the amounts and timing of fertilizer applications and crop uptake. A difficulty in this approach is that nitrogen mineralization rates may be difficult to determine. The economics of irrigated agriculture can increase the potential for groundwater pollution problems. When water or chemicals are inexpensive compared to the economic returns they can create, overuse is more likely. Pollution potential is greatest when high-value crops are grown and water and chemicals are relatively inexpensive. Farmers and managers sometimes perceive that liberal application of water and chemicals is inexpensive insurance for maximum profit. Leaching of chemicals to groundwater can be viewed as either positive or negative, depending on one’s perspective. Leaching of salts from the root zone is necessary to maintain acceptable crop growth and health and provide for sustained irrigated crop productivity. Leaching of nutrients and pesticides result in loss of valuable resources and reduced availability for crop growth and pest control. Leaching of salts, nutrients, and pesticides to groundwater may adversely affect groundwater resources. 4.3.4 Strategies for Reduction and Mitigation of Leaching Leaching of chemicals to groundwater from irrigated fields can be reduced by decreasing water percolation amounts, by decreasing the amount of chemical available for transport, or both. Tools available to achieve these reductions include the selection, design, and management of irrigation systems, as well as the chemical management and agricultural practices used in crop production. When irrigation and crop management do not perform as desired, groundwater remediation may be necessary.

Design and Operation of Farm Irrigation Systems

91

4.3.4.1 Irrigation system selection, design, and management. Irrigation system design or management changes that increase water distribution uniformity potentially decrease deep percolation losses and chemical leaching. Good irrigation scheduling (Chapter 8) reduces chemical leaching by avoiding excessive water applications and reducing deep percolation. Proper irrigation scheduling also helps ensure healthy plants and maximum nutrient uptake, leaving smaller nutrient amounts—particularly nitrogen—available for leaching. Where groundwater contamination is a concern, intentional deficit irrigation (not refilling the soil profile to field capacity with each irrigation event) can reduce deep percolation, often with only slight yield reductions. Surface, subsurface, micro-, and sprinkler irrigation systems have different characteristics relative to water percolation and chemical leaching. Systems with the ability to apply small, uniform applications have the potential to decrease deep percolation losses. In general, systems that use pressure distribute water more precisely and with better control than those that use gravity for distribution across the soil surface. Thus, well-engineered drip irrigation systems have higher potential irrigation efficiencies than sprinkler systems, and sprinkler systems are potentially better than surface systems. Generally speaking, smaller fields or irrigation zones can be irrigated more uniformly than larger ones. Irrigation systems also differ in the fraction of the soil surface that is wetted and the rate of application. Partially wetting the soil surface may allow purposely placing chemicals to avoid downward movement with the irrigation water. Surface irrigation always results in saturated flow at the soil surface and hence has greater potential for preferential flow. Drip and sprinkler irrigation may not result in saturated flow conditions if application rates are lower than the infiltration rate of the soil. Subsurface irrigation systems use water table control to meet crop water requirements. With very high water tables, the potential for chemicals to reach shallow groundwater is high. However, with subirrigation, since water movement should be primarily upward through the root zone, downward movement of chemicals during irrigation should be minimized. 4.3.4.2 Chemical management and agricultural practices. In addition to irrigation management, good agronomic practices can help reduce the amount of nutrients and chemicals leached to groundwater. These management practices can include reduced chemical usage, use of less mobile chemicals, chemical application timing when leaching is least likely and when usage or transformation is most rapid, and chemical placement where it is least likely to be leached. Nutrient applications can be scheduled to meet crop needs through in-season soil and plant sampling. Nitrogen content of the irrigation water should be measured and considered to avoid over-fertilization (Smith and Cassel, 1991). Nutrient application based on realistic potential yields rather than maximum yields may significantly reduce nutrient application and leaching with only small reductions in yield. Cropping systems management may be used to mitigate N leaching. Examples of cropping systems management include accounting for the N contribution of legumes in subsequent years; using cover crops such as rye, vetch, and wheat, planted after harvest of other crops to immobilize residual N in the soil (Herbert et al., 1995; Wilson et al., 1995); and planning crop sequences and timing to use nitrogen released by the breakdown of the previous crop.

92

Chapter 4 Environmental Considerations

Precisely scheduled and targeted pesticide applications such as are used in integrated pest management programs can result in reduced pesticide applications. Additional practices that can reduce chemical availability for leaching include using alternate-furrow irrigation and chemical placement for row crops (Hefner and Tracy, 1995); placing herbicides away from anhydrous ammonia application zones (Clay et al., 1995) to avoid the high pH regions that reduce herbicide adsorption on the soil and increase potential for herbicide leaching (Hunter, 1986); using encapsulated, slowrelease, or controlled-release herbicides (Hickman and Vail, 1995; Williams et al., 1995); using paraffinic oil as a herbicide carrier to increase spreading on contact and reduce the amount of herbicide needed (Hanks, 1995); and combining banding of herbicides and cultivation for weed control (Baker et al., 1995). Since the initial irrigation often produces the largest deep percolation loss with surface irrigation systems, application of mobile chemicals before the first irrigation should be avoided when possible. Chemicals are usually applied as evenly as possible even though the needs may vary spatially. One promising solution to spatial variability is site-specific technology and management, which divides fields into sub-field management zones and applies chemicals based on localized needs. Banding nutrients, placement away from the irrigated furrow or dripper, and using slow-release or lessmobile forms of nitrogen and nitrification inhibitors can reduce nitrate leaching (Crumpton and Baker, 1993; Dahab and Sirigina, 1994). The unpredictable nature of weather can increase the potential for leaching chemicals to groundwater. For example, unexpected rainfall following irrigation can result in deep percolation. If evaporative demand is significantly smaller than normal in a given growing season due to cool weather, crop growth may be retarded compared to that expected for a warmer season and N fertilizer may be underutilized. High N residuals present in soil at the end of the growing season increase the potential for leaching to groundwater due to off-season precipitation. Since weather cannot be predicted with certainty, flexible chemical and water management are essential to minimize leaching of chemicals to groundwater. 4.3.4.3 Mitigation of impacts. Groundwater remediation may be necessary when irrigation and agronomic management practices are not sufficient to keep chemical leaching losses below acceptable levels, or groundwater concentrations already exceed acceptable levels. Technologies for groundwater remediation may be grouped into four categories: containment, collection, treatment, and disposal (Hunter, 1986). Examples of the technologies are given in Table 4.3. Typical treatment processes employ one or more of the technologies from each of the four categories in an overall treatment scheme. For example, a treatment scheme could involve containment and collection of contaminated groundwater via a well, above-ground biological treatment through irrigation of a crop, and disposal via ET and percolation back to groundwater. In-situ treatment of groundwater can be accomplished by injecting biological organisms, chemicals, or both into the contaminated area. For example, Hunter and Follet (1995) proposed injecting vegetable oil as a carbon source into the porous media around well screens in high-nitrate groundwaters to accelerate biological denitrification in situ. Robertson et al. (2000), used various reactive passive carbon barriers to remediate nitrates in groundwater over a six- to seven-year period. Carbon sources for groundwater remediation have also been studied by Obenhuber and Lowrance (1991), Schipper and Vojvodic-Vukovic (1998, 2000), and Robins et al. (2000).

Design and Operation of Farm Irrigation Systems

93

Table 4.3. Technologies for remediation of groundwater. Technologies potentially applicable to agricultural settings are in italics (from Hunter, 1986).[a] Category Containment slurry wall, vibrating beam, sheet piling, concrete wall, extraction wells, subsurface drains, ditches and trenches Collection

extraction wells only, subsurface drains, ditches and trenches, extraction wells combined with injection wells

Treatment

[a]

In Situ

microbial degradation, limestone treatment bed, activated carbon bed, chemical treatment, neutralization

Biological

activated sludge, trickling filter, rotating biological contactor, land application, aerated lagoon, anaerobic/facultative lagoon

Physical/Chemical

ion exchange, membrane separation, oxidation, reduction, liquid/liquid extraction, carbon adsorption, air stripping, steam stripping, spray evaporation, wet air oxidation, incineration

Disposal

deep well injection, direct discharge, evaporation ponds, percolation ponds, land application

Reprinted by permission of the National Ground Water Association. Copyright 1986. All rights reserved.

A variety of other technologies have been studied for groundwater remediation. The use of zero-valent metals has been studied for the removal of atrazine (Singh et al., 1998), and for sulfate reduction and denitrification (Chew and Zhang, 1998; Gu et al., 2002). Electrokinetic methods for nitrate remediation have been studied by Chew and Zhang (1998) and by Eid et al. (2000). Ma and Spalding (1997a), using artificial recharge of an aquifer with river water, reported tenfold reductions in nitrate and atrazine concentrations beneath and downgradient from the infiltration sites. Weier et al. (1994) found that additions of ethanol to high-nitrate irrigation water may have potential to remediate groundwater nitrates. Biological, physical, and chemical treatment may require collecting groundwater and bringing it above ground for treatment, and then disposing of the treated water. The technologies for biological treatment are well established in the fields of domestic and industrial water and wastewater treatment (Hunter, 1986; Dahab and Sirigina, 1994, and references therein). Land application of contaminated water can involve irrigation to disperse the contaminated water for uptake or removal by plants or soil organisms. Thus, land application can be both a treatment method and a disposal method.

4.4 SURFACE WATER RUNOFF 4.4.1 Impacts of Irrigation-Induced Erosion and Sediment Discharge Water running across the land surface can erode soil. Soil erosion caused by rainfall is a commonly recognized problem, but irrigation can also cause erosion. Sprinkler irrigation is similar to rainfall. Irrigation furrows are similar to rills with a controlled water application. The extent of irrigation-induced erosion is not well documented, although measurements in Idaho, Wyoming, Washington, and Utah show that it is a serious problem in some areas of the western U.S. Koluvec et al. (1993) reviewed

94

Chapter 4 Environmental Considerations

Figure 4.2. Aerial views of furrow irrigated fields in southern Idaho showing white soil areas near the inflow ends where the top soil has been eroded away. (Photo credit: Dave Carter.)

those and other irrigation-induced erosion studies carried out over the last 50 years. A 1985 USDA-Soil Conservation Service survey indicated that about 21% of irrigated cropland in the U.S. is affected by erosion to some degree (Koluvec et al., 1993). Soil erosion reduces soil depth and productivity. In southern Idaho, measurements indicate that 80 years of irrigation-induced erosion of the erodible silt loam soils has reduced crop yield potentials by 25% (Carter, 1993) (Figure 4.2). For many soils, it is very difficult to return eroded soil to its pre-erosion production potential. When irrigation water is applied to the land faster than it is infiltrated, a portion of the water is redistributed within the field and may run off the field. Twenty to fifty percent of the water applied to most furrow-irrigated fields with slopes greater than 0.5% runs off the downstream end (Trout and Mackey, 1988). This runoff water carries eroded sediment as well as nutrients, pests, or chemicals that are attached to the sediment, off the field and into surface drains, rivers, and lakes. Annual sediment runoff between 4 and 40 Mg/ha is commonly measured from furrow-irrigated fields with slopes greater than 1% (Koluvek et al., 1993). Sediment and its adsorbed constituents negatively impact downstream water bodies and users. It fills drains and downstream reservoirs and irrigation canals. Some irrigation companies spend a large portion of their annual maintenance budget mechanically removing sediment deposits from reservoirs, drains, and canals (Koluvek et al., 1993). Often, runoff water becomes the irrigation water supply for downstream farms. Sediment-laden irrigation water increases maintenance costs of ditches, pipelines, and ponds; increases sprinkler head and nozzle wear; and may make water treatment costs for microirrigation prohibitively expensive. Weed seed and other soil-borne pests such as crop viruses and nematodes can be spread from farm to farm with runoff sediment (Sojka and Entry, 2000). Sediment from irrigated fields has degraded many western U.S. rivers (Koluvek et al., 1993). Irrigation return-flow sediments deposit in rivers and may cover fish spawning beds and other natural habitats (Figure 4.3). Sediment accumulation is often severe in low rainfall areas because river flow rates (and thus carrying capacity) are usually low during the irrigation season in irrigated valleys, and spring flushing flows may be reduced by upstream storage facilities. Sediments that are transported through the rivers accumulate in downstream reservoirs, reducing reservoir storage capacity, or at the river mouths where they may interfere with shipping or recreation facilities. Agricultural sediments often carry sufficient adsorbed phosphorus to promote plant growth in rivers and lakes (Westermann et al., 2001). The sediments may also carry trace amounts of persistent agricultural chemicals such as pyrethroids and organophosphates that can accumulate in river and lake beds, and reduce populations of aquatic macroinvertebrates (Anderson et al., 2003).

Design and Operation of Farm Irrigation Systems

95

Figure 4.3. Sediment deposition in the Snake River in southern Idaho from irrigation tailwater return flow. (Photo credit: Dave Carter.)

4.4.2 Runoff and Erosion Processes Soil erosion occurs when soil particles are detached, either by the force of water droplet impact on the surface, or by the shear of water flowing over the surface, and are then transported by flowing water to another location. Detachment depends on both soil properties and condition (erodibility), and on the erosiveness of the water drops or flows. Sediment transport depends on sediment particle sizes and densities, and transport capacity of the flow. Trout and Neibling (1993) discuss irrigationinduced erosion processes in detail. Irrigation-induced erosion is most common with furrow irrigation, in which flows are concentrated in small channels and directed downslope. Where closely spaced crops are grown and flows are spread uniformly across the surface (border irrigation), or where field slopes are very small (basin irrigation), little erosion occurs. When sprinkler irrigation water is applied at rates high enough to produce water runoff, erosion can also occur. Runoff is most common with low-pressure center pivot irrigation where instantaneous water application rates are high at the outer end of the lateral. Furrow irrigation is similar to concentrated flow in rills. The flow erosiveness increases with the shear exerted by water on the soil, which in turn increases flow rate and furrow slope. Doubling the flow rate more than doubles erosion. Doubling the slope more than triples erosion (Kemper et al., 1985). Furrow flow rates decrease in the downstream direction as water is infiltrated. Thus, the erosion rates are highest at the inflow ends of furrows, and a portion of the eroded soil may be deposited in the downstream half of the furrow length (assuming uniform furrow slope). The portion of the eroded sediment that discharges from the field depends on the portion of the flow that runs off the field. In southern Idaho furrows, most erosion is from the upstream one-third of the furrow length, and about two-thirds of the soil is deposited in the downstream half of the furrow length (Trout, 1996), so only about one-third of the eroded soil leaves the field. Sprinkler irrigation erosion is similar to rainfall-induced erosion. Water drops detach soil particles and shallow overland flow transports them to rills where concentrated flow carries the particles down-slope. With center pivot irrigation, only a small portion of the field is being irrigated at any one time, so runoff is often infiltrated within the field and concentrated flow channels are often not long. Sediment discharge from sprinkler irrigated fields is seldom a problem. The discharge of agricultural chemicals with surface irrigation runoff depends on the source and nature of the materials. Soluble materials in the irrigation water, such as salts and nitrate, will discharge with runoff water. However, soluble substances that

96

Chapter 4 Environmental Considerations

are in the soil generally move into the profile with infiltrating water and are not carried with surface flows or eroded surface soils. Thus nitrogen and salt concentrations in runoff are generally not higher than in the applied water. An exception is possible with salt accumulation from surface evaporation at the soil surface and flow through cracks that intermingles with surface flow (Rhoades et al., 1997). Substances that are adsorbed to surface soil particles or precipitated on the soil surface will be transported with eroding soil particles (Bjorneberg et al., 2002). Total phosphorus in runoff water with sediment is usually higher than that in the applied water (Carter et al., 1974; Westermann et al., 2001). Recently applied or persistent pesticides may also be transported off fields with sediment (Spencer et al., 1985). Small soil particles have large specific surface compared to large particles, and thus have more capacity per unit mass to adsorb agricultural chemicals. Thus, a large proportion of the phosphorus and other chemicals that move with sediment is associated with the smallest sediment particles (Brown et al., 1981; Aggasi et al., 1995). Erosion control techniques that preferentially control large sediments are less effective on agricultural chemical discharge. 4.4.3. Evaluating Erosion and Sediment Discharge Problems Erosion causes two problems: loss of soil and productivity on the field and downstream damage from discharged sediment. Furrow erosion can cause both problems. Center pivot irrigation can cause erosion damage on the field but seldom has significant sediment discharge from the field. The problem being evaluated will determine the evaluation method. Quantifying center pivot erosion is difficult because it varies greatly in space and time. Visual observation of water ponding and surface flows indicates potential for erosion. Rills and sediment deposition fans show there is a problem. Measurements of soil depth or crop yield variation between sloped and flat field sections may show erosion has occurred in the past. Erosion in irrigation furrows is evident from furrows that down cut and evolve to gulleys in high slope areas or near the furrow inflow end (Figure 4.4), depositional areas where slope decreases, and very turbid water or visible bed load. As with pivot irrigation, past erosion damage may be evident from decreased soil depth or crop yield in eroded areas. Furrow erosion and sediment discharge can be quantified by measuring the flow rate and sediment concentration in the flow (Sojka et al., 1994; Trout, 1996). Sediment concentration of periodic volumetric samples can be evaluated

Figure 4.4. Eroded irrigation furrows.

Design and Operation of Farm Irrigation Systems

97

Figure 4.5. Collecting a furrow water sample from a furrow flume and pouring it into an Imhoff cone to determine sediment concentration.

gravimetrically or with Imhoff (sedimentation) cones (Figure 4.5). Volumetric samples can likewise be collected for chemical analysis. It is critical that a representative sample be collected, because sediment is not evenly distributed in the flow. This usually requires collecting the complete furrow flow for a short time period, without disturbing the upstream flow, or by sampling a well-mixed turbulent flow in a drain ditch. Sediment discharge is measured at the outflow point of the furrow or field. Erosion or sedimentation is determined as the difference in sediment discharge between two in-field measurement points. Because sediment often deposits on fields where flows are low (downstream sections) or slopes are flat, field average erosion based on field discharge measurements usually underestimates actual erosion damage. The runoff water collection ditch at the downstream end of the field can strongly affect sediment discharge from the field. Where water tends to pond up and flow slowly at the field downstream end, much of the sediment in the water may be deposited before leaving the field. Where field cross-slope exceeds 0.5% and farmers cut a runoff ditch across the lower end of the field to discharge runoff, serious erosion can occur in the ditch and at the end of the furrows, resulting in a downward sloping (convex) field end and greatly increased sediment discharge. Although much is known about erosion and sedimentation processes, they cannot yet be accurately modeled and predicted. Soil erodibility cannot be accurately predicted. Soil erodibility varies with time and management practices, as well as with soil texture and constituents. The Revised Universal Soil Loss Equation, RUSLE (USDAARS, 1997), does not apply to irrigation-induced erosion. The Water Erosion Prediction Project (WEPP) model (Nearing et al., 1989), although adapted for sprinkler and furrow irrigation conditions, has not successfully predicted field measured erosion (Bjorneberg et al., 1999; Kincaid, 2002). All sediment discharging from farm fields does not usually reach downstream rivers or lakes (Brown et al., 1974; Brockway and Robison, 1992). Some of the sediment may deposit in drains or sediment basins. The runoff water may be rediverted by downstream farmers where much of the sediment may deposit on irrigated fields. Field measurements must be supplemented by return flow quality and quantity measurements to assess sediment and chemical inflows to water bodies. Assessing damages to rivers and lakes and their complex aquatic biological systems requires a thorough understanding of those systems as well as an accounting of the various pollutants. 4.4.4 Strategies for Reduction of Erosion and Sediment Discharge 4.4.4.1 Reducing flow erosiveness. Erosion under center pivot irrigation can be eliminated by eliminating overland water flow. This requires reducing the water appli-

98

Chapter 4 Environmental Considerations

cation rate, or increasing the soil infiltration rate. The water application rate can be reduced by increasing application width with spray booms or nozzles with larger wetting patterns. Reducing total application per pass will also reduce runoff, since infiltration rates decline with application amount. Infiltration enhancement practices include preservation of surface residues, reduced tillage, and, for sodic soils, reducing the sodium adsorption ratio of the soil with the addition of gypsum. Nozzles that produce small droplets with low kinetic energy may also sustain higher infiltration by reducing surface sealing. If the application rate cannot be reduced sufficiently, surface pitting (diking, reservoir tillage) can store excess water in place until it can infiltrate. If runoff cannot be avoided, surface residue and close-growing crops can reduce the exposure of the soil to water droplet impact and reduce overland flow shear. Reduced tillage may also reduce the erodibility of soil. The erosiveness of furrow flows can be reduced by reducing flow rates. Reducing flow rate usually results in longer time required to spread water across the field, and thus lower irrigation water distribution uniformity. Thus, there is usually a trade-off between reducing erosion and reducing irrigation uniformity. Management practices that reduce infiltration, such as furrow compaction and surge irrigation, may counteract the impact of reduced flow rates on uniformity. Shortening furrow lengths by subdividing fields reduces required flow rates. However, if the number of runoff points is increased when fields are subdivided, the amount of runoff and sediment discharge may increase. Mid-field gated pipelines can reduce run lengths without increasing field runoff. In many cases, average flow rates are set to insure all portions of all furrows are adequately irrigated. Furrow-to-furrow inflow and infiltration variability creates the need to increase average flow rates to adequately irrigate most of the field (Trout and Mackey, 1988). Reducing flow rate to reduce erosion damage and allowing a small portion of the field to be inadequately irrigated may be beneficial in the long term. Furrow application systems that facilitate uniform furrow flows allow reduced average flow rates and greatly reduce runoff rates. Reduced flow rate after stream advance is complete (cutback) will result in reduced erosion. Irrigation scheduling may result in smaller total application amounts and times, although erosion reduction will likely be proportionately less than the application reduction. Flow velocity, and thus erosiveness, is also reduced by increasing furrow roughness. Furrow roughness can be increased by leaving or placing crop residue in the furrow (Aarstad and Miller, 1981; Carter and Berg, 1991). However, roughness also slows water advance and may reduce irrigation uniformity. Furrow residue is a good option for steep sections of furrows where erosion is greatest and water advance is rapid (Brown and Kemper, 1987). Miller et al. (1987) used straw mulching in combination with surge irrigation to reduce erosion and maintain irrigation uniformity. Carter and Berg (1991) found that no-tillage resulted in lower infiltration during early season irrigations so irrigation uniformity was maintained while surface residue essentially eliminated erosion. Erosion can be reduced by reducing furrow slope, but substantially changing field slopes is usually not practical. In some cases, the furrow direction can be oriented across the slope (contour furrows) to reduce effective furrow slope. This practice can result in severe concentrated flow erosion if water overtops rows. On fields with a convex downstream end, if water flow velocity in the runoff ditch can be reduced, sediment deposition may gradually fill in the depression. Carter and

Design and Operation of Farm Irrigation Systems

99

Berg (1983) devised a buried pipe runoff collection system that collects sediment and eliminates convex field ends. Eliminating runoff collection ditches and planting closegrowing crops on the convex end can slow the flow and reduce erosion and may result in sediment deposition on convex ends. Portable canvas dam checks across eroding runoff collection ditches can reduce ditch erosion. 4.4.4.2 Reducing soil erodibility. Our understanding of soil aggregate stability, cohesiveness, and erodibility is poor. Few techniques are available to reduce erodibility. Erosion tends to be higher after tillage, so reducing the number and depth of tillage operations reduces erosion (Carter and Berg, 1991). Sodium disperses clays and can increase erosion (Lentz et al., 1996). Decreasing the sodium adsorption ratio of soil or irrigation water can reduce erosion. Polyacrylamide (PAM) applied in the irrigation water has been shown to dramatically reduce furrow erosion. PAM has two effects: it acts as a soil stabilizer and reduces erodibility, and it flocculates suspended sediment particles inducing them to deposit. Low concentrations of PAM (0

exp(h*)

Θ = [exp(h*/2)(1–h*/2)]2/(m+2)

01

Gardner (1958) (GR) van Genuchten (1980)[a] (VG)

[c]

(m − 1)Θ 2 (m − Θ )

Linear None Θ [a] Use p = 0.5 and commonly use n = 1/(1 – m). [b] Use v = 2m + 3 (sometimes v = 2m + 1 or 2m + 2). [c] Also may have kr(h) = dθ/dh.

Θ = |h*|-m/(1-m) ; h* >1 = 1 ; 1< h* < 0 1 1 ⎡ ( m − 1)Θ ⎤ h* = 1 − + ln ⎢ Θ m ⎣ m − Θ ⎥⎦

h* = ln Θ

The significance of the parameters α, n, and m used in Table 6.2 in connection with the van Genuchten (1980) (abbreviated VG) and Brooks and Corey (1964) (BC) functions can best be seen in Figure 6.3, which is a log-log plot of these retention relations for specific values of m. The log slope of the asymptote is mn or m/(1 – m), often called the pore-size distribution index, and n determines the degree of curvature in the region near the intercept. The asymptote intercept is 1/α and is often referred to as the air-entry head, he. As n becomes large, the shoulder curvature near he becomes sharper, and the VG expression approaches the more simple BC relation as a limit. It should be noted that the BC relation is a special case of the VG expression. A generalized form of the BC relation, called the transitional Brooks-Corey relation (TBC), has been introduced by Smith (1990). This is functionally equivalent to the VG relation but retains the same parameters as the BC expression. The relation of m to n often used in the VG expression (see Table 6.2) is not retained.

Figure 6.3. The parameters in the TBC or VG retention relation have specific relationship to the shape of the curve, as illustrated here.

126

Chapter 6 Soil Water Relationships

The term soil water capacity, C(h), refers to the slope (dθ/dh) of the soil water characteristic at any point on the curve. This value represents the change in water content per unit change in matric potential and represents an important property for soil water storage and release. The soil water characteristic can be used to estimate the amount of water “released” between any two potentials. Although the soil water potential largely determines the ease with which a plant can obtain water, it is also important to know how much water is in the soil at potentials above a given critical level. This, along with crop water requirements, allows one to estimate the need for irrigation. The soil water potential will decrease as a plant withdraws water. If hc is considered a critical level below which it is not desired to deplete water, then the amount of water available at a potential h1, h1 > hc , will be θ(h1) – θ(hc). Many soils swell and shrink with wetting and drying, so that all of the water does not come from a constant volume of soil. This is especially important at high potentials where soil structure influences the characteristic. 6.2.3.1 Hysteresis. The soil water characteristics for sorption and desorption will often differ because the water content in a soil at a given potential depends upon the wetting and drying history of the soil. This history dependence in the relationship between potential and water content is called hysteresis. A schematic example of desorption and sorption curves for a soil is given in Figure 6.4. When the desorption curve is obtained by drying an initially saturated sample and the sorption curve is obtained by wetting an initially dry sample, the two moisture characteristics are known as the primary hysteresis loops or main branches (main drying curve, MD, and wetting curve, MW, respectively). If the soil is not completely dry before rewetting or not completely wet before drying, the resulting curves will fall between the two primary curves, and they are known as scanning curves (wetting scanning curve, WS, and drying scanning curve, DS). Wherever the starting point is within the main curves, a drying condition will approach the MD curve, and a wetting condition will approach the MW curve. At any given potential the water content will be greater in a drying soil (desorption) than in a wetting soil (sorption). Field soil is rarely either completely wet before drying, or completely dry before wetting, so measured primary wetting or drying reten-

Figure 6.4. Definition diagram of the hysteresis loops which can occur during wetting and drying of a soil.

Design and Operation of Farm Irrigation Systems

127

tion curves can be used only with reservation in interpreting soil water status. The water content or potential can only be estimated from measurement of the water content and the main curves, unless the wetting history is accurately known. However, the amount of error involved is relatively small, compared with other errors involved such as soil variability, climatic changes, and plant variabilities. Excellent discussions of hysteresis are given by Jury et al. (1991), Hillel (1971), and Nielsen et al. (1972). 6.2.3.2 Methods of determining the soil water characteristic. The soil water characteristic is usually determined in the laboratory using tension tables or pressure plates (Figure 6.5). In all of the techniques used, a porous membrane or plate hydraulically connects the soil water with water in the lower chamber. The pores in the membrane are small enough that, under the imposed pressure, water but not air can pass through. In all cases P1 > P2, so that water is forced from the soil into the lower chamber. At equilibrium, the imposed pressure (expressed in suitable terms) can be considered as the potential of the water remaining in the soil. For high potentials the membrane may be blotter paper, fine sand, sintered glass, porous steel, or similar materials. In this case P1 is often atmospheric and P2 is obtained with a hanging water column (Figure 6.5a) or with regulated vacuum. At lower potentials of about –1 bar or less, the pores of these materials are too large to remain water filled and air will pass through the membrane. A fine-pored ceramic is then used as the membrane, P2 is atmospheric, and P1 is obtained with compressed gas, usually air or nitrogen. Ceramic membranes are available with bubbling pressures of 100 bars and more. The air entry value of the plate should be somewhat matched to the soil water potential of interest as the finer-pored ceramics necessary for higher pressures tend to restrict flow. In addition to porous ceramics, porous stainless steel and plastic materials can be appropriate for the wet range. Regulated Pressure

Cover to prevent evaporation Soil

P1

Porous Membrane

Soil

P1

P2 Hanging Water Column

P2 = atmospheric

Figure 6.5. Two methods of determining water retention relations for a soil sample: (left) hanging column and (right) pressure plate.

128

Chapter 6 Soil Water Relationships

In practice, a sample of soil is placed in the pressure chamber in a retaining ring and saturated overnight. The desired pressure is then applied until outflow ceases and the soil water is considered to be in equilibrium with the applied pressure. The amount of water in the soil is then determined, usually by oven drying. The process is repeated, with a second sample being subjected to a different pressure. The resulting soil water contents are plotted against applied pressure or vacuum, expressed as potential units (usually cm or millibars) to form the water characteristic. Details of apparatus and procedure are given in Dane and Topp (2002). Because the pore size distribution has such a large influence on water retention at high potentials, disturbed samples (dried and sieved) often give erroneous results. Use of so-called undisturbed soil cores is preferable, but even with these, some error is inevitable because of the swelling and shrinking that accompanies wetting and drying of many soils. At low potentials (approaching –15 bars) the soil-specific surface dominates water retention and the error introduced by using disturbed soil samples is quite small. Determining the approximate local soil water characteristic in the field may be done at sites where both soil water potential and water content are measured, using apparatus discussed below. 6.2.4 Approximate Soil Water Parameters Field capacity and permanent wilting point once were considered to be soil water constants. They are now recognized as very imprecise but qualitatively useful terms. After infiltration ceases, water within the wetted portion of the profile will redistribute under the influence of potential gradients. Downward movement is relatively rapid at first, but decreases rapidly with time. Field capacity refers to the water content in a field soil after the drainage rate has become small and it estimates the net amount of water stored in the soil profile for plant use. While field capacity was formerly accepted as a physical property characterizing each soil, now it is used as only a very rough measure of the soil water content a few days after it has been wetted. For most soils this is a near-optimum condition for growing plants. However, soil water will continue to move downward for many days after irrigation. Figure 6.6 (Gardner et al., 1970) is a good illustration of the continuous nature of profile water redistribution, contradicting the idea of a definable point associated with this common term. Indeed, Gardner et al. (1970), referring to their experiments, stated that “the soil exhibited nothing resembling a field capacity.” The field capacity concept is perhaps more applicable to coarse than to finetextured soils because in coarse soils most of the pores empty soon after irrigation and the capillary conductivity becomes very small at relatively high potentials. Fine soils, however, retain more water than coarse soils as well as drain longer at significant rates. Any interface of soil layers will inhibit water movement across the interface (more or less, depending on the relative hydraulic properties) and thus restrict redistribution and increase apparent “field capacity” (see Section 6.5.1). Other factors that may influence soil water redistribution rate are organic matter content, depth of wetting, wetting history, and plant uptake pattern. Even the cultural practices are important: for example, the appropriate “field capacity” for dryland farming on a soil is probably much lower than for a frequently irrigated farming system.

Design and Operation of Farm Irrigation Systems

129

Figure 6.6. After irrigation, redistribution in a soil profile extends for many days, as demonstrated here by measurements of Gardner et al. (1970).

The permanent wilting point is the soil water content below which plants growing in the soil remain wilted even when transpiration is nearly eliminated. It represents a condition where the rate of water supply to the plant roots is very low. The water content corresponding to the wilting point applies to the average water content of the bulk soil and not to the soil adjacent to the root surfaces. The soil next to the root surfaces will usually be drier than the bulk soil (Gardner, 1960), because water cannot move toward the root surfaces fast enough to supply plant demands and a water content gradient develops near the root. Like field capacity, permanent wilting is not a soil constant nor a unique soil property. There is no single soil water content at which plants cease to withdraw water. However, for a given soil, the range of water contents for wilting may be quite small, since soil water contents change little with matric potential at very low potentials. Plant wilting is a function of demand as well as soil conditions: plants growing under low atmospheric demand can dry soil to lower water contents than if the demand is high, because more time is allowed for water to move through the soil to the roots. When atmospheric demands are high, plants may temporarily wilt even though soil water contents are considered adequate; an example is sugar beet wilting in midday during the summer. In the wilting range, almost all soil pores are empty of water and the water content is determined largely by the specific surface area and the interparticle contact points. The water content in soil subjected to a pressure potential of –15 bars is closely correlated with the permanent wilting percentage for a wide range of soils (see Romano and Santini, 2002). Because of its simplicity and the availability of reliable equipment, the –15 bar percentage is now commonly used to estimate the permanent wilting point. Formerly, sunflowers were the standard test plant used for determining permanent wilting percentage (Romano and Santini, 2002). The amount of water released by a soil between whatever is considered “field capacity” and permanent wilting is traditionally called the available water. The term implies that the available water can be used by plants, but this is misleading. If the soil water content approaches the wilting range, especially during periods of high atmos-

130

Chapter 6 Soil Water Relationships

pheric demands for water or during flowering and pollination, the yield and/or quality of most crops will decrease significantly. This concept is more important for dryland agriculture than for irrigated conditions. Inasmuch as the difference between field capacity and available water can be no more meaningful than either of the terms, available water is only an estimate of the amount of water a crop can use from a soil. Many farmers irrigate when the available water has been depleted a certain amount, depending upon the crop. For high-waterrequiring crops such as potatoes, irrigation may be scheduled at 15% to 25% depletion (85% to 75% available water remaining in the soil); for many other crops, the depletion may go to 50% to 75% before irrigation. This bank-account type of irrigation ignores any relation between depletion and water potential. For a given soil, the degree of depletion allowed before irrigation may be roughly related to potential through the characteristic curve. As with field capacity, available water is a useful concept, providing that its limitations are recognized, such as variations with soil depth, the influence of climatic factors on evapotranspiration, and the effects of soil profile characteristics. 6.2.5 Methods for Characterizing Soil Water The soil water characteristics at locations in the field may be obtained by water potential measuring devices in combination with soil water measurement. Because of natural soil variability and the variation in soil water content with wetting history (hysteresis), the field-determined water characteristic curve is not precise and is difficult to duplicate. Sorption curves are more difficult and tedious to determine then desorption curves because equilibrium is reached very slowly. 6.2.5.1 Measuring water content. Topp and Ferre (2002) and Rawlins (1976) have discussed the various methods and associated error for measuring soil wetness. The discussion here will be limited to those techniques considered most useful in the field. Gravimetric. The accepted standard for soil water measurement is the gravimetric method, which involves weighing a sample of moist soil, drying it to a constant weight at a temperature of 105° to 110°C, and reweighing to determine the amount of water lost on drying. The results are often expressed as the ratio of mass of water lost to mass of dry soil. The required drying time depends upon the soil texture, soil wetness, loading of the oven, sample size, whether the oven is a forced draft or convection type, and other factors. Usually 24 h is sufficient but the required time is obtained by repeatedly weighing a sample after various periods of drying. Microwave ovens have been used to reduce drying times (Horton et al., 1982). The bulk density of soil may be measured by drying and weighing a known volume of soil, or by using the clod, core, or excavation method (Grossman and Reinsch, 2002). The core method is the most commonly used. A cylindrical metal sampler of a known volume is forced into the soil at the desired depth. The resulting soil core is dried and the bulk density is found by dividing the mass by the volume of the cylinder. Samples may be taken at successive depths from the surface by cleaning out the sample hole to the desired depths with an auger and then forcing the sampler into the soil at the bottom of the hole. Alternatively, samples may be taken in a trench by forcing the sampler into the soil horizontally at the desired depth. Obviously, this latter method involves more labor but the sampling zones can be better observed. Excellent core samplers are available commercially. Neutron scattering. The neutron scattering procedure to estimate soil water content has gained wide acceptance in recent years. A source of high energy or fast neutrons is

Design and Operation of Farm Irrigation Systems

131

lowered to the desired soil depth into a previously installed access tube. Radiumberyllium has been used, but current equipment uses americium. The fast neutrons are emitted into the soil and gradually lose energy by collision with various atomic nuclei. Hydrogen, present almost entirely in soil water, is the most effective element in the soil in slowing down the neutrons. Thus, the degree of the slowing down of neutrons is a measure of the soil water content. The slowed or thermalized neutrons form a cloud around the source and some of these randomly return to the detector, causing ionization of the gas within the detector and creating an electrical pulse. The number of such pulses is measured over a given interval of time with a scalar or the rate of pulsation can be measured with a ratemeter. The count rate is almost linearly related to the water content. When not in use the radiation source is housed in a shield that contains a material high in hydrogen, such as polyethylene. This material serves as a standard by which proper operation of the instrument can be verified. Inasmuch as instrument variations and source decay occur, it is more satisfactory to use the count ratio method rather than just a count. The ratio of sample count/standard count is plotted versus water content. This eliminates any systematic errors due to instrumentation that may vary from day to day. The volume of soil measured depends upon the energy of the initial fast neutrons and upon the wetness of the soil. With the americium-beryllium source the volume of soil measured is a sphere of about 150 mm diameter in a wet soil and up to 500 mm or more in a dry soil (Grossman and Reinsch, 2002). Neutron scattering has some advantages over the gravimetric method because repeated measurements may be made at the same location and depth, thus minimizing the effect of soil variability on successive measurements. It also determines water content on a volume basis, with the measured soil volume influenced by the instrument used, the soil type, and wetness. Disadvantages of neutron scattering are the initial high investment in equipment, the time required per site because of the need to install access tubes, and the training, licensing, and testing required for possession of a radioactive device. Also, measurements near the soil surface are not accurate because neutrons are lost through the surface, and it is difficult to accurately detect any sharp delineation such as a wetting front or soil layering effect. The manufacturer usually supplies a calibration curve, but one should verify whether it applies to a given soil. If changes in water content are desired, rather than absolute values, a single curve is more widely applicable because the bias will be the same in successive readings. Two calibration procedures have been used: field calibration in natural soil profiles, and laboratory calibration in large prepared soil standards. Calibration should be done with the same type of access tubes as used in the field. For details of the method, see Topp (2002). Time domain reflectometry (TDR). This is a relatively new technique used to measure volumetric soil water content and soil bulk salinity based on the high-frequency electrical properties of soil and water. It offers a variety of advantages including rapid, reliable, and repeatable measurements with a minimum of soil disturbance. In theory, TDR requires no calibration and the soil moisture determination is independent of soil texture, structure, salinity, density, or temperature. In reality, calibration may be necessary for different soil types and TDR probes. Additionally, soil salinity can be evaluated with the technique. More than 50 papers discussing the state-of-the-art developments in TDR are published in the proceedings of a 1994 conference (U.S. Bureau of Mines, 1994).

132

Chapter 6 Soil Water Relationships

B

A +3000

Voltage

TDR

Coaxial cable

0

-3000

C

t1

t2 Time

Soil surface Lp

TDR probe

Waveguide

Figure 6.7. (a) Installation diagram for time domain reflectometry (TDR) measurement of surface soil water content. (b) The instrument trace and its interpretation, discussed in text.

TDR refers to both the overall technique and to the electronic device that generates and measures the electrical signal used to measure the soil water content and salinity. A sketch of a typical TDR probe is given in Figure 6.7a. It consists of a coaxial cable, handle, and two or three parallel metal rods. The metal rods are also referred to as waveguides. A voltage pulse is sent down a cable and the returning signal monitored over time. The velocity of the electrical pulse is proportional to the dielectric coefficient (κ) of the soil in contact with the probes. κ is a dimensionless ratio related to the degree of orientation of dipoles in a material when subjected to an oscillating electric field (see Table 6.3 for some typical values of κ). For soil, κ is considered independent of density, texture, structure, temperature, salts (not necessarily true at very low water contents), and others. This is mainly because changes in κ due to changes in water content are very large compare to changes in κ for the other common constituents in soils. As a result, it is possible to create a calibration curve relating κ to soil water content. Topp et al. (1980) conducted a number of experiments to relate the dielectric constant of a wide variety soil types to volumetric water content. They found a close fit for a variety of soils and soil-like materials using a single polynomial equation: θ = –5.3 × 10-2 + 2.9 × 10-2 κ – 5.5 × 10-4 κ2 + 4.3 × 10-6 κ3

(6.5)

This calibration curve is still used, but recent work confirms that it is best to make a calibration curve for each soil and probe to get an optimal fit. Table 6.3. Typical values of dielectric coefficient κ. Material κ (dimensionless) Perfect metal conductor ∞ Water 70 to 80 Dry soils 2 to 5 Perfect vacuum 1

Design and Operation of Farm Irrigation Systems

133

The velocity, vs, of the electromagnetic wave as it travels through the soil is related to the dielectric constant κ by vs = c κ-0.5

(6.6)

where c is the velocity of light. vs is defined as twice the physical length of the probes (2Lp) divided by the time (2ts) for the wave to travel the length of the probe and back, so that ⎛ ct κ =⎜ s ⎜ Lp ⎝

⎞ ⎟ ⎟ ⎠

2

(6.7)

Therefore, the dielectric constant is a function of the speed of light, the length of the probe, and the wave travel time. The art of finding soil water content with TDR is in knowing how to find the two distance end points on the trace. An ideal trace is given as Figure 6.7b. Point t1 represents the time when the signal enters the soil and t2 the time when the wave has traveled to the end of the probe and back to the soil surface. For locating t1 and t2, some users choose to pick the numerical maxima and minima of the trace. For purposes of automation and for noisy signals, special procedures are adapted for identifying points on the trace. Also shown on Figure 6.7b is a voltage height C. The value of C is related to the effective conductivity and hence to soil salinity. For nonsaline conditions, recovery height C is nearly to the same level as at t1; for a saline condition the height C will be less. The analysis of the trace in this manner allows for the simultaneous determination of water content and salinity Other methods. The attenuation of a beam of gamma rays of known intensity passed through a soil column is related to the mass of material through which the beam passes. If the soil bulk density remains constant or varies in a known way, the water content may be inferred, or if the water content is known, the soil bulk density may be inferred. This technique has been used primarily in laboratory studies, although a double-tube attenuation unit has been commercially developed for use in the field (Reginato and Van Bavel, 1964). It is potentially accurate but only for carefully controlled conditions. Transient heat pulse measurements have long been used to evaluate soil thermal properties. Water content is closely related to thermal properties. In particular, the volumetric heat capacity, ρc, is a sum of that due to the water present and the dry soil matrix: ρc = ρw θ cw + ρb cm (6.8) where ρw = density of the water ρb = density of the bulk soil θ = volumetric water content cw = mass specific heat of the water cm = mass specific heat of the dry soil. The value of ρc is found using principles of Fourier heat flow to match temperature rise due to an applied heat pulse. Thus, if the properties of the dry soil are known, then θ follows. A commercially available device is available similar to that described by Bristow et al. (1993).

134

Chapter 6 Soil Water Relationships

6.2.5.2 Measuring soil water potential. It is often desirable to measure soil water potential in addition to, or instead of, soil water content. The estimation of soil water potential from water content data via the characteristic curve may not be sufficiently accurate. Tensiometer. Tensiometers are widely used for measuring the higher ranges of soil water potential in the field and laboratory. The name is derived from the term “tension” that was initially applied to the energy of retention of soil water. Many commercial models are available, or the necessary parts can be purchased and assembled at a significant savings. The theory and use of tensiometers have been discussed by Young and Sisson (2002). Schematic diagrams of tensiometers are shown in Figure 6.8. A tensiometer consists of a porous ceramic cup filled with water and connected through a water-filled tube to a suitable vacuum measuring device. The cup, when saturated with water, must be capable of withstanding air pressures of about 1 bar without leaking air. For normal field applications, the vacuum is measured with a reliable vacuum gauge (Bourdon type). For special conditions where rapid response time is needed, the vacuum measurement is made with pressure transducers and almost no water flow through the cup is required. The pressure transducers can also be used to continuously record the vacuum in the tensiometer, and to read many tensiometers through a switching system. Tensiometers have been used as sensors for automating irrigation systems to maintain a desired soil water range. Tensiometers fitted with a septum and read with a portable pressure transducer attached to a hypodermic needle are also commercially available (see Young and Sisson, 2002). The major criticism of the tensiometer is that it functions reliably only in the wetter soil range at potentials of about –0.8 bar or higher, and its range is limited by the depth

Pressure transducer Removable air-tight cap Vacuum guage Barrel

Porous cup

Figure 6.8. Two types of tensiometer, using a gauge (left) or a pressure transducer (right). With multiple tensiometers, readings may be taken by a single transducer with a needle through septums.

Design and Operation of Farm Irrigation Systems

135

of installation. At lower potentials the water inside the tensiometer vaporizes (boils) and readings are not reliable. In drier sandy soils, when hydraulic conductivities are very low, tensiometers may not function properly. Soil water flow away from the cup as the soil dries may be so slow that hydraulic equilibrium with the bulk soil will not be achieved. Under such conditions the bulk soil water potential may be much lower than the tensiometer indicates. If it is necessary to measure water potentials at great depths, tensiometers equipped with integral pressure transducers can be used and only electrical leads need come to the soil surface (Watson, 1967). Porous ceramic blocks. The water content of porous blocks in equilibrium with the soil water may also be used as an approximate measure of soil water potential or soil water content. Two electrodes are imbedded within a gypsum block and the resistance between them measured with an AC ohmmeter (alternating current is used to prevent polarization). Modern resistance blocks utilize an inert material saturated with gypsum. The effect of soil solution salinity levels is masked because the electrolyte within the block is essentially a saturated solution of calcium sulfate. Gypsum blocks are relatively cheap and easy to use. Several companies supply them, as well as inexpensive resistance meters. Even though the accuracy is not good, they do indicate soil water conditions qualitatively and can monitor changes in θ (Topp and Ferre, 2002). Psychrometric methods. A detailed discussion of the use of psychrometry to measure water potential has been given by Andraski and Scanlon (2002). The technique measures the sum of the matric and osmotic potentials. The method most widely adopted for in situ measurement of soil water potential is the measurement of wet-bulb temperature, with a small thermocouple serving as the wet bulb. Water is condensed on the thermocouple by passing a current through it to cool it below the dew point by the Peltier effect. The current is then removed and the wet-bulb temperature is measured. Ambient temperature is also measured with the same thermocouple to allow corrections to be made for temperature effects on the calibration curve. Units are calibrated against potentials of standard solutions. This technique is most useful for measurement of very low potentials, since the dew point temperature is very near to ambient (i.e., high relative humidity) at high potentials.

6.3 SOIL HYDRAULIC CONDUCTIVITY 6.3.1 Conductivity and Darcy’s Law The basic relationship for describing soil water movement was derived from experiments by Darcy who found in 1856 that the flow rate in porous materials is directly proportional to the hydraulic gradient. This relation was originally set forth by Buckingham (1907), although it is better known as Darcy’s law (Swartzendruber, 1969), and may be written as: q = −K

ΔH Δs

(6.9)

where q is the volume of water moving through the soil in the s-direction per unit area per unit time and ΔH/Δs is the hydraulic gradient in the same direction. The proportionality factor, K, is the hydraulic conductivity, which depends on properties of both the fluid and the porous medium. H is the hydraulic head which is the sum of the pressure head, h, and the elevation head, z (Section 6.2.2). The negative sign in Equation 6.9 indicates flow is in the direction of decreasing H.

136

Chapter 6 Soil Water Relationships

Figure 6.9. Unsaturated hydraulic conductivity as a function of water potential (left) and water content (right). As illustrated, Ks is less than the fully saturated conductivity since soil naturally traps a small quantity of air within its pores.

For a saturated soil K is constant, but for regions of the soil which are only partially saturated the hydraulic conductivity varies significantly with water content, K = K(θ). Since θ is a function of h, we may also write K = K(h). Recall that H = h + z (Section 6.2.2) where z is the vertical distance from the datum. Then for flow in the vertical direction, dh ⎛ dh ⎞ q = − K ( h)⎜ − K (h) + 1⎟ = − K ( h ) (6.10) dz ⎠ ⎝ dz

We noted earlier (Section 6.2.3) that soils are usually not completely saturated in nature because of air entrapment during the wetting process. Thus, even for apparently saturated regions below the water table the volumetric water content may not be equal to total porosity, but to θs, the water content at residual air saturation. The value of K corresponding to θs is Ks (Figure 6.9, right), which may still be considered constant in regions below the water table and is sometimes referred to as the apparent saturated conductivity. Further discussions will assume Ks is effective saturated hydraulic conductivity. 6.3.2 Measuring Saturated Hydraulic Conductivity Various methods for measuring saturated hydraulic conductivity in the field have been described in detail by Reynolds et al. (2002). There are methods suitable for soils with high water tables, similar to well pumping methods for obtaining aquifer transmissivities, and other methods suitable for unsaturated soils. The reader is referred to Reynolds et al. (2002) or to Bouwer and Jackson (1974) for details concerning these methods. Again all of these methods provide measurement of Ks at a point, so, due to field variability, numerous measurements may be required to obtain a field effective Ks value. Further, some methods measure the approximate horizontal conductivity, which may differ significantly from the vertical conductivity. 6.3.3 Unsaturated Hydraulic Conductivity For unsaturated soils the water moves primarily in small pores and through films located around and between solid particles. As the water content decreases, the crosssectional area of the films also decreases and the flow paths become more limited. The

Design and Operation of Farm Irrigation Systems

137

result is a hydraulic conductivity function that decreases very rapidly with water content as shown schematically in Figure 6.9(right). Several functions for this relation were given above in Table 6.2. In most cases hysteresis in the K(θ) relationship is small. However, when K = K(h) is used as in Equation 6.10, shown in Figure 6.9(left), hysteresis may be pronounced due to hysteresis in the h(θ) relationship (Figure 6.4). 6.3.4 Measuring Unsaturated Hydraulic Conductivity The measurement of unsaturated hydraulic conductivity is considerably more difficult than measuring saturated hydraulic conductivity. Since the K value is dependent on water content, both the hydraulic gradient and water content or potential must be determined for a range of water contents to adequately define the hydraulic conductivity function. Most of the reported measurements for unsaturated soils have been conducted in the laboratory where boundary conditions can be carefully controlled and soil water content and flow rates precisely measured. Field measurements have also been reported but are much more difficult because of the number of variables that must be measured and the variability of soil in the field. A major problem with both field and laboratory methods for determining K(θ) is the time and expense required. Measurement of a single K(θ) function by a well trained technician may require several days. Furthermore, several measurements may be needed to adequately characterize K(θ) for a given soil type because of field variability of the soil properties. 6.3.4.1 Laboratory methods. Clothier and Scotter (2002) describe steady state methods for measuring K(θ) based on the defining relationship given in Equation 6.10. Essentially the method consists of setting up boundary conditions to obtain steady, one-directional flow for adjustable pressure heads. In one method, a soil sample is placed in an airtight cavity between two horizontal porous plates through which water flows into and out of the sample. The average pressure head in the sample is controlled by the air pressure in the cavity. The mean hydraulic gradient between two points in the sample is determined by using tensiometers to measure the difference in pressure head. Alternatively, the potential and the potential gradient may be controlled by changing the elevations of the water source and the outlet. In either case, by measuring the steady state flow rate, q, the conductivity may be calculated directly from Equation 6.9. Then the potential is changed and the procedure repeated for another value of pressure head. Although simple in concept, this method has some disadvantages. Since soil conductivities are small in general, particularly at the lower water contents, long times are required to approach steady flow, especially for imbibition. Also the conductivity function obtained using the above method represents a point determination, or at most a determination for a sampled soil section. In order to incorporate some of the heterogeneities of natural soils in the conductivity function, it is better if conductivity determinations can be made on several rather large soil samples. Transient methods utilize a controlled boundary condition with careful measurements during water movement to infer conductivity relationships by optimization with solution of Richards’ equations. A good example of such methods is described by Hudson et al. (1996), in which upward flow into a cylindrical soil sample is caused by a fixed flux at the lower boundary, and water content and tension measurements are made with TDR and tensiometer methods, respectively.

138

Chapter 6 Soil Water Relationships

6.3.4.2 Field methods. A method originally described by Nielsen et al. (1964) (also called the instantaneous profile method) has been frequently used (e.g., Nielsen et al., 1973; Cassel, 1974) to determine K(θ) and h(θ) for soils. Water is applied to saturate the surface of a field plot approximately 2 to 4 m square. After a predetermined amount of water has infiltrated, usually 50 to 150 mm, application is ceased and the plot covered with plastic film to prevent evaporation from the surface. Changes in soil water pressure head during subsequent redistribution of the infiltrated water are measured using tensiometers located at 150-mm depth increments to a total depth of about 1.5 m below the surface. Changes in the soil water content with time are inferred from tensiometer readings and the soil water characteristics. The soil water characteristics are obtained from cores taken at each tensiometer depth. The flux at a given depth is calculated from changes in the soil water contents above that depth. Then the hydraulic gradient is determined directly from tensiometer readings and the conductivity calculated from Equation 6.10. Tension infiltrometer or disc permeameter. Figure 6.10 schematically illustrates a disk permeameter. Like the ring infiltrometer, water is applied to a small circular area, but the water is applied under a small tension and no guard ring is used (Hussen and Warrick, 1994). The mariotte siphon is an integral part of the device, controlling the entry water potential, and the disk has a fine nylon mesh surface through which water passes to the soil. Small irregularities in the undisturbed soil surface are dealt with by a shallow layer of fine sand which forms a contact interface. The tension may be adjusted from 10 or 20 up to 150 mm. An alternative arrangement, using a ring with minimum soil penetration, allows use of the same device for application at zero to

Nylon Screen

Air Inlet

Figure 6.10. Schematic of a tension infiltrometer or disk permeameter. The column at right controls the air pressure at the entrance to the supply tube at center, and thus controls the water pressure at the disk/soil interface.

Design and Operation of Farm Irrigation Systems

139

small positive heads. The radially symmetric flow under the permeameter is analyzed by using the method of Wooding (1968). The method can obtain local estimates of Ks for soils that approach steady flow quickly, and can obtain points on the K(h) curve only for soils with low values of he. Discussion of application and data analysis for this device is given in Smith et al. (2002). Borehole permeameters. Similar to the disk permeameter, this method employs a constant head point source at some depth in the soil, which may be the bottom of a bore. The Guelph permeameter is one such device (Reynolds and Elrick, 2002), and another has been introduced by Shani et al. (1987). These methods are used in conjunction with analytic solutions for three-dimensional flow, and can estimate K(h) for homogeneous soils and a specific K(h) relation. 6.3.4.3 Calculation of K(θ) from the soil water characteristic. To evade the difficulty of directly measuring K(θ), numerous attempts have been made to formulate a computational scheme so that the partial or unsaturated hydraulic conductivity may be estimated through the knowledge of other soil properties that are easier to measure. Such properties should be representative of the geometry of pores and their distribution in space. Since the microscopic structure of a porous medium is too complicated to deal with in exact mathematical terms, simplifying assumptions are necessary. Kosugi et al. (2002) has presented a good review of these methods. The fundamental approach, as explained by Mualem (1974), is that the soil retention curve is an analog for the distribution of pore radii that are effective in the flow of water at any overall water content, and that the flow in those filled pores may be described by analogy to a laminar Hagen-Poisouille formula. In terms of an effective pore radius, re, and an effective area of flowing pores, ae, the general formula can be expressed as follows:

∫ re dae 2

K r (θ) =

filled _ pores re 2 dae all _ pores

∫

(6.11)

where Kr is relative hydraulic conductivity, K/Ks. The relations for K(h) shown in Table 6.2 for Brooks and Corey (1964) and van Genuchten (1980) come from applications of such formulas.

6.4 WATER MOVEMENT IN SOIL 6.4.1 Mass Balance and Flow Equations The Darcy-Buckingham flux equations (6.9 or 6.10) are adequate for steady flow in unsaturated soils, but steady flow is not the common condition. To describe the dynamics of soil water, these equations must be combined with an expression for dynamic mass balance to obtain what is commonly called Richards’ equation (Richards, 1931):

∂θ = −∇ ⋅ q + e ∂t

(6.12)

where q is the flux vector, ∇ ⋅ represents the divergence operation (generalized spatial differentiation) and e is a gain/loss term representing root uptake (–), for example.

140

Chapter 6 Soil Water Relationships

For flow in the vertical z direction only, Equation 6.12 may be written as ∂θ ∂q =− +e ∂t ∂z

(6.13)

This expression is combined with Equation 6.10 to obtain the more common form of Richards’ equation: ∂θ ∂ ⎡ ∂ h⎤ ∂ K (6.14) = K (h) ⎥ + +e ⎢ ∂t ∂z ⎣ ∂ z⎦ ∂ z Water content θ may be made the independent variable by substituting the soil water diffusivity, D(θ) ≡ K(h) dh/dθ. Using this variable the equation takes the form of a nonlinear diffusion equation: ∂θ ∂ ⎡ ∂ θ⎤ ∂ K (6.15) = D(θ ) ⎥ + +e ⎢ ∂t ∂z⎣ ∂ z⎦ ∂ z These equations are nonlinear because of the functional dependence of the coefficients, D(θ) or K(h). Given the soil water capacity, C(h), defined above, the variables D(θ) and K(h) are related by D = K/C. Both parameters vary markedly with water content or pressure head as discussed in Section 6.2.3. The significant nonlinearity of the soil parameters is the prime source of difficulty in solving the equation for agricultural water flow conditions. Note that use of Equation 6.14 assumes that there is no resistance to soil air movement and the air pressure remains constant throughout the profile. It is also usually assumed that the soil matrix is rigid and does not change with time, so that the soil water characteristic and hydraulic conductivity relationships are not time variant. These assumptions do not always hold and may cause significant errors in predicted results, as will be discussed below. 6.4.1.1 Two- and three-dimensional flow. In several types of irrigation, water is delivered to a point, a line, or a confined region representing only a small fraction of the total surface area. Examples include trickle and furrow irrigation, subsurface irrigation, and bubbler systems with a separate outlet for individual trees or shrubs. Once the water enters the soil, the flow is governed by gravity and capillarity, just as for the one-dimensional cases Solutions to water flow equations in two and three dimensions have been found for a variety of point and line source geometries. One of the simplest, but still useful, is the solution for a steady point source that is “buried” in a uniform soil and assuming a hydraulic conductivity to be an exponential of the pressure head (Table 6.2). The solution is (Philip, 1968; Warrick, 2003) ϕb ( R, z ) =

(

K s exp(αh) ⎛ q ⎡ exp⎜⎜ 0.5α ⎢ z − R 2 + z 2 = 2 2 0 . 5 α ⎣ 8π( R + z ) ⎝

where ϕb = matric flux potential Ks = saturated conductivity α = m from Table 6.2 h = pressure head

)

0.5 ⎤ ⎞

⎥ ⎟⎟ ⎦⎠

(6.16)

Design and Operation of Farm Irrigation Systems

141

3

q = source strength [L /T] z = depth below the source R = cylindrical radius from the source. For any value of z and R, Equation 6.16 can be used to find the pressure head h. The pressure head distribution based on Equation 6.16 is shown as Figure 6.11a for Ks = 15 mm h-1 and α = 0.0050 mm-1 after Or (1995). The wettest soil (h—> 0) is near the source, i.e., at R = z =0. The contours away from the source, –0.6, –0.7, –0.8, are somewhat elliptical shaped and reach a deeper depth than radius due to the influence of gravity. In Figure 6.11b results are calculated assuming a point source at the surface rather than buried. This solution is

(

)

0.5 ⎤ ⎞ K s exp(αh) αq exp(αz ) ⎛ ⎡ = 2ϕb ( x, y ) − E1 ⎜⎜ 0.5α ⎢ z − R 2 + z 2 ⎥ ⎟⎟ α 4π ⎣ ⎦⎠ ⎝

(6.17)

where ϕb(x,y) is calculated for the buried source from Equation 6.16 and E1 is the exponential integral function (e.g., Abramowitz and Stegun, 1964, Equation 5.1.1). The wettest region is now on the surface near the source. The contours are similar to ellipses descending into the profile, but with the tops truncated. For the deeper depths, the pressure heads are very similar to those for the buried source. For example, at a depth 0.7 m below the source, the values of h in each case is between –0.7 and –0.8 with the surface point source showing a slightly wetter result (i.e., closer to the –0.7 contour than for the buried source). For more complex geometries and time-dependent cases, it is generally necessary to go to purely numerical solutions. Computer packages to do this have generally become much more available and easier to use with increasing use of microcomputers. For example, HYDRUS-2D (Simunek and van Genuchten, 1999) is widely used. Radial Distance (m) 0.0 0.3

0.2

0.4

Radial Distance (m) 0.0 0.0

0.2

0.4 Head (m)

0.1

-0.2

0.2

Depth (m)

0.0

0.0

-0.1

-0.4

-0.3

-0.6

-0.2 -0.4 -0.6 -0.8

-0.5 -0.7

-1.0

-0.8

A

-1.0

-1.2

B

Figure 6.11. Steady-state distribution of soil water pressure head for (a) buried and (b) surface point sources (after Or, 1995; Warrick, 2003).

142

Chapter 6 Soil Water Relationships

6.4.2 The Infiltration Process Infiltration, usually defined as the surface entry of water into the soil profile, is a process of great practical importance to irrigation design. Failure to adequately consider the infiltration process may result in nonuniform distribution of water in the field as well as excessive water loss due to deep percolation or runoff. For surface irrigation, the most efficient furrow or border length depends in part on the infiltration capacity. Many of the soil-related factors that control infiltration also govern soil water movement and distribution during and after the infiltration process. Hence, an understanding of infiltration and the factors affecting it is important to the design and operation of efficient irrigation systems. 6.4.2.1 Infiltrability. Consider a hypothetical experiment with a water depth maintained on the surface to create infiltration into a deep, homogeneous soil column with a uniform initial water content. The flux or the rate water enters the soil surface is called the infiltration rate, f. If ponded water is maintained on the surface, we will find that f decreases with time as shown schematically in Figure 6.12. This decrease is due to reduction in the hydraulic gradients in the soil, but in nature may also be affected by soil changes such as surface sealing and crusting. If the experiment is continued for a sufficiently long time the infiltration rate will approach a constant rate, fz. For homogeneous soil profiles fz is for practical purposes equal to Ks, the hydraulic conductivity at residual air saturation. For layered or surface-crusted profiles, fz is different than Ks, as discussed below. Since water is always ponded on the surface in our hypothetical experiment, the infiltration rate is limited only by soil-related factors. At any time during infiltration the maximum rate that water will infiltrate, as limited by soil properties, has often been called the infiltration capacity of the soil, fc. Hillel (1971) noted that the term capacity is generally used to denote an amount or volume and can be misleading when applied to a time-rate process. He proposed the term soil infiltrability rather than infiltration capacity, and we employ that terminology here.

Figure 6.12. Time dependency of infiltration flux under two supply rates. Although the amounts are equal, the lower rate succeeds in putting more water in the soil, which is the area defined by r – f.

Design and Operation of Farm Irrigation Systems

143

Now consider the same soil column as described above with irrigation water or rainfall applied at a constant rate, r, to the surface. For this case the infiltration rate will initially be equal to r and is limited by the application rate rather than the soil head gradient. As long as the application rate is less than the infiltrability, the infiltration rate will be controlled by the application rate; f = r. However, if r is sufficiently greater than fz or Ks, the infiltrability may become less than r after a period of time. Then the infiltration rate will again be controlled by the soil profile and water will pond on the surface. The infiltration patterns for two rates r are shown in Figure 6.12. Water supplied in excess of the infiltrability will become available for surface storage and/or runoff. Ponding or runoff cannot occur when r < Ks, unless by saturating the soil above a restrictive layer or a high water table. The infiltration rate is normally expressed in units of depth of water per unit time (or volume per unit area per unit time), e.g., mm/h. Cumulative infiltration, I = I(t), is the total amount of water infiltrated at any time t since the start of application, and may be expressed as: t (6.18) I (t ) = f (t ) dt

∫

0 where f is the infiltration rate which may or may not be equal to infiltrability, as discussed above. The soil water distribution during ponded infiltration into a uniform unsaturated soil profile will be approximately as shown by the dotted line in Figure 6.13. Only the surface will be saturated, and water content will decrease with depth toward a relatively sharp front known as the wetting front. Viewed through the side of a glass container, this wetting front appears to be a sharp line, but is in fact a transition not distinguishable by eye. The soil often will not be completely saturated behind the wetting front, in some places, due to air entrapment and possible local counterflow of the air phase.

Figure 6.13. Typical shapes of water after a uniform infiltration event, as well as the shapes after a period of distribution.

144

Chapter 6 Soil Water Relationships

One class of formulas to describe infiltration are those derived by using Darcy’s law (Equation 6.10), plus an expression of mass balance at the soil surface, with appropriate simplifying assumptions on the hydraulic soil characteristics. These are discussed in detail by Smith et al. (2002). The most well known of this class of infiltration equations is that of Green and Ampt (1911). This equation was first derived by a more simple conceptual approach, and originally dealt with ponded upper boundary conditions alone. It arises from Equation 6.15 by assuming that D(θ) is a step function, constant within the wetting zone. This allows integration of Equation 6.15 to obtain: fc = K s

Δθ (G + d ) + I I 0

where G = effective capillary drive: G = ∫ k r (h)dh

(6.19) (6.20)

−∞

d = depth of water on the surface kr = relative conductivity, K(h)/Ks Δθ = (θs – θi), which is the saturation deficit, or the available water storage in the profile. G is an important integral measure of the capillary suction of a soil. From its definition, one can see that it can be thought of as the kr-weighted mean value of capillary potential, plus the surface water driving head. Generalized values of G for various soil types have been presented by Woolhiser et al. (1990). Note that in Equation 6.19 infiltrability is expressed in terms of infiltrated depth I. While in many cases one would find it more convenient to have fc(t), the expression of fc in terms of I is significant for two reasons. First, it unifies the calculation of infiltrability for both sprinkling (or rainfall) and ponding conditions. Second, it allows calculation of ponding and subsequent infiltrability patterns with a single function. Higher application rates cause ponding for smaller values of I, and vice versa. The ponding depth, Ip, can be found by substituting r for f in Equation 6.19 and solving for I = Ip. For subsequent times, as long as rate r > f, the infiltrability is described by this equation. Over the last decades several investigators have verified that ponding times can be accurately modeled by relating rainfall rates to infiltrated depths (Mein and Larson, 1973; Reeves and Miller, 1975; Smith and Parlange, 1978). Infiltration relations with physical meaning have been presented by Smith and Parlange (1978) and Philip (1957), as well as Green and Ampt (1911) (abbreviated G-A). These are summarized in Table 6.4 in dimensionless terms using simple scaling relations. These scaling functions (Smith et al., 2002) are as follows: f Ks

(6.21a)

I* =

I GΔθ

(6.21b)

t* =

tK s GΔθ

(6.21c)

f* =

Design and Operation of Farm Irrigation Systems

145

The step function D(θ) assumption (G-A) is most appropriate for relatively coarse soils and relatively uniform particle size distributions, and represents one limiting type of soil behavior. At the other end of the spectrum of soil hydraulic behavior are well graded soils which may exhibit roughly exponential behavior of D(θ) in the region near saturation (Parlange et al., 1982). This assumption leads from Equation 6.18 to the infiltration relation of Smith and Parlange (1978) (abbreviated S-P) (Table 6.4). Philip (1957) was the first to deal with solutions to the Richards’ equation, and solved Equation 6.19 for surface ponding conditions using a series expansion. He also produced an approximate infiltration equation, by truncating the series solution. Because of the truncation it is not as accurate an approximation as others except at very small or very large times. The Philip expression, using scaled terms defined above, is also given in Table 6.4. These expressions are compared graphically in Figure 6.14. The region between the S-P and the G-A expression is where the infiltration curves for most soils, ideally, should lie. The relative error of the truncated series of Philip in the intermediate values of I is apparent in Figure 6.14, but the expressions are generally in agreement. Since many irrigation applications are simple in terms of time variation, an expression of fc(to) (to being opportunity time) is often useful. For that purTable 6.4. Scaled forms of several analytically derived infiltration capacity functions. Base Infiltration Relationships Equation fc*(t*) fc*(I*) I*(t*) t* (1 − α ) = t* (1 − α ) = Smithα ⎛ f c * ⎞ f c* = 1 + ⎛ exp(αI * ) − 1 + α Parlange 1 ⎛⎜ f c* − 1 + α ⎞⎟ exp( α I * ) − 1 I − ln⎜ ⎜ ⎟ ln⎜ − ln⎜ * ⎜ ⎟ ⎟ and α ⎝ f c* − 1 ⎠ α ⎝ ⎝ f c* − 1 ⎠ Green2 I* + 1 Ampt[a] 1 f c* = 1 + I * = t* + 2t* f c* = Philip 2t* 2 I* + 1 − 1 [a]

⎞ ⎟⎟ ⎠

0 1 dS/m) and SAR is high, soilapplied amendments are often more effective. Gypsum is the most commonly used and widely available amendment. In practice it is unusual to get more than 0.5 to 2 mol/m3 of calcium dissolved from gypsum into an irrigation water supply. These relatively small amounts of calcium in a low-salinity water, however, may increase infiltration by as much as 100% to 300%. If the water is relatively saline, these small amounts of calcium are far less effective and increase infiltration to a much smaller degree. The rate at which gypsum goes into solution is dependent upon its particle size. Finely ground gypsum (less than 0.25 mm in diameter) dissolves much more rapidly than coarse material. The coarsely ground, lowpurity forms of gypsum are more satisfactory for application to the soil; given time for dissolution, low grades of gypsum have been applied successfully in the irrigation water. Tillage to create a rough, yet thoroughly disturbed, soil surface is frequently an effective practice for improving infiltration. Cultivation prior to an irrigation roughens the soil and opens cracks and air spaces that increase the surface area exposed for infiltration. Deep tillage (chiseling, subsoiling) for many soils, particularly fine-textured soils, improves deep water penetration for only a few irrigations because the soil surTable 7.10. Chemical properties of various amendments for reclaiming sodic soil. Amount Solubility in Equivalent to Cold Water 1 kg of 100% Chemical Physical Amendment Composition Description (kg/m3) Gypsum (kg) . CaSO4 2H2O Gypsum white mineral 2.4 1.0 Sulfur S8 yellow element 0 0.2 Sulfuric acid H2SO4 corrosive liquid Very High 0.6 Lime sulfur 9% Ca + 24% S yellow-brown Very high 0.8 solution white mineral 0.014 0.6 Calcium carbonate CaCO3 white salt 977 0.9 Calcium chloride CaCl2.2H2O Ferrous sulfate FeSO4.7H2O blue-green salt 156 1.6 Pyrite FeS2 yellow-black mineral 0.005 0.5 Ferric sulfate Fe2(SO4).9H2O yellow-brown salt 4400 0.6 corrosive granules 869 1.3 Aluminum sulfate Al2(SO4)3.18H2O

198

Chapter 7 Controlling Salinity

face quickly returns to its original condition. Although improvement is not permanent, deep tillage may temporarily allow sufficient water to enter to make an appreciable improvement in stored soil water and in crop yield. Deep tillage, done only when the soil is relatively dry to prevent compaction, physically tears, shatters, and rips the soil. It is done prior to planting or during dormancy of permanent crops. The preservation or application of crop residue, livestock manure, and other organic matter on the soil surface will improve infiltration and is a widely accepted practice. The fibrous and less easily decomposed crop residues, such as from barley, rice, wheat, corn, and sorghum, have improved water penetration markedly, whereas residues from legumes and vegetable crops have been less effective. Relatively large quantities of residues are needed; for instance, manure has been added at rates of 40 to 400 Mg/ha. An application of organic residues in the range of 10% to 30% by soil volume in the upper soil layer may be needed to be effective. 7.6.4 Reclamation of Salt-Affected Soils Soils may be naturally saline or suffer from salination because of irrigation mismanagement or inadequate drainage. Such soils require reclamation before irrigated agriculture can be profitable. The only proven method of reclaiming salt-affected soils is by leaching. Sodic soils normally require the addition of an amendment or tillage to promote the leaching process. Soils high in boron are particularly difficult to reclaim because of the tenacity by which boron is held in the soil. Leaching can be enhanced by growing very salt-tolerant plants during the reclamation process. Reclaiming saline soils by harvesting plants with the salts they have taken up is not practical. Likewise, flushing water over the soil surface to remove visible crusts of salt is not effective. Adequate drainage and suitable disposal of the leaching water are absolute prerequisites for reclamation. 7.6.4.1 Saline soils. The amount of water required for the removal of salts from a saline soil depends on the initial level of salinity, soil physical characteristics, technique of applying water, and soil water content. Leaching by continuous flooding is the fastest method but it requires larger quantities of water than leaching by sprinkler or by intermittent flooding. Intermittent flooding or sprinkling, though slower, is more efficient and will require less water, particularly on finer-textured soils, than continuous ponding. The relationship between the fraction of the initial salt concentration remaining in the soil profile, c/co, and the amount of water leaching though the profile per unit depth of soil, dL/dS, when water is ponded continuously on the soil surface can be described by (c/co) (dL/dS) = K (7.12) where K is a constant that differs with soil type. Equation 7.12 defines the curves in Figure 7.17 for organic (peat) soils when K = 0.45, for fine-textured (clay loam) soils when K = 0.3, and for coarse-textured (sandy loam) soils when K = 0.1. Equation 7.12 is valid when dL/dS exceeds K. Figure 7.17 summarizes the results of nine field experiments (Hoffman, 1986). The amount of water required for leaching soluble salts, particularly for finetextured soils, can be reduced by intermittent applications of ponded water or by sprinkling. The results of three leaching trials by intermittent ponding where no water table was present are summarized in Figure 7.18. Agreement among experiments is excel-

Design and Operation of Farm Irrigation Systems

Figure 7.17. Depth of leaching water per unit depth of soil required to reclaim a saline soil by continuous ponding (adapted from Hoffman, 1986).

Figure 7.18. Depth of leaching water per unit depth of soil required to reclaim a saline soil by ponding water intermittently (adapted from Hoffman, 1986).

199

200

Chapter 7 Controlling Salinity

lent, considering the variety of soil textures and the depth of water applied each cycle (50 to 150 mm) with corresponding ponding intervals varying from weekly to monthly. The relationship between c/co and dL/dS for intermittent ponding can be approximated by (c/ co) (dL/dS) = 0.1 (7.13) For fine-textured soils, intermittent ponding requires only about one-third as much water to remove about 70% of the soluble salts initially present, compared to continuous ponding. Reclamation by flooding may be relatively inefficient when tile or open drains are present as most of the water flow will take place through the soil near the drains, while midway between the drains there will be far less leaching. The regions between drains should be leached by basins separated from the regions above the drains, or leaching should be done by sprinklers to improve efficiency (Luthin et al., 1969). When a soil is both saline and sodic, the initial removal of salts may create sodic conditions and reduce permeability. The equivalent dilution method, which is based on the principle that divalent cations tend to replace monovalent cations on the exchange complex as a result of the dilutions of the soil solution, was proposed to overcome this problem (Reeve and Doering, 1966). By successive dilution of initially applied highsalt water, sodic soils can be reclaimed without a reduction in infiltration capacity. 7.6.4.2 Sodic soils. The reclamation of sodic soils is very difficult, if not impossible, without the use of chemical amendments to replace exchangeable sodium by calcium. In some cases, the equivalent dilution method can be used to reclaim sodic soils. Properties of typical amendments are given in Table 7.10. In a calcarious soil, the addition of acids or acid-forming materials may dissolve sufficient lime to provide sufficient exchangeable calcium. Gypsum, however, is the most common additive because of its low cost, good solubility, and availability. The gypsum requirement to reclaim a sodic soil depends on the amount of exchangeable sodium (Na) to be replaced by calcium. It can be calculated from (Keren and Miyamoto, 1990) GR = 0.86 dsDb (CEC) (ENai – ENaf) (7.14) where GR = gypsum requirement dS = depth of soil to be reclaimed, m Db = soil bulk density, Mg/m3 CEC = cation exchange capacity of the soil, mol/kg ENai and ENaf = initial and desired final exchangeable sodium fractions. According to this equation about 10 Mg/ha of gypsum will replace 30 mol/kg of Na to a depth of 0.3 m for a soil having a bulk density of 1.32 Mg/m3. Typically, a 1-cm depth of water per hectare will dissolve about 0.25 Mg of gypsum. Thus, a 40-cm depth of water will be required to reclaim the soil in the above example to a depth of 30 cm. It is difficult to reclaim a deep sodic soil profile in a single leaching operation. The usual procedure is to partially reclaim the soil during the first year, then plant a shallow-rooted crop and continue the reclamation process in the following years until the entire profile is reclaimed (Keren and Miyamoto, 1990). 7.6.4.3 Boron leaching. Like other salts, excess boron must be removed. More water is needed to leach boron than other salts because it is tightly adsorbed on soil particles. Similar to leaching soluble salts, the relationship between c/co and dL/dS can be approximated by

Design and Operation of Farm Irrigation Systems

(c/co) (dL/dS) = 0.6

201

(7.15)

In addition, periodic leachings may be required to remove additional boron released from the soil particles over time (Oster et al., 1984).

7.7 SUMMARY AND CONCLUSIONS The design and management of irrigation systems must be adequate to prevent harmful accumulations of salt in the crop root zone. Timely irrigations must be of sufficient quantity and uniformity to meet the crop’s needs and leach salts adequately, without excessive surface runoff or deep percolation. All irrigation water contains salt; as evapotranspiration proceeds, the salt concentration increases in the remaining soil water. Without proper irrigation management, the land can become waterlogged and salinized. Regardless of management, drainage water from irrigated lands carries salt that requires appropriate disposal. The response of crops to salinity, sodicity, and toxicity varies widely among plant species. The relationship between crop yield and soil salinity has been quantified for many crops under typical growing conditions. This relationship, however, depends on a number of soil, crop, and environmental factors. Sodicity typically reduces infiltration, which leads to reduced crop yields. Crops can also be sensitive to specific solutes, such as chloride, sodium, and boron. With proper crop selection and appropriate irrigation management, economic yields are usually possible under low to moderate saline conditions. Salt-affected soils can be reclaimed through leaching. Sodic soils normally require a chemical amendment that supplies calcium or magnesium to replace sodium. Copious amounts of water are frequently required to reclaim a salt-affected soil.

REFERENCES Armillas, P. 1961. Land use in pre-columbian America. In A History of Land Use in Arid Regions, 225-276. L. D. Stamp, ed. Paris, France: UNESCO Arid Zone Research, 17. Aubertin, G. M., W. R. Rickman, and J. Letey. 1968. Differential salt-oxygen levels influence plant growth. Agron. J. 60: 345-353. Ayars, J. E. and R. A. Schoneman. 1986. Use of saline water from a shallow water table by cotton. Trans. ASAE 29: 1674-1678. Ayars, J. E., R. B. Hutmacher, R. A. Schoneman, S. S. Vail, and D. Felleke. 1986. Drip irrigation of cotton with saline drainage water. Trans. ASAE 29(6): 1668-1673. Ayers, A. D., C. H. Wadleigh, and O. C. Magitad. 1943. The interrelationship of salt concentration and soil moisture content with the growth of beans. American Soc. Agron. J. 35: 796-810. Ayoub, A. T. 1977. Some primary features of salt tolerance in senna (Cassia acutifolia). J. Exp. Bot. 28: 484-492. Bernstein, L., and M. Fireman. 1957. Laboratory studies on salt distribution of furrow irrigated soil with special reference to pre-emergence period. Soil Sci. 83: 249-263. Bernstein, L., and L. E. Francois. 1973. Leaching requirement studies: Sensitivity of alfalfa to salinity of irrigation and drainage waters. Soil Sci. Soc. America Proc. 37: 931-943. Bernstein, L., and L. E. Francois. 1975. Effect of frequency of sprinkling with saline waters compared with daily drip irrigation. Agron. J. 67: 185-190.

202

Chapter 7 Controlling Salinity

Bernstein, L., M. Fireman, and R. C. Reeve. 1955. Control of salinity in the Imperial Valley, California. USDA, ARS 41(4). Bingham, F. T., and M. J. Garber. 1970. Zonal salinization of the root system with NaCl and boron in relation to growth and water uptake of corn plants. Soil Sci. Soc. America Proc. 34: 122-126. Bingham, F. T., J. E. Strong, J. D. Rhoades, and R. Keren. 1985. An application of the Maas-Hoffman salinity response model for boron toxicity. Soil Sci. Soc. America J. 69: 672-674. Bower, C. A., G. Ogata, and J. M. Tucker. 1969. Root zone salt profiles and alfalfa growth as influenced by irrigation water salinity and leaching fraction. Agron. J. 61: 783-785. Boyden, S. 1987. Western Civilization in Biological Perspective: Patterns in Biohistory. Oxford, U.K.: Oxford Univ. Press. Bresler, E. 1975. Two-dimensional transport of solutes during non-steady infiltration from a trickle source. Soil Sci. Soc. America Proc. 39: 604-613. Bressler, M. B. 1979. The use of saline water for irrigation in the U.S.S.R. Joint Commission on Scientific and Technical Cooperation. Water Resources. Cervinka, V., C. Finch, J. Beyer, F. Menezens, and R. Ramirez. 1987. The agro-forestry demonstration program in the San Joaquin Valley. Progress Report. Sacramento, Calif.: Calif. Dept. of Food and Agriculture, Agricultural Resources Branch. Childs, S. W., and R. J. Hanks. 1975. Model for soil salinity effects on crop growth. Soil Sci. Soc. America Proc. 39: 617-622. Dalton, F. N. 1992. Development of time-domain reflectometry for measuring soil water content and bulk soil electrical conductivity. In Advances in Measurement of Soil Physical Properties Bringing Theory into Practice, 143-167. C. Topp, and R. Green, eds. ASA Special Publication No. 30. Madison, Wis.: Soil Sci. Soc. America. de Wit, C. T. 1958. Transpiration and Crop Yields. Verslag Landbouck Onderzoek No. 64.6.Wageningen, The Netherlands. Dhir, R. D. 1977. Investigations into use of highly saline waters in an arid environment. I: Salinity and alkali hazard conditions in soil under a cyclic management system. In Proc. Intl. Salinity Conf. on Managing Saline Water for Irrigation, 608-609. Lubbock, Tex.: Texas Tech. Univ. Drew, M. C., J. Guenther, and A. Lauchli. 1988. The combined effect of salinity and root anoxia on growth and net Na+ and K+- accumulation in Zea mays growth in solution culture. Ann. Bot. 62: 41-53. Eaton, F. M. 1944. Deficiency, toxicity, and accumulation of boron in plants. J. Agric. Res. 69: 237-277. Ehlig, C. F. 1960. Effect of salinity on four varieties of table grapes grown in sand culture. American Soc. Hort. Sci. 76: 323-331. Epstein, E., J. D. Norlyn, D. W. Rush, R. W. Kingsbury, R. W. Kelley, G. A. Cunningham, and A. F. Wrona. 1980. Saline culture of crops: A genetic approach. Science 210: 399-404. Fang, S., Y. Tian, and D. Xin. 1978. Comprehensive control of drought, waterlogging salinization, and saline groundwater. In Selected Works of Symposium on the Reclamation of Salt-Affected Soil in China. Shandong Publ. House of Scientific Technology. Francois, L. E. 1981. Alfalfa management under saline conditions with zero leaching.

Design and Operation of Farm Irrigation Systems

203

Agron. J. 73: 1042-1046. Francois, L. E. 1982. Narrow row cotton (Gossypium hirsutum L.) under saline conditions. Irrig. Sci. 3: 149-156. Francois, L. E. 1984. Effect of excess boron on tomato yield, fruit size, and vegetative growth. J. American Soc. Hort. Sci. 109: 322-324. Francois, L. E. 1986. Effect of excess boron on broccoli, cauliflower, and radish. J. American Soc. Hort. Sci. 111: 494-498. Francois, L. E., and E. V. Maas. 1994. Crop response and management on salteffected soils. In Handbook of Plant and Crop Stress, 149-181. M. Passaraklie, ed. New York, N.Y.: Marcel Dekker. Francois, L. E., C. M. Grieve, E. V. Maas, and S. M. Lesch. 1994. Time and salt stress affect growth and yield components of irrigated wheat. Agro. J. 86: 100-107. Gardner, W. R. 1957. Some steady-state solutions of the unsaturated moisture flow equation with application to evaporation from a water table. Soil Sci. 85(4): 228-232. Gardner, W. R., and M. Fireman. 1958. Laboratory studies of evaporation from soil columns in the presence of a water table. Soil Sci. 85(5): 244-249. Gelburd, D. E. 1985. Managing salinity lessons from the past. J. Soil Water Cons. 40: 329-331. Goldberg, D., and M. Shimueli. 1971. Sprinkler and trickle irrigation of green pepper in an arid zone. Hort. Sci. 6: 559-564. Grattan, S. R., and J. D. Rhoades. 1990. Irrigation with saline ground water and drainage water. In Agricultural Salinity Assessment and Management, 432-449. K. K. Tanji, ed. New York, N.Y.: American Soc. Civil Engineers. Greenway, H. 1965. Plant response to saline substrates. VII Aust. Biol. Sci. 18: 763-779. Grimes, D. W., R. L. Sharma, and D. W. Henderson. 1984. Developing the Resource Potential of a Shallow Water Table. Contribution No. 188. Davis, Calif.: Calif. Water Resource Center, Univ. of California. Hanson, B. R., and S. R. Grattan. 1990. Field sampling of soil, water, and plants. In Agricultural Salinity Assessment and Management, 186-200. K. K. Tanji, ed. ASCE Manuals and Reports on Engineering Practices. No. 71. New York, N.Y.: American Soc. Civil Engineers. Hanson, B. R., and S. W. Kite. 1984. Irrigation scheduling under saline high water tables. Trans. ASAE 27: 1430-1434. Hoffman, G. J. 1985. Drainage required to manage salinity. J. Irrig. Drain. Eng. 111: 199-206. Hoffman, G. J. 1986. Guidelines for reclamation of salt-affected soils. Applied Agric. Res. 1(2): 65-72. Hoffman, G. J., and J. A. Jobes. 1978. Growth and water relations of cereal crops as influenced by salinity and relative humidity. Agron. J. 70: 765-769. Hoffman, G. J., and M. T. van Genuchten. 1983. Soil properties and efficient water use: Water management for salinity control. In Limitations to Efficient Water Use in Crop Production, 73-85. H. M. Taylor, W. Jordan, and T. Sinclair, eds. Madison, Wis.: American Soc. Agronomy. Hoffman, G. J., J. A. Jobes, and W. J. Alves. 1983a. Response of tall fescue to irrigation water salinity, leaching, fraction and irrigation frequency. Agric. Water Mgmt. 7: 439-456. Hoffman, G. J., E. V. Maas, and S. L. Rawlins. 1973. Salinity-ozone interactive effect

204

Chapter 7 Controlling Salinity

on yield and water relation of pinto beans. J. Environ. Quality 2: 148-152. Hoffman, G. J., E. V. Maas, T. L. Prichard, and J. L. Meyer. 1983b. Salt tolerance of corn in the Sacramento-San Joaquin Delta of California. Irrig. Sci. 4: 31-44. Hoffman, G. J., J. A. Jobes, Z. Hanscom, and E. V. Maas. 1978. Timing of environmental stress affects growth, water relation and salt tolerance of Pinto beans. Trans. ASAE 21: 713-718. Hoffman, G. J., S. L. Rawlins, M. J. Garber, and E. M. Cullen. 1971. Water relations and growth of cotton as influenced by salinity and relative humidity. Agron. J. 63: 822-826. Hoffman, G. J., J. D. Rhoades, J. Letey, and F. Sheng. 1990. Chapt. 18: Salinity management. In Management of Farm Irrigation Systems, 667-715. G. J. Hoffman, T. A. Howell, and K. H. Solomon, eds. St. Joseph, Mich.: ASAE. John, C. D., V. Limpinuntana, and H. Greenway. 1976. Interaction of salinity and anaerobiosis in barley and rice. J. Exp. Bot. 28: 133-141. Keren, R., and S. Miyamoto. 1990. Reclamation of saline, sodic, and boron-affected soils. In Agricultural Salinity Assessment and Management, 410-431. K. K. Tanji, ed. ASCE Manuals and Reports on Engineering Practices No. 71. New York, N.Y.: American Soc. Civil Engineers. Keren, R., A. Meiri, and Y. Kalo. 1983. Plant spacing effect on yield of cotton irrigated with saline water. Plant Soil 74: 461-465. Kriedemann, P. E., and R. Sands. 1984. Salt resistance and adaptation to root zone hypoxia in sunflower. Aust. J. Plant Physiol. 11: 287-301. Kruse, E. G., D. A. Young, and D. F. Champion. 1985. Effects of saline water tables on corn irrigation. Proc. Specialty Conference, 444-453. New York, N.Y.: American Soc. Civil Engineers. Kruse, E. G., R. E. Yoder, D. L. Cuevas, and D. F. Champion. 1986. Alfalfa water use from high, saline water tables. ASAE Paper No. 86-2597. St. Joseph, Mich.: ASAE. Lee, E. W. 1990. Drainage water treatment and disposal solutions. In Agricultural Salinity Assessment and Management, 450-468. K. K. Tanji, ed. New York, N.Y.: American Soc. Civil Engineers. Letey, J., A. Dinar, and K. C. Knapp. 1985. Crop water production function model for saline irrigation water. Soil Sci. Soc. America J. 49: 1005-1009. Letey, J., C. Roberts, M. Penberth, and C. Vasek. 1987. An agricultural dilemma: Drainage water and toxic disposal in the San Joaquin Valley. Special Publication 3319. Div. of Agric. and Natural Resource, Univ. of Calif. Lunin, J., and M. H. Gallatin. 1965. Zonal salinization of the root system in relation to plant growth. Soil Sci. Soc. America Proc. 29: 608-612. Lunt, O. R., J. J. Oertli, and K. C. Kohl. 1960. Influence of environmental conditions on the salinity tolerance of several plant species. 7th Int. Congr. Soil Sci. 1: 560-570. Luthin, J. N., P. Fernandez, J. Woerner, and F. Robinson. 1969. Displacement front under ponded leaching. J. Irrig. Drain. Div., ASCE 95(IR1): 117-125. Mass, E. V. 1990. Crop salt tolerance. In Agricultural Salinity Assessment and Management, 262-304. K.K. Tanji, ed. ASCE Manuals and Reports on Engineering Practices, No. 71. New York, N.Y.: American Soc. Civil Engineers. Maas, E. V., and G. J. Hoffman. 1977. Crop salt tolerance: Current assessment. J. Irrig. Drain. Div., ASCE 103: 115-134.

Design and Operation of Farm Irrigation Systems

205

Maas, E. V., and R. H. Nieman. 1978. Physiology of plant tolerance to salinity. In Crop Tolerance to Suboptimal Land Conditions, 277-299. G. A. Jung, ed. Spec. Publ. 32. Maas, E. V., and J. A. Poss. 1989a. Salt sensitivity of wheat at various growth stages. Irrig. Sci.10: 313-320. Maas, E. V., and J. A. Poss. 1989b. Sensitivity of cowpea to salt stress at three growth stages. Irrig. Sci. 10: 29-40. Maas, E. V., S. R. Grattan, and G. Ogata. 1982. Foliar salt accumulation and injury in crops sprinkled with saline water. Irrig. Sci. 3: 157-168. Maas, E. V., J. A. Poss, and G. J. Hoffman. 1986. Salinity sensitivity of sorghum at three growth stages. Irrig. Sci. 7: 1-11. Maas, E. V., G. J. Hoffman, G. D. Chaba, J. A. Poss, and M. C. Shannon. 1983. Salt sensitivity of corn at various growth stages. Irrig. Sci. 4: 45-57. Magistad, O. C., A. D. Ayers, C. H. Wadleigh, and H. G. Gauch. 1943. Effect of salt concentration, kind of salt and climate on plant growth in sand cultures. Plant Physiol. 18: 151-166. Massoud, F. I. 1981. Salt affected soils at a global scale and concepts for control. Technical Paper. Rome, Italy: Land and Water Development Div., FAO. Meiri, A., G. J. Hoffman, M. C. Shannon, and J. A. Poss. 1982. Salt tolerance of two muskmelon cultivars under two radiation levels. J. American Soc. Hort. Sci. 170: 1168-1172. Meiri, A., J. Shalhevet, L. H. Stoley, G. Sinai, and R. Steinhardt. 1986. Managing multi-source irrigation water of different salinities for optimum crop production. BARD technical report #1-402-81. Bet Sagan, Israel: Volcani Center. Micklin, P. P. 1991. The Water Management Crisis in Soviet Central Asia. Pittsburgh, Pa.: Univ. Pittsburgh, Center for Russian and East European Studies. Miles, D. L. 1977. Salinity in the Arkansas Valley of Colorado. Interagency Agreement Report EPA-AIC-D4-0544. Denver, Colo.: EPA. Moore, J., and J. V. Hefner. 1977. Irrigation with saline water in the Pecos Valley of West Texas. In Proc. Intl. Salinity Conf. on Managing Saline Water for Irrigation, 339-344. Lubbock, Tex.: Texas Tech. Univ. Namken, L. N., C. L. Wiegand, and R. B. Brown. 1969. Water use by cotton from low and moderately saline static water tables. Agron. J. 61: 305-310. Nolan, S. L., T. H. Ashe, R. S. Lindstrom, and D. C. Martens. 1982. Effect of sodium levels on four foliage plants grown at two light levels. HortSci. 17: 815-817. Oster, J. D., and F. W. Schroer. 1979. Infiltration as influenced by irrigation water quality. Soil Sci. Soc. America J. 43: 444-447. Oster, J. D., and L. S. Willardson. 1971. Reliability of salinity sensors for the management of soil salinity. Agron. J. 63: 695-698. Oster, J. D., G. J. Hoffman, and F. E. Robinson. 1984. Management alternatives: Crop, water, and soil. California Agric. 38: 29-32. Peck, A. J., C. D. Johnston, and D. R. Williamson. 1981. Analyses of solute distributions in deeply weathered soils. Agric. Water Mgmt. 4: 83-102. Pratt, P. F. 1973. Quality criteria for trace elements in irrigation waters. Calif. Agric. Exp. Station Bulletin. Pratt, P. F., and D. L. Suarez. 1990. Irrigation water quality assessments. In Agricultural Salinity Assessment and Management, 220-236. K. K. Tanji, ed. New York, N.Y.: American Soc. Civil Engineers.

206

Chapter 7 Controlling Salinity

Raats, P. A. C. 1974. Steady flows of water and salt in uniform soil profiles with plant roots. Soil Sci. Soc. America Proc. 38: 717-722. Rains, D. W., S. Goyal, R. Weyranch, and A. Lauchli. 1987. Saline drainage water reuse in a cotton rotation system. Calif. Agric. 41(9 and 10): 24-26. Reeve, R. C., and E. J. Doering. 1966. The high salt water dilution method for reclaiming sodic soils. Soil Sci. Soc. America Proc. 30: 498-504. Rhoades, J. D. 1977. Potential for using saline agricultural drainage waters for irrigation. In Proc. Water Mgmt. for Irrig. & Drain., 85-116. Reston, Va.: American Soc. Civil Engineers. Rhoades, J. D. 1979. Inexpensive four-electrode probe for monitoring soil salinity. Soil Sci. Soc. America J. 43: 817-818. Rhoades, J. D. 1987. Use of saline water for irrigation. Water Quality Bull. 12: 14-20. Natl. Water Res. Inst., Ontario, Canada special bulletin, Water Quality, Burlington. Rhoades, J. D., and D. L Corwin. 1981. Determining soil electrical conductivity depth relations using an inductive electromagnetic soil conductivity meter. Soil Sci. Soc. America J. 45: 225-260. Rhoades, J. D., and R. D. Ingvalson. 1971. Determining salinity in field soils with soil resistance measurements. Soil Sci. Soc. America Proc. 35: 54-60. Rhoades, J. D., and J. van Schilfgaarde. 1976. An electrical conductivity probe for determining soil salinity. Soil Sci. Soc. America J. 40: 647-651. Rhoades, J. D., P. J. Shouse, W. J. Aloes, N. A. Monteghi, and S. M. Lesch. 1990. Determining soil salinity from soil electrical conductivity using different models and estimates. Soil Sci. Soc. America J. 54: 46-54. Rhoades, J. D., F. T. Bingham, J. Letey, A. R. Dedrick, M. Bean, G. J. Hoffman, W. J. Alves, R. V. Swain, P. G. Pacheco, and R. D. LeMert. 1988. Reuse of drainage water for irrigation: Results of Imperial Valley study. I: Hypothesis, experimental procedures and cropping results. Hilgardia 56: 1-16. Rolston, D. E., D. W. Rains, J. W. Biggar, and A. Lauchli. 1988. Reuse of saline drain water for irrigation. Presented at UCD/INIFAP Conference. Guadalajara, Mexico. Schwartz, M., and J. Gale. 1984. Growth response to salinity at high levels of carbon dioxide. J. Exp. Bot. 35: 193-196. Shalhevet, J. 1984. Management of irrigation with brackish water. In Soil Salinity under Irrigation, Processes and Management, Ecological Studies, 298-318. I. Shainberg, and J. Shalhevet, eds. New York, N.Y.: Springer-Verlag. Shalhevet, J., and L. Bernstein. 1968. Effects of vertically heterogeneous soil salinity on plant growth and water uptake. Soil Sci. 106: 85-93. Shalhevet, J., B. Heuer, and A. Meiri. 1983. Irrigation interval as a factor in salt tolerance of eggplant. Irrig. Sci. 4: 83-93. Shalhevet, J., P. Reiniger, and D. Shimshi. 1969. Peanut response to uniform and nonuniform soil salinity. Agron. J. 1: 384-387. Shalhevet, J., A. Vinten, and A. Meiri. 1986. Irrigation interval as a factor in sweet corn response to salinity. Agron. J. 78: 539-545. Shannon, M. C. 1980. Differences in salt tolerance within ‘Empire’ lettuce. J. American Soc. Hort. Sci. 105: 944-947. Shannon, M. C., and L. E. Francois. 1978. Salt tolerance of three muskmelon cultivars. J. American Soc. Hort. Sci. 103: 127-130. Shannon, M. C., and C. L. Noble. 1990. Genetic approaches for developing economic

Design and Operation of Farm Irrigation Systems

207

salt tolerant crops. In Agricultural Salinity Assessment and Management, 161-185. K. K. Tanji, ed. ASCE Manuals and Reports on Engineering Practices, No. 71. New York, N.Y.: American Soc. Civil Engineers. Shennan, C., S. Grattan, D. May, R. Burau, and B. Hanson. 1987. Potential for the long-term cyclic use of saline drainage water for the production of vegetable crops, 142-146. Technical Progress Report. U.C. Salinity/Drainage Task Force, Div. of Ag. and Natural Resources, Univ. of California. Stewart, J. I., R. M. Hagan, and W. O. Pruitt. 1976. Salinity effects on corn yield, evapotranspiration, leaching fraction and irrigation efficiency in managing saline water for irrigation. In Proc. Intl. Salinity Conf., 316-332. Suarez, D. L. 1981. Relationship between Ph and SAR and an alternative method of estimating SAR of soil or drainage water. Soil Sci. Soc. America J. 45: 469-475. Sylvester, M. A., J. P. Deason, H. R. Feltz, and R. A. Engberg. 1988. Preliminary results of the Department of Interior’s irrigation drainage studies. In Planning Now for Irrigation and Drainage in the 21st Century, Proc. of a conference sponsored by the Irrigation and Drainage Division of the American Society of Civil Engineers, 665-677. D. L. Hay, ed. New York, N.Y.: American Soc. Civil Engineers. Szabolcs, I. 1989. Salt Affected Soils. Boca Raton, Fla.: CRC Press, Inc. Taylor, A. C. 1965. Water history and the Indus Plain. Natural History 24(5): 40-49. Torres, C. B., and F. T. Bingham. 1973. Salt tolerance of Mexican wheat. I: Effects of NO3 and NaC1 on mineral nutrition, growth and grain production of four wheat varieties. Soil Sci. Soc. America Proc. 37: 711-715. Trofimenko, S. 1985. The state and density of the Aral Sea. Ambio. 14(3): 181-182. USDA. 1988. Water quality education and technical assistance plan. USDA–Soil Conservation Service and USDA–Extension Service Report. Washington, D.C.: USDA. Vinten, A., J. Shalhevet, A. Meiri, and J. Peretz. 1986. Water and leaching requirements of industrial tomatoes with brackish water. Irrig. Sci. 7: 13-25. Wadleigh, C. H., H. G. Gauch, and O. C. Magistad. 1946. Growth and rubber accumulation in guayule. USDA Tech. Bulletin No. 925. Washington, D.C.: USDA. Wadleigh, C. H., H. G. Gauch, and D. G. Strong. 1947. Root penetration and moisture extraction in saline soil by crop plants. Soil Sci. 63: 341-349. Wagenet, R. J., W. P. Campbell, A. M. Bamatraff, and D. L. Turner. 1980. Salinity, irrigation frequency, and fertilization effect on barley growth. Agron. J. 72: 969-974. Wallender, W. W., D. W. Grimes, D. W. Henderson, and L. K. Stromberg. 1979. Estimating the contribution of a perched water table to the seasonal evapotranspiration of cotton. Agron. J. 71: 1056-1060. West, D. W., and J. A. Taylor. 1984. Response of six grape cultivars to the combined effects of salinity and root zone waterlogging. J. American Soc. Hort. Sci. 109: 844-851. Wichelns, D., D. Nelson, and T. Weaver. 1988. Farm-level analyses of irrigated crop production in areas with salinity and drainage problems. San Joaquin Valley Drainage Program. Sacramento, Calif.: U.S. Bureau of Reclamation. Willey, G. R. 1953. Prehistoric settlement patterns in the Viru Valley, Peru. Smithsonian Inst., Bureau of American Ethnology, Bull. 155. Wash. D.C.: Smithsonian Inst. Wu, L., Z. Huang, and R. G. Burau. 1988. Selenium accumulation and selenium-salt co-tolerance in five grass species. Crop Sci. 28: 517-522.

CHAPTER 8

WATER REQUIREMENTS Richard G. Allen (University of Idaho, Kimberly, Idaho) James L. Wright (USDA-ARS, Kimberly, Idaho) William O. Pruitt (University of California, Davis, California) Luis S. Pereira (University of Lisbon, Lisbon, Portugal) Marvin E. Jensen (USDA-ARS, Fort Collins, Colorado) Abstract. Evapotranspiration (ET) calculation guidelines are based on the crop coefficient-reference evapotranspiration method (Kc ETref). Equations for the ASCEEWRI standardized Penman-Monteith method are provided for grass and alfalfa references, where the grass reference standardization follows the FAO Penman-Monteith procedure. Linearized FAO-style crop coefficients from FAO-56 and curvilinear coefficients from Wright are presented as both mean and as dual (basal) crop coefficients. ET coefficients for landscape utilize a decoupled procedure similar to that summarized by the Irrigation Association Water Management Committee. Guidelines for calculating irrigation water requirements and peak system design rates are described. Keywords. Evapotranspiration, Evaporation, Penman, Penman-Monteith, Irrigation requirement, Effective precipitation, Crop coefficient, Landscape coefficient.

8.1 INTRODUCTION Evapotranspiration (ET) from vegetation is the primary water requirement for agricultural crops. The quantification of ET is necessary for design and sizing of irrigation system components, for operating irrigation and water resources systems, for conducting water balances, and for conducting hydrologic analyses. Many types of systems are available for measuring ET, although direct measurement of ET is generally difficult or expensive. The primary focus of this chapter is estimation of ET using historical or real-time weather data, because design and operation of irrigation systems generally requires long and often continuous records of irrigation water requirements. Most standard, operational procedures for determining ET utilize the crop coefficient (Kc)-reference evapotranspiration (ETref) approach due to its simplicity, reproducibility, relatively good accuracy, and transportability among locations and climates. Two families of Kc curves are presented: the linear procedure of the FAO-56 Irrigation and Drainage Paper (Allen et al., 1998) and the curvilinear procedure of

Design and Operation of Farm Irrigation Systems

209

Wright (1982). In addition, within each family, two types of Kc are defined: mean or time-averaged Kc used for irrigation systems design and general planning, and dual Kc where basal transpiration and evaporation from soil are separated to increase accuracy. Two reference ET types are defined and described: clipped cool-season grass and fullcover alfalfa, since these are both in common usage within the U.S. Numerous ETref equations have been introduced during the past 50 years. Many of these have been described and evaluated in prior literature (Jensen 1974; Burman et al. 1980; Hatfield and Fuchs, 1990; Jensen et al. 1990). The three reference ET equations described in this chapter are the Penman-Monteith (PM) equation, the classical Penman equation and the 1985 Hargreaves equation. These equations are considered to be the most commonly used methods today and are appropriate and applicable to irrigation systems design and operation under a wide range of application situations and climates. The PM equation has been standardized to both clipped grass and alfalfa references by ASCE-EWRI (2005), and is considered to be the best method when a full complement of weather data is available (solar radiation, air temperature, humidity, and wind speed data). The 1985 Hargreaves equation and the PM equation with estimated weather data are recommended when only air temperature data are available. The FAO-24 pan evaporation method is also useful for situations where goodquality pan evaporation data are measured in conjunction with some weather data. That method is described in Doorenbos and Pruitt (1977), Jensen et al. (1990), and Allen et al. (1998). Other methods for estimating ET include crop simulation computer models that “grow” and simulate the evaporation and transpiration components of ET separately. These may also simulate the photosynthesis and respiration of plants (Ritchie and Otter, 1985, Jones et al., 1987). Complex ET models include “multi-layer” resistance equations that consider the effects of stomatal, leaf boundary, and aerodynamic characteristics for various portions of the plant canopy (Shuttleworth and Wallace, 1985; Dolman, 1993; Huntingford et al., 1995). However, data and time requirements to prepare model parameters for new or localized applications often place these models outside the realm of irrigation systems design and operation. The Kc-ETref method is a tried and true method that empirically incorporates many of the physiological and aerodynamic variables governing crop evapotranspiration. The Kc-ETref method, when applied carefully, can produce estimates of ET that are sufficiently accurate for irrigation systems design and operation (Allen et al., 2005c).

8.2 DEFINITIONS Several important quantities are defined in this section. Most of these definitions are common to agricultural literature. 8.2.1 Evapotranspiration The term evapotranspiration, abbreviated ET, is defined as the combined process by which water is converted from liquid or solid forms via evaporation from soil and wet plant surfaces and via evaporation of water from within plant tissue. The latter process is known as transpiration. ET can be expressed as the energy consumed as latent heat energy per unit area (and denoted as LE) or as the equivalent depth of evaporated water. Units for ET are typically mm t-1 where t denotes a time unit (hour, day, month, growing season, or year) and units for LE are typically W m-2 or MJ m-2 t-1. The term reference evapotranspiration has been used as a standardized and reproducible index approximating the climatic demand for water vapor and is generally

210

Chapter 8 Water Requirements

abbreviated ETref. Reference ET is the ET rate from an extensive surface of reference vegetation having a standardized uniform height and that is actively growing, completely shading the ground, has a dry but healthy and dense leaf surface, and is not short of water. This definition is applied to the standardized reference crops of grass (ETo) and alfalfa (ETr). The advantage of using the reference concept is that it enables the measurement and validation of estimated reference ET using living, standardized crops. In addition, because stomatal control of the alfalfa reference crop approximates that of most agricultural crops at full cover, ET from the reference crop is more like crop ET than is potential ET (Pereira et al., 1999). Crop evapotranspiration, abbreviated ETc, is defined as the rate of ET from an extensive surface of a specific crop. The ETc rate is influenced by crop growth stages, amount and frequency of wetting of the soil surface, environmental conditions, and crop management. Crop ET is usually less than about 1.3 ETo or 1.0 ETr when crop foliage does not completely shade the ground or when the crop has begun to mature and senesce.1 Crop ET may exceed ETo when the crop has substantially more leaf area, is taller, or has less stomatal control as compared to the clipped grass reference, or has a wet leaf or soil surface. Crop ET is normally expressed in units of mm h-1, mm d-1, mm month-1, or mm season-1, and is synonymous with the term consumptive use. ETc representing ET under any condition, ideal or nonideal, is termed “actual ETc” and is denoted as ETc act. The “extensive surface” in the definition of ETc and calculation methodologies implies that the crop covers a large enough area that the energy exchange at the crop surface, and the wind speed, temperature and humidity profiles above the crop, are in equilibrium. Only when this equilibrium exists do energy-based flux-profile equations such as the Penman and Penman-Monteith equations attain the highest accuracy. The extensive surface condition applies to field sizes having dimensions greater than about 200 m (equivalent to 4 ha or 10 acres). Landscape evapotranspiration, abbreviated ETL, is ET from residential and urban landscapes. Procedures for estimating ETL are similar to those for ETc, with two distinctions: (1) landscape systems are nearly always comprised of a mixture of multiple types and species of vegetation, thereby complicating the estimation of the landscape coefficient and ET; and (2) typically, the objective of landscape irrigation is to promote appearance rather than biomass production as is the case in agriculture. Therefore, target ET for landscape may include an intentional “stress” factor in the baseline value for ETc act. This adjustment can result in significant water conservation. A multicomponent approach for estimating ETL is described that covers a wide range of landscape vegetation and environmental conditions. This approach is similar to that adopted by the Irrigation Association (2003). 8.2.2 Effective Precipitation Effective precipitation, abbreviated Pe, was defined by Dastane (1974) in the context of irrigation water management as precipitation that “is useful or usable in any phase of crop production.” This definition implies that precipitation must infiltrate and remain resident in the effective root zone of a cropped field long enough to be ex1 Senescence describes the natural aging process of leaves whereby leaves begin to yellow and die, and stomatal function and exchange of carbon dioxide and water vapor reduce. Senescence may be accelerated by environmental stresses such as disease and water shortage.

Design and Operation of Farm Irrigation Systems

211

tracted by crop roots during the ET process or where evaporation from wet soil reduces some transpiration demand. Effective precipitation is discussed further in Sections 8.10 and 8.12, and in the context here includes evaporation from soil incidental to production agriculture and irrigation. 8.2.3 Irrigation Water Requirements The designer or operator of an irrigation system most frequently estimates irrigation water requirements (IR) for short periods and on a seasonal basis. Short-term estimates are needed for systems sizing and for day to day operations. Seasonal estimates are needed for allocation of water supplies and in water rights administration. The units for IR are often expressed as volume per unit area per unit time (for example, m3 ha-1 d-1) or as depth per unit time (for example, mm d-1). The “net” irrigation water requirement, abbreviated IRn, is defined as the depth of water needed to fulfill the ET requirement in excess of any effective precipitation for a disease-free crop growing in large fields under non-restricting soil and soil water conditions and under adequate fertility. In addition, IRn considers contributions of shallow ground water (GW) and change in stored soil water during the period of interest. In equation form: IRn = ETc − Pe − GW − Δ θ z s

(8.1) where Pe = effective precipitation during the period of calculation Δθ = change in soil water content in the root zone during the period of calculation zs = depth of soil experiencing the change in water content. Units for all terms are the same. In essence, IRn is the beneficially consumed portion of an irrigation application. The “gross” irrigation water requirement, abbreviated IR, includes the water required for the IRn in addition to water for leaching salts and to satisfy delivery and field system losses. Delivery system losses include spillage and seepage. Field system losses include surface runoff and deep percolation. In equation form: IR =

IRn ETc − Pe − GW − Δ θ z s = (1 − LR ) CF (1 − LR) CF

(8.2)

where LR is the leaching requirement (a fraction) and CF is the consumed fraction of applied water, generally equivalent to the so-called “irrigation efficiency.” Usually, LR is considered only when irrigation uniformity is high. Under low to moderate uniformity, deep percolation incidental to irrigation is often sufficient to fulfill LR. Calculation of the leaching requirement is described in Chapter 7.

8.3 DIRECT MEASUREMENTS System designers and operators obtain ET data from either direct field measurements or from estimates that are based on weather and crop data. Use of direct, realtime field measurements for ET in design and operation is rare due to the expense and operation and maintenance requirements of equipment. Direct measurements are primarily used to provide data for calibrating climatic and weather-based ET methods and for monitoring soil water conditions. Although direct measurements are essential for ET equation development or validation, the user must exercise caution when using these systems, data collected by these systems, and also coefficients calibrated from such data. Errors involved can be subtle, yet large. For example, eddy covariance data are prone to under-measurement by up to 30% (Twine

212

Chapter 8 Water Requirements

et al., 2000; Wilson et al., 2002). Bowen ratio (BR)-based estimates are prone to error and bias in net radiation and soil heat flux measurements (Payero et al., 2003). Both LE and BR estimates are sensitive to instrument siting relative to the roughness sublayer and evaporative footprint. Users should familiarize themselves with the problems associated with design, installation, maintenance, and operation of direct ET measuring devices that may influence the values and integrity of direct ET measurement data. One caution on measurement of ET is that measurements must represent conditions for which the data are applied. For irrigation systems design, measurements should be from or representative of ET from large expanses of vegetation where the ET process is essentially one-dimensional (upward). ET measurements should not be made from small groups of plants that are spatially isolated. Spatial isolation can increase transfer of heat and radiation energy from outside the measured group of plants to the plants, thereby increasing measured ET and making it unrepresentative of ET from large fields (Pruitt and Lourence, 1985; Meyer and Mateos, 1990; Pruitt, 1991; Allen et al., 1991b; Grebet and Cuenca, 1991). The reader is referred to other literature sources for background and description of various measurement methods, including lysimeters: WMO (1966), Tanner (1967), Aboukhaled et al. (1982), Pruitt and Lourence (1985), Allen et al. (1991a), and Howell et al. (1991); eddy covariance: Swinback (1951), Businger et al. (1967), Brutsaert (1982), Campbell and Tanner (1985), Tanner (1988), Twine et al. (2000), Wilson et al. (2002), and Munger and Loescher (2004); and Bowen ratio: Sellers (1965), Tanner (1967), Blad and Rosenberg (1974), Dyer (1974), Sinclair et al. (1975), Brutsaert (1982), Pruitt et al. (1987), and Payero et al. (2003).

8.4 ESTIMATION OF REFERENCE ET Evaporation of water requires relatively large amounts of energy obtained either from transfer of sensible heat from the air stream or from radiant energy. Therefore, the ET process is largely governed by energy exchange at the vegetation surface and is limited by the amount of energy available. Because of this limitation, it is possible to estimate the rate of ET based on a net balance of energy fluxes. This is the basis for the Penman and Penman-Monteith equations. 8.4.1 Reference Evapotranspiration Two types of reference crops have been widely applied. These are clipped, coolseason grass, such as fescue and perennial ryegrass, and full-cover alfalfa. Reference ET for clipped grass is commonly denoted as ETo and reference ET for alfalfa is commonly denoted as ETr. 8.4.2 Standardized Definitions and the Penman-Monteith Method Grass reference ETo was defined in FAO-24 (Doorenbos and Pruitt, 1977) as “the rate of evapotranspiration from an extensive surface of 8 to 15 cm tall, green grass cover of uniform height, actively growing, completely shading the ground and not short of water.” It is generally accepted that the grass reference crop is a “coolseason,” C-3 type of grass with roughness, density, leaf area, and bulk surface resistance characteristics similar to perennial ryegrass (Lolium perenne L.) or Alta fescue (Festuca arundinacea Schreb. ‘Alta’). Alfalfa reference ETr was defined by Wright and Jensen (1972) as: “... ET from well watered, actively growing alfalfa with 8 in. (20 cm) or more of growth ...” and by Wright (1982) as “... when the alfalfa crop was well watered, actively growing, and at

Design and Operation of Farm Irrigation Systems

213

least 30 cm tall; so that measured ET was essentially at the maximum expected level for the existing climatic conditions.” The height of alfalfa (Medicago sativa L., v. Ranger) in the data set used to develop the 1982 Kimberly Penman (Wright, 1982) and used in developing resistance algorithms for the Penman-Monteith method (Allen et al., 1989) ranged from about 0.15 to 0.80 m in height and averaged 0.47 m (unpublished data from Wright, 1985). Generally, ETr is about 1.1 to 1.4 times that of ETo due to the increased roughness and leaf area of alfalfa. The higher value (1.4) represents the ratio of ETr to ETo under extremely arid and windy conditions (minimum daytime relative humidity [RH] < 20% and wind speed > 5 m s-1 [11 mph]) and the lower value (1.1) represents the ratio of ETr to ETo under humid, calm conditions. Wright (referenced in Jensen et al. (1990), Table 6.9) has measured an average ratio of ETr to ETo at Kimberly, Idaho, equal to about 1.25. This value represents conditions common to the semiarid mountain states in the U.S. Because of the challenges in growing and maintaining a living reference crop, the PM equation has been used by the Food and Agriculture Organization (FAO) (Smith et al., 1991, 1996; Allen et al., 1998) and ASCE-EWRI (2005) to represent a fixed definition for ETo. The FAO definition for ETo in terms of the PM equation is “the rate of evapotranspiration from a hypothetical reference crop with an assumed crop height of 0.12 m, a fixed surface resistance of 70 s m-1 and an albedo of 0.23, closely resembling the evapotranspiration from an extensive surface of green grass of uniform height, actively growing, completely shading the ground and with adequate water” (Allen et al. 1998). ASCE-EWRI (2005) has adopted this same definition for standardization of ETo, with the provision for lower surface resistance (50 s m-1), when calculating on hourly or shorter timesteps. This lower resistance for hourly calculations has subsequently been adopted by FAO (Allen et al., 2006). ASCE-EWRI additionally defined a standardized ETr for the alfalfa reference, also based on the PM, where the standardized height is 0.5 m. The ASCE-EWRI (2005) standardized PM method has the form: Cn u2 (es − ea ) T + 273 Δ + γ (1 + Cd u 2 )

0.408 Δ ( Rn − G ) + γ ETref =

(8.3)

where ETref = standardized reference ET for short (ETos) or tall (ETrs) surfaces, mm d-1 for daily timesteps or mm h-1 for hourly timesteps Rn = calculated net radiation at the crop surface, MJ m-2 d-1for daily timesteps or MJ m-2 h-1 for hourly timesteps G = soil heat flux density at the soil surface, MJ m-2 d-1 for daily timesteps or MJ m-2 h-1 for hourly timesteps T = mean daily or hourly air temperature at 1.5 to 2.5-m height, °C u2 = mean daily or hourly wind speed at 2-m height, m s-1 es = saturation vapor pressure at 1.5- to 2.5-m height, kPa, calculated for daily timesteps as the average of saturation vapor pressure at maximum and minimum air temperature ea = mean actual vapor pressure at 1.5 to 2.5-m height, kPa Δ = slope of the saturation vapor pressure-temperature curve, kPa °C-1 γ = the psychrometric constant, kPa °C-1

214

Chapter 8 Water Requirements

Table 8.1. Values for Cn and Cd in Equation 8.3 (from ASCE-EWRI, 2005). Tall Reference, Short Reference, Units for Units for ETrs ETos ETos, R n, (clipped grass) (alfalfa) ETrs G Calculation Timestep Cn Cd Cn Cd -1 Daily 900 0.34 1600 0.38 mm d MJ m-2 d-1 -1 Hourly during daytime 37 0.24 66 0.25 mm h MJ m-2 h-1 -1 Hourly during nighttime 37 0.96 66 1.7 mm h MJ m-2 h-1

Cn = a numerator constant that changes with reference type and calculation time step, K mm s3 Mg-1 d-1 or K mm s3 Mg-1 h-1 Cd = a denominator constant that changes with reference type and calculation timestep, s m-1. Table 8.1 provides values for Cn and Cd. The values for Cn consider the timestep and aerodynamic roughness of the surface (i.e., reference type). The constant in the denominator, Cd, considers the timestep, bulk surface resistance, and aerodynamic roughness of the surface. Cn and Cd were derived by simplifying several terms within the ASCE PM equation of ASCE Manual 70 (Allen et al., 1989; Jensen et al., 1990) and rounding the result. Units for the 0.408 coefficient are m2 mm MJ-1. Daytime is defined as occurring when Rn during an hourly period is positive. The ASCE-EWRI definition uses smaller values for surface resistance for hourly or shorter calculation timesteps (during daytime) than for daily calculation timesteps. The FAO-PM (Allen et al., 1998) is equivalent to Equation 8.3, where Cn = 900 and Cd = 0.34. ASCEEWRI (2005) and Allen et al. (2006) recommended applying the FAO-PM equation for ETo using the Cn and Cd coefficients for hourly and shorter calculation timesteps in Table 1. Daily ETref are obtained by summing ETref values from hourly or shorter periods or by applying the ETref equations on a 24 h-timestep. Generally, summing hourly calculations is considered to be more accurate. ASCE-EWRI recommended application of ETos and ETrs for calculating reference evapotranspiration and development of new crop coefficients, and for facilitating transfer of crop coefficients. 8.4.3 The Classical Penman Equation Besides the standardized Penman-Monteith equation that is now recommended by FAO (Allen et al., 1998) and by ASCE-EWRI (2005), some forms of the original Penman equation (Penman, 1948) are still in use that perform similarly to the ASCE PM. The classical form of the Penman equation is: ET ref = 0.408

Δ (Rn − G ) + K w γ (a w + bw u 2 ) (es − ea ) Δ +γ Δ +γ

(8.4)

where terms and definitions are the same as those used for the PM equation in Equation 8.3. Parameter Kw is a units parameter, with Kw = 2.62 for ETref in mm d-1 and Kw = 0.109 for ETref in mm hour-1. The aw and bw terms are empirical wind coefficients that have often been locally or regionally calibrated. The values for aw and bw were 1.0 and 0.537, respectively, in the original Penman equation (Penman, 1948, 1963) for grass ETo, for wind speed in m s-1, (es – ea) in kPa and ETo in mm d-1. Generally, local calibration of the Penman method is neither required nor recommended and was often an artifact of earlier studies required to overcome biases or faulty measurements in

Design and Operation of Farm Irrigation Systems

215

weather and ET data. Currently the commonly used Penman values for aw and bw are those for the hourly CIMIS Penman method for ETo and daily 1982 Kimberly Penman method for ETr. 8.4.3.1 CIMIS Penman equation for grass reference. Pruitt (Pruitt and Doorenbos 1977a; Pruitt et al., 1984; Snyder and Pruitt, 1992) developed aw and bw for estimating grass ETo for hourly periods. These coefficients have been used since the 1980s for standard ETo estimation in the California Irrigation Management Information Service (CIMIS). The CIMIS Penman ETo equation uses aw = 0.29 and bw = 0.53 for Rn > 0, and aw = 1.14 and bw = 0.40 for Rn < 0, and is applied hourly, where in Equation 8.4, ETo = mm h-1, Rn = MJ m-2 h-1, and Kw = 0.109. Standard CIMIS calculations assume G = 0. When applied hourly, ETo from the CIMIS Penman equation is summed over 24-h periods to obtain daily totals, similar to what is done using the PM method. 8.4.3.2 1982 Kimberly-Penman alfalfa reference. The 1982 Kimberly-Penman equation was developed from intensive studies of ET using precision weighing lysimeters at Kimberly, Idaho (Wright and Jensen 1972; Wright, 1981, 1982, 1988) and is intended for application with 24-hour timesteps (Kw = 2.62 in Equation 8.4). The Kimberly-Penman wind function is calibrated to estimate ET from full-cover alfalfa, where the aw and bw wind function coefficients vary with time of year (Jensen et al. 1990): ⎧⎪ ⎡ J − 173 2⎤ ⎛ ⎞ ⎟ ⎥ a w = 0.4 + 1.4 exp ⎨− ⎢ ⎜ ⎪⎩ ⎢⎣ ⎝ 58 ⎠ ⎥⎦ ⎧ ⎪ b w = 0.605 + 0.345 exp⎨− ⎪⎩

⎫⎪ ⎬ ⎪⎭

⎡⎛ J − 243 ⎞ 2⎤ ⎫⎪ ⎢⎜ ⎟ ⎥⎬ ⎢⎣⎝ 80 ⎠ ⎥⎦ ⎪ ⎭

(8.5)

(8.6)

where J is the day of the year. For latitudes south of the equator, Equation 8.5 and 8.6 are applied using J' in place of J, where J' = (J – 182) for J ≥ 182 and J' = (J + 182) for J < 182. The (es – ea) term in the 1982 Kimberly Penman is computed the same as for the standardized PM equation (as the average of es computed at daily maximum and minimum temperatures; Equation 8.7). 8.4.4 Computation of Parameters for the Penman-Monteith and Penman Equations It is recommended that standardized procedures and equations be used to calculate parameters in ETref equations. This insures agreement among independent calculations and simplifies calculation verification. Equations presented in this section follow procedures standardized by FAO-56 and ASCE-EWRI (2005). 8.4.4.1 Saturation vapor pressure of the air. For 24-hour or longer calculation timesteps, es, the saturation vapor pressure of the air, is computed as:

eo (T max) + eo (T min) (8.7) 2 where Tmax and Tmin are daily maximum and minimum air temperature, °C, at the measurement height (1.5 to 2 m), and e° is the saturation vapor pressure function. For hourly applications, es is calculated as e°(T) where T is average hourly air temperature. The saturation vapor pressure function is: es =

216

Chapter 8 Water Requirements

⎛ 17.27 T ⎞ e o (T ) = 0.6108 exp ⎜ ⎟ ⎝ T+ 237.3 ⎠

(8.8)

where e°(T) is in kPa and T is in °C (Tetens, 1930). 8.4.4.2 Actual vapor pressure of the air. Actual vapor pressure of the air, ea, is equivalent to saturation vapor pressure at the dew point, Td. For 24-h or longer timesteps, Td is taken as mean daily or early morning dewpoint temperature, °C. Humidity of the air can be measured using many different methods, including relative humidity sensors, dewpoint sensors, and wet bulb/dry bulb psychrometers, so that ea can be calculated using several ways. The recommended procedures, in order of what are considered to be the most reliable to the least reliable, are (ASCE-EWRI, 2005): 1. For 24-h periods, averaging ea measured or computed hourly over the 24-h period. 2. For 24-h periods, calculating ea from dewpoint, Td, that is measured or computed hourly over the 24-h period: ⎛ 17.27 T d ⎞ ⎟⎟ e a = eo ( T d ) = 0.6108 exp ⎜⎜ ⎝ T d + 237.3 ⎠

(8.9)

where ea is in kPa and Td is in °C. 3. Psychrometer measurements using dry and wet bulb thermometers, where (Bosen, 1958):

(

ea = e o (Twet ) − γ psy Tdry − Twet

)

(8.10)

where e°(Twet) is saturation vapor pressure at the wet bulb temperature in kPa (Equation 8.8), γpsy is the psychrometric constant for the psychrometer, in kPa °C-1, and Tdry – Twet is the wet bulb depression, where Tdry is dry bulb temperature and Twet is the wet bulb temperature, both in °C. For 24-h periods, Tdry and Twet are preferably averaged over the 24-h period. Alternatively, once-per-day readings of Tdry and Twet can be used, for example, readings taken at 7 or 8 a.m. Tdry and Twet must be measured simultaneously. The psychrometric constant for the psychrometer is:

γ psy = a psy P

(8.11)

where apsy is a coefficient depending on the type of ventilation of the wet bulb, in °C-1, and P is the mean atmospheric pressure, in kPa. The coefficient apsy depends primarily on the design of the psychrometer and on the rate of ventilation around the wet bulb. The following values are often used: apsy = 0.000662 for ventilated (Asmann type) psychrometers having air movement between 2 and 10 m s-1 for Twet > 0 and 0.000594 for Twet < 0 = 0.000800 for naturally ventilated psychrometers having air movement of about 1 m s-1 = 0.001200 for non-ventilated psychrometers installed in glass or plastic greenhouses. 4. For 24-hour or longer timesteps, relative humidity (RH) measurements taken twice daily (early morning, corresponding to Tmin and early afternoon, corresponding to Tmax) can be combined to yield an approximation for 24-hour average ea:

Design and Operation of Farm Irrigation Systems

217

RHmax RH + e o (Tmax ) min 100 100 ea = 2 (8.12) where RHmax is daily maximum relative humidity (%) (during early morning), and RHmin is daily minimum relative humidity (%) (during early afternoon, around 1400 hours). For hourly calculations, ea is calculated from relative humidity as: e o (Tmin )

ea =

RH o e (T ) 100

(8.13)

where RH is mean relative humidity for the hourly or shorter period, %, and T is mean air temperature for the hourly or shorter period, °C. 5. From daily RHmax and Tmin: RHmax ea = e o (Tmin ) 100

(8.14)

6. From daily RHmin and Tmax:

RHmin (8.15) 100 7. If daily humidity data are missing or are of questionable quality, ea can be approximated for the reference environment assuming that Td is near Tmin: Td = Tmin − K o (8.16) ea = e o (Tmax )

where Ko is approximately 2° to 4°C in dry seasons in semiarid and arid climates and Ko is approximately 0°C in humid to subhumid climates and the rainy season in semi-arid climates (ASCE-EWRI, 2005). 8. In the absence of RHmax and RHmin data, but where daily mean RH data are available, ea can be estimated as: ea =

RH mean o e (Tm ean ) 100

(8.17)

where RHmean is mean daily relative humidity, generally defined as the average between RHmax and RHmin, and Tmean is mean daily air temperature. For daily or longer periods, Equation 8.17 is generally less desirable than methods 1-7 for calculating ea because the e°(T) relationship is nonlinear. 8.4.4.3 Psychrometric constant. The psychrometric constant (γ) in the Penman and PM equations is calculated as (Brunt, 1952):

γ = 0.000665 P

(8.18)

-1

where P has units of kPa and γ has units of kPa °C . 8.4.4.4 Atmospheric pressure. For purposes of ET estimation, mean atmospheric pressure, P, can be based on elevation using an equation simplified from the universal gas law relationship (Burman et al., 1987, ASCE-EWRI, 2005): P = (2.406 - 0.0000534 z ) 5.26

(8.19)

where P has units of kPa and z is the weather site elevation (m) above mean sea level.

218

Chapter 8 Water Requirements

8.4.4.5 Slope of the saturation vapor pressure-temperature curve. The slope of the saturation vapor pressure-temperature curve, Δ, is computed as:

Δ=

⎛ 17.27 T ⎞ ⎟⎟ 2503 exp ⎜⎜ ⎝ T + 237.3 ⎠ (T + 237.3)2

(8.20)

where Δ has units of kPa °C-1 and T is daily or hourly mean air temperature in °C. 8.4.4.6 Wind speed at 2 meters. Wind speed varies with height above the ground surface. For the calculation of standardized ETref, the wind speed measurement must be made equivalent to that 2 m above the surface. Therefore, wind measured at other heights is adjusted by: 4.87 (8.21) u2 = u z ln(67.8 z w − 5.42) where u2 = wind speed at 2 m above ground surface, m s-1 uz = measured wind speed at zw m above ground surface, m s-1 zw = height of wind measurement above the ground surface, m. Equation 8.21 is used for measurements taken above a short grass (or similar) surface, based on the logarithmic wind speed profile equation. For wind measurements made above surfaces other than clipped grass, the user should apply a full logarithmic equation that considers the influence of vegetation height and roughness on the shape of the wind profile. These alternative adjustments are described in Allen and Wright (1997) and ASCE-EWRI (2005). Wind speed data collected at heights above 2 m are acceptable to use in the standardized equations following adjustment to 2 m, and are preferred if vegetation adjacent to the weather station commonly exceeds 0.5 m. Measurement at a greater height, for example 3 m, reduces the influence of the taller vegetation surrounding the weather measurement site. 8.4.4.7 Net radiation. Net radiation, Rn, in the context of ET, is the net amount of radiant energy available at a vegetation or soil surface for evaporating water, heating the air, or heating the surface. Rn includes both short and long wave radiation components. Net radiation flux density can be measured directly using hemispherical net radiometers. However, measurement is difficult because net radiometers are problematic to maintain and calibrate, creating the likelihood of systematic biases. Therefore, Rn is often estimated from observed short wave (solar) radiation, vapor pressure, and air temperature. This estimation is routine and generally highly accurate for well defined vegetated surfaces. If Rn is measured, then care and attention must be given to the calibration of the radiometer, the surface over which it is located, maintenance of the sensor domes, and level of the instrument. The condition of the vegetation surface is as important as the sensor. For purposes of calculating reference ET, the measurement surface for Rn is generally assumed to be clipped grass or alfalfa at full cover. When calculated, Rn is estimated from the short wave and long wave components: Rn = Rns – Rnl (8.22) -2 -1 -2 -1 where Rns is net short-wave radiation (in MJ m d or MJ m h ), defined as being positive downwards and negative upwards, and Rnl is net outgoing long-wave radiation (in MJ m-2 d-1 or MJ m-2 h-1), defined as being positive upwards and negative downwards. Rns and Rnl are generally positive or zero in value.

Design and Operation of Farm Irrigation Systems

219

Net short-wave radiation resulting from the balance between incoming and reflected solar radiation is given by: Rns = Rs − α Rs = (1 − α ) Rs (8.23) where α is albedo, fixed in the ASCE-EWRI standardization at 0.23 for both daily and hourly timesteps (dimensionless), and Rs is incoming solar radiation (MJ m-2 d-1 or MJ m-2 h-1). Users may elect to use a different estimation for albedo; however, it is essential to ascertain the validity and accuracy of an alternative method using good measurements of incoming and reflected solar radiation. The standardized procedure of FAO-56 and ASCE-EWRI for estimating Rnl is based on the Brunt (1932, 1952) approach for estimating net emissivity. Users may choose to utilize a different approach for calculating Rnl; however, as with albedo it is essential to ascertain the validity and accuracy of the Rn method using net radiometers in excellent condition and that are calibrated to some dependable and recognized standard. Rnl, net long-wave radiation, is the difference between upward long-wave radiation from the standardized surface (Rlu) and downward long-wave radiation from the sky (Rld), so that Rnl = Rlu – Rld. The ASCE-EWRI procedure for daily Rnl is: ⎡ TK4 max + TK4 min ⎤ ⎥ (8.24) daily Rnl = σ f cd 0.34 − 0.14 ea ⎢ 2 ⎢ ⎥ ⎣ ⎦

(

and for hourly Rnl is:

)

(

)

hourly Rnl = σ f cd 0.34 − 0.14 ea TK4 hr

(8.25)

where Rnl = net outgoing long-wave radiation, MJ m-2 d-1or MJ m-2 h-1 σ = Stefan-Boltzmann constant, 4.901 × 10-9 MJ K-4 m-2 d-1 or 2.042 × 10-10 MJ K-4 m-2 h-1 fcd = a cloudiness function (dimensionless) and limited to 0.05 < fcd < 1.0 (see below) ea = actual vapor pressure , kPa TK max = maximum absolute temperature during the 24-hour period, K (recall that K = °C + 273.15) TK min = minimum absolute temperature during the 24-hour period, K TK hr = mean absolute temperature during the hourly period, K. The superscripts “4” in Equations 8.24 and 8.25 indicate the need to raise the air temperature, expressed in Kelvin units, to the power of 4. For daily and monthly calculation timesteps, fcd is calculated as: f cd = 1.35

Rs − 0.35 Rso

(8.26)

Rs is measured or calculated solar radiation and Rso is calculated clear-sky radiation, both in MJ m-2 d-1. The relative solar radiation, Rs/Rso, represents relative cloudiness and is limited to 0.3 < Rs /Rso < 1.0 so that fcd has limits of 0.05 < fcd < 1.0. For hourly periods during daytime when the sun is more than about 15° above the horizon, fcd is calculated using Equation 8.26 with the same limits applied. For hourly periods during nighttime, Rso, by definition, equals 0, and Equation 8.26 is undefined. Therefore, fcd during periods of low sun angle and during nighttime is defined using fcd

220

Chapter 8 Water Requirements 2

from prior periods. When sun angle (β) above the horizon at the midpoint of the hourly or shorter time period is less than 0.3 radians (~17°), then: fcd = fcd β >0.3

(8.27)

where fcd β >0.3 is the cloudiness function (dimensionless) for the time period prior to when sun angle β (in the afternoon or evening) falls below 0.3 radians. If the calculation timestep is shorter than one hour, then fcd from several periods can be averaged into fcd β >0.3 to obtain a representative average value. In mountain valleys where the sun may set near or above 0.3 radians (~17°), the user should increase the sun angle at which fcd β >0.3 is computed and imposed. For example, for a location where mountain peaks are 20° above the horizon, a period should be selected for computing fcd β >0.3 where the sun angle at the end of the time period is 25° to 30° above the horizon. The same adjustment is necessary when deciding when to resume computation of fcd during morning hours when mountains lie to the east. Only one value for fcd β >0.3 is calculated per day for use during dusk, nighttime, and dawn periods. That value for fcd β >0.3 is then applied to the time period when β at the midpoint of the period first falls below 0.3 radians (~17°) and to all subsequent periods until after sunrise when β again exceeds 0.3 radians. Equations 8.26 and 8.27 will not apply at latitudes and times of the year when there are no hourly (or shorter) periods having sun angle of 0.3 radians or greater. These situations occur at latitudes of 50° for about one month per year (in winter), at latitudes of 60° for about five months per year, and at latitudes of 70° for about seven months per year (ASCE-EWRI, 2005). Under these conditions, the application can average fcd β> 0.3 from fewer time periods or, in the absence of any daylight, can assume a ratio of Rs/Rso ranging from 0.3 for complete cloud cover to 1.0 for no cloud cover. Under these extreme conditions, the estimation of Rn is only approximate. The application of Equation 8.27 presumes that cloudiness during periods of low sun angle and nighttime is similar to that during late afternoon or early evening. This is generally a reasonable assumption and is commensurate with the relative simplicity and moderate accuracy of the procedure. 8.4.4.8 Clear-sky solar radiation (Rso). Clear-sky solar radiation (Rso) is used in the calculation of net radiation (Rn). Rso is defined as the amount of short-wave radiation that would be received at the weather measurement site under conditions of clearsky (i.e., cloud-free). Daily Rso is a function of the time of year and latitude and is impacted by station elevation (affecting atmospheric thickness and transmissivity), the amount of precipitable water in the atmosphere (affecting the absorption of some short-wave radiation), and the amount of dust or aerosols in the air. Rso, for purposes of calculating Rn, can be computed as: (8.28) Rso = (0.75 + 2 × 10-5 z) Ra where z is station elevation above sea level in m. More complex estimates for Rso, which include impacts of turbidity and water vapor on radiation absorption, can be used for assessing integrity of solar radiation data and are described in Appendix D of ASCE-EWRI (2005). 2 The sun angle β is defined as the angle of a line from the measurement site to the center of the sun’s disk relative to a line from the measurement site to directly below the sun and tangent to the earth’s surface. This definition assumes a flat surface.

Design and Operation of Farm Irrigation Systems

221

8.4.4.9 Exoatmospheric radiation. Exoatmospheric radiation, Ra, (also known as extraterrestrial radiation) is defined as the short-wave solar radiation in the absence of an atmosphere, and is a well-behaved function of the day of the year, time of day, and latitude. Ra is needed for calculating Rso, which is in turn used in calculating Rn. For daily (24-hour) periods, Ra is estimated from the solar constant, the solar declination, and the day of the year: Ra =

24 Gsc d r [ω s sin(ϕ ) sin(δ ) + cos(ϕ ) cos(δ ) sin(ω s )] π

(8.29)

where Ra = exoatmospheric radiation, MJ m-2 d-1 Gsc = solar constant, 4.92 MJ m-2 h-1 dr = inverse relative distance factor (squared) for the earth-sun, unitless ωs = sunset hour angle, radians ϕ = station latitude, radians, positive for the northern hemisphere and negative for the southern hemisphere δ = solar declination, radians. Parameters dr and δ are calculated as: ⎛ 2π ⎞ (8.30) d r = 1 + 0.033 cos ⎜ J⎟ ⎝ 365 ⎠ ⎛2π ⎞ J − 1.39 ⎟ ⎝ 365 ⎠

δ = 0.409 sin ⎜

(8.31)

where J is the number of the day in the year between 1 (1 January) and 365 or 366 (31 December). The constant 365 in Equations 8.23 and 8.24 is held at 365 even during a leap year. J can be calculated as: M ⎞ ⎛ ⎛ 3 ⎞ ⎛ M Mod(Y , 4) ⎞ J = DM - 32+ Int ⎜ 275 ⎟ + 2 Int ⎜ + 0.975 ⎟ ⎟ + Int ⎜ 9 ⎠ 4 ⎝ ⎝ M + 1⎠ ⎝ 100 ⎠

(8.32a)

where DM is the day of the month (1-31), M is the number of the month (1-12), and Y is the number of the year (for example 1996 or 96). The “Int” function in Equation 8.32 finds the integer number of the argument in parentheses by rounding downward. The “Mod(Y,4)” function finds the modulus (remainder) of the quotient Y/4. For monthly periods, the day of the year at the middle of the month (Jmonth) is approximately: J month = Int(30.4 M − 15 ) (8.32b) The sunset hour angle, ωs, is given by:

ω s = arccos [ − tan (ϕ ) tan (δ )]

(8.33)

The “arccos” function is the arc-cosine function and represents the inverse of the cosine. This function is not available in all computer languages, so that ωs can alternatively be computed using the arc-tangent (inverse tangent) function:

ωs = where

⎡ − tan(ϕ ) tan(δ ) ⎤ π − arctan ⎢ ⎥ 2 X 0.5 ⎣ ⎦

X = 1 − [tan(ϕ )] 2 [tan(δ )] 2

(8.34) (8.35)

222

Chapter 8 Water Requirements

and X = 0.00001 if X ≤ 0

For hourly time periods, the solar time angle at the beginning and end of the period serve as integration endpoints for calculating Ra: Ra =

12 Gsc d r [(ω2 −ω1 )sin(ϕ)sin(δ )+cos(ϕ)cos(δ ) (sin(ω2 )−sin(ω1) )] π

where Ra = exoatmospheric radiation, MJ m-2 hour-1 Gsc = solar constant, 4.92 MJ m-2 h-1 ω1 = solar time angle at beginning of period, radians ω2 = solar time angle at end of period, radians. πt ω1 and ω2 are given by ω1 = ω − 1 24

ω2 = ω +

π t1 24

(8.36)

(8.37)

(8.38)

where ω is solar time angle at the midpoint of the period (radians), and tl is the length of the calculation period (hour): i.e., 1 for hourly periods or 0.5 for 30-minute periods. The solar time angle at the midpoint of the hourly or shorter period is: π ω = {[t + 0.06667( Lz − Lm ) + Sc ] − 12} (8.39) 12 where t = standard clock time at the midpoint of the period (hour) after correcting time for any daylight savings shift. For example, for a period between 1400 and 1500, t = 14.5 hours. Lz = longitude of the center of the local time zone, expressed as positive degrees west of Greenwich, England. In the U.S., Lz = 75°, 90°, 105°, and 120° for the Eastern, Central, Rocky Mountain, and Pacific time zones, respectively, and Lz = 0° for Greenwich, 345° for Paris, France, and 255° for Bangkok, Thailand. Lm = longitude of the solar radiation measurement site, expressed as positive degrees west of Greenwich, England Sc = a seasonal correction for solar time, hours. Because ωs is the sunset hour angle and –ωs is the sunrise hour angle (noon has ω = 0), values of ω < –ωs or ω > ωs from Equation 8.39 indicate that the sun is below the horizon, so that, by definition, Ra and Rso are zero and their calculation has no meaning. When the values for ω1 and ω2 span the value for –ωs or for ωs, this indicates that sunrise or sunset occurs within the hourly (or shorter) period. In this case, the integration limits for applying Equation 8.36 should be correctly set using the following conditionals: If ω1 < –ωs then ω1 = –ωs If ω2 < –ωs then ω2 = –ωs If ω1 > ωs then ω1 = ωs If ω2 > ωs then ω2 = ωs If ω1 > ω2 then ω1 = ω2

(8.40)

Design and Operation of Farm Irrigation Systems

223

The above conditionals can be applied for all timesteps to insure numerical stability of the application of Equation 8.36 as well as correctly computing the theoretical quantity of solar radiation early and late in the day. Where there are hills or mountains, the hour angle when the sun first appears or disappears may increase for sunrise or decrease for sunset; however, generally, no corrections are necessary. The seasonal correction for solar time is: S c = 0.1645 sin(2 b) − 0.1255 cos(b) − 0.025 sin(b) b=

2π(J −81) 364

(8.41) (8.42)

where J is the number of the day in the year and b has units of radians. The user should confirm accurate setting of the datalogger clock. If clock times are in error by more than 5–10 minutes, estimates of exoatmospheric and clear sky radiation may be significantly impacted. This can lead to errors in estimating Rn on an hourly or shorter basis, especially early and late in the day. A shift in “phase” between measured Rs and Rso estimated from Ra according to the data logger clock can indicate error in the reported time. Discussion is given in Appendix D of ASCE-EWRI (2005). The angle of the sun above the horizon, β, at the midpoint of the hourly or shorter time period is computed as:

β = arcsin[sin(ϕ )sin(δ )+cos(ϕ )cos(δ )cos(ω) ]

(8.43)

where β = angle of the sun above the horizon, radians ϕ = station latitude, radians δ = solar declination, radians ω = solar time angle at the midpoint of the period, radians. The “arcsin” function is the arc-sine function and represents the inverse of the sine. This function is not available in all computer languages, but β can alternatively be computed using the arc tangent (inverse tangent) function: ⎡

β = arctan ⎢

Y

(

⎢ 1−Y ⎣

where

)

2 0.5

⎤ ⎥ ⎥ ⎦

Y = sin(ϕ )sin(δ ) + cos(ϕ )cos(δ )cos(ω )

(8.44) (8.45)

and all other parameters are as defined previously. 8.4.4.10 Soil heat flux. The magnitude of soil heat storage or release can be significant in the surface energy balance over a period of a few hours, but is usually small day to day because heat stored early in the day as the soil warms is lost late in the day or at night when the soil cools. Since the magnitude of 24-hour average soil heat flux under a crop canopy over 10- to 30-day periods is relatively small, G normally can be neglected for daily and longer timesteps. The total G over a complete growing period, however, may be significant, especially for 30 days or longer. As defined here and by ASCE-EWRI (2005), G is positive when the soil is warming and negative when the soil is cooling. For daily periods, the magnitude of G averaged over 24 hours beneath a fully vegetated grass or alfalfa reference surface is relatively small in comparison with Rn. Therefore, it is ignored in the standardized ET calculations so that

224

Chapter 8 Water Requirements

(8.46) Gday = 0 where Gday is the daily soil heat flux density, MJ m-2 d-1. Over a monthly period, G for the soil profile can be significant, especially during spring and fall. Assuming a constant soil heat capacity of 2.0 MJ m-3 °C-1 and an effectively warmed soil depth of 2 m, G for monthly periods in MJ m-2 d-1 is estimated from the change in mean monthly air temperature as: Gmonth,i = 0.07 (Tmonth,i +1 − Tmonth,i−1 )

(8.47)

or, if Tmonth,i+1 is unknown, Gmonth,i = 0.14 (Tmonth,i − Tmonth,i−1 )

(8.48)

where Tmonth,i is mean air temperature of month i, Tmonth,i-1 is mean air temperature of the previous month, and Tmonth,i+1 is the mean air temperature of the next month (all °C). For application to short-periods, e.g., when hourly data for G are required, other approaches must be used. One method common to research uses heat flux plates installed near the soil surface, usually at a depth of about 0.1 to 0.15 m. The total heat flux density is determined by performing a calorimetric balance of the soil layer above the plate. Because soil heat flux plates are not a common measurement at weather stations, the reader is referred to other references (Tanner, 1960; Brutsaert, 1982; and Allen et al., 1996) for application descriptions. For hourly or shorter time periods, G, in the standardized calculation is expressed as a function of net radiation for the two reference types. For the standardized short reference ETos : Ghr daytime = 0.1 Rn (8.49a) (8.49b) Ghr nightime = 0.5 Rn where G and Rn have the same measurement units (MJ m-2 h-1 for hourly or shorter time periods). For the standardized tall reference ETrs : (8.50a) Ghr daytime = 0.04 Rn (8.50b) Ghr nightime = 0.2 Rn For standardization, nighttime is defined as when measured or calculated hourly net radiation Rn is < 0 (i.e., negative). The amount of energy consumed by G is subtracted from Rn when estimating ETos or ETrs. The coefficient 0.1 in Equation 8.49a represents the condition of only a small amount of dead thatch underneath the leaf canopy of the short (clipped grass) reference. Large amounts of thatch insulate the soil surface, reducing the daytime coefficient for grass to about 0.05. However, the 0.1 coefficient is part of the EWRI and FAO standardizations. 8.4.5 1985 Hargreaves Grass Reference Equation The Hargreaves equation (Hargreaves and Samani, 1982, 1985; Hargreaves et al., 1985) is suggested as a means for estimating ETo in situations where weather data are limited and only maximum and minimum air temperature data are available. The form of the 1985 Hargreaves equation is: ETo = 0.0023 (Tmax – Tmin)0.5 (Tmean + 17.8) (Ra) (8.51) where Tmax and Tmin = maximum and minimum daily air temperature, °C,

Design and Operation of Farm Irrigation Systems

225

Tmean = mean daily air temperature, (Tmax + Tmin)/2 Ra = average daily exoatmospheric radiation (Equation 8.29). ETo in Equation 8.51 has the same units as Ra and can be converted from MJ m-2 d-1 to mm d-1 by dividing by λ = 2.45 MJ kg-1. The Hargreaves equation was the highest-ranked temperature-based method in the ASCE Manual 70 analysis (Jensen et al., 1990). Droogers and Allen (2002) and Hargreaves and Allen (2003) reported Equation 8.51 to estimate well over a wide range of latitudes and climates for periods of 5 days or longer without significant error, provided mean wind speeds averaged between 1 and 3 m s-1. Itenfisu et al. (2003) compared daily ETo by Equation 8.51 and that by the standardized ASCE PM method for 49 sites in 16 states in U.S. and found mean ratios of Hargreaves ETo to ASCE PM ETos to range from 1.43 to 0.79, with a mean of 1.06 and a standard deviation of 0.13. The Hargreaves equation tended to estimate higher than the ASCE PM method when mean daily ETo was low, and vice versa. One advantage of an equation such as the Equation 8.51 relative to more complex equations such as the Penman or PM equation, which is often overlooked, is the reduced data requirement and therefore reduced chance for data error. This is advantageous in regions where solar radiation, humidity, and wind data are lacking or are of low or questionable quality (Droogers and Allen, 2002). Generally, air temperature can be measured with less error, with less sophisticated equipment, and by less trained individuals than can the other three parameters required by combination equations. Equation 8.51 can be calibrated against the PM equation (Equation 8.3) when data are available to produce a regionally calibrated temperature equation. Examples of this type of calibration in Spain include Martinez-Cob and Tejero-Juste (2004), Vanderlinden et al. (2004), and Gavilan et al. (2005). An alternative to using Equation 8.51 when data are lacking is to employ the PM equation using estimates for missing variables. 8.4.6 Effect of Timestep Size on Calculations The Penman and Penman-Monteith equations can be applied to hourly and 24-h timesteps. The 24-h timesteps can use daily, weekly, 10-d, and monthly averages for weather data. Under many climatic conditions, calculating ETo or ETr using hourly timesteps and then summing over 24 hours provides estimates that closely equal ETo or ETr calculated using 24-h average data with 24-h calculation timesteps, especially when applying the standardized ASCE-EWRI PM method (Itenfisu et al., 2003; ASCE-EWRI, 2005). Generally, 24-h ETo and ETr have potential for higher accuracy when computed using hourly or shorter timesteps and then summed to 24-hour totals. Hourly calculation is better able to consider impacts of abrupt and gradual changes in weather parameters during the course of a day on ET (Irmak et al., 2005; Allen et al., 2006). Examples of this are high wind conditions during afternoons with low humidity, overpass of cloud fronts and rain events, wintertime, and nighttime calm. 8.4.7 Limited Data Availability Many historical weather data sets include only measurements of daily air temperature. When ETo estimates are desired, one of three approaches is recommended: 1. When approximate calculations of ETo are suitable, apply Equation 8.51. 2. When more accuracy is required, calibrate Equation 8.51 at a regional station or stations that have Rs and u2 data. If Td or other humidity data are not available, Td can be estimated from Tmin as described in Equation 8.16. The desired standard reference method can be used as the calibration basis (e.g., Equation 8.3) at

226

Chapter 8 Water Requirements

the regional station. The calibration can be of the form:

b=

ET ref ET equation

or b =

ET ref - a ET equation

(8.52)

where ETref is the reference ET against which to calibrate and ETequation is ET estimated by the equation being calibrated. Coefficients b and a are calibration factors that may vary by month. These coefficients can be determined by linear regression. Reference ET is then calculated at locations with limited data as: ETref = b ETequation or ETref = a + b ETequation (8.53) As an example of regional calibration, Allen et al. (1983) determined values for b by month for the state of Idaho when calibrating the FAO-Blaney-Criddle equation against ETr. The user is cautioned, however, that use of faulty regional weather data for calibration can create more error in the ET estimate than using an original equation. 3. The user may elect to use the Penman or PM equation, estimating Rs, Td, and/or u2. This is the approach recommended by FAO-56 for most situations. The benefit is that a physically correct equation for ETo is used, so that local or regional calibration should be unnecessary. The accuracy of the ETo estimate depends on how well missing data parameters are estimated. FAO-56 and ASCE-EWRI (2005) describe procedures for estimating Rs, Td and u2 , where Rs is estimated from Tmax, Tmin, and Ra, Td is estimated from Tmin, and u2 is estimated from regional averages or from some regional weather station. In some parts of the U.S., gridded estimates of weather data are available from weather forecast systems such as by NCEP (McQueen et al., 2004). Allen and Robison (2007) applied the ASCE PM method at 107 National Weather Service stations in Idaho where only daily maximum and minimum air temperature data were available for long historic periods. They estimated Rs using a procedure by Thornton and Running (1999) following regional calibration. 8.4.8 Weather Data Integrity ETo and ETr estimates are only as good as the weather data used in their estimation. Weather data should be screened to insure integrity and representativeness. This is especially important with electronically collected data, since human oversight and maintenance may be limited. Solar radiation can be screened by plotting measurements against a clear sky Rso envelope provided by Equation 8.28, or a more accurate and complicated method in ASCE-EWRI (2005) Appendix D, illustrated in Allen (1996a). Humidity data (Td, RH, ea) can be evaluated by examining daily maximum RH (RHmax) or by comparing Td with Tmin. Under reference conditions, RHmax generally exceeds 80% during early morning and Td approaches Tmin (Allen 1996a; ASCEEWRI, 2005). Weather data should be representative of the reference condition. Data collected at or near airports can be negatively influenced by the local aridity, especially in arid and semiarid climates. Data from dry or urban settings may cause overestimation of ETo or ETr due to air temperature measurements that are too high and humidity measurements that are too low, relative to the reference condition. Allen et al. (1998) and ASCEEWRI (2005) suggest simple adjustments for “nonreference” weather data to provide

Design and Operation of Farm Irrigation Systems

227

data more reflective of well-watered settings. Allen and Gichuki (1989) and Ley et al. (1996) have suggested more sophisticated approaches. Often, substituting Td = Tmin – Ko for measured Td, as suggested in Equation 8.16, can improve ETo estimates made with the combination equation when data are from a nonreference setting. Using the nonreference data in Equation 8.16 will overestimate the true Td and ea that would occur under reference conditions, since Tmin will be higher in the dry setting and, consequently, so will estimated Td. However, because es and ea in the nonreference setting are both inflated when calculated using Tmax, Tmin, and Td estimated from Equation 8.16, the es – ea difference in the combination equation is often brought more in line with that expected for the reference condition and a more accurate estimate for ETref results.

8.5 ESTIMATING ET FOR CROPS The crop coefficient, Kc, has been developed over the past half-century to simplify and standardize the calculation and estimation of crop water use. The Kc is defined as the ratio of ET from a specific surface to ETref. The specific surface can be comprised of bare soil, soil with partial vegetation cover, or full vegetation cover. The Kc represents an integration of effects of crop height, crop-soil resistance, and surface albedo that distinguish the surface from the ETref definition. The value for Kc often changes during the growing season as plants grow and develop, as the fraction of ground covered by vegetation changes, as the wetness of the underlying soil surface changes, and as plants age and mature. The potential crop ET is calculated by multiplying ETref by the crop coefficient: ETc = Kc ETref (8.54) The reference crop corresponds to a living, agricultural crop (clipped grass or fullcover alfalfa) and it incorporates the majority of the effects of variable weather into the ETref estimate. Because ETref represents an index of climatic evaporative demand, the Kc varies predominately with specific crop characteristics and a small amount with climate in the case of ETo. This enables the transfer of standard values for Kc between locations and between climates. The transfer has led to the widespread acceptance and usefulness of the Kc-ETref approach. Values for Kc based on ETo tend to average 1.2 to 1.4 times Kc based on ETr due to differences in magnitudes between the two ETref types. For this reason, ASCE-EWRI (2005) has recommended differentiation between the two families of Kc by referring to ETo-based Kc as Kco and to ETr-based Kc as Kcr. The Kc and ETc in Equation 8.54 represent ET under potential growing conditions with no stresses caused by shortage of soil water or salinity. These are the general conditions for agricultural production. Both water and salinity stress reduce transpiration and thus ET by causing plant canopies to reduce stomatal opening and water loss. When water or salinity-induced reductions are considered, ET is calculated as: ETc act = Kc act ETref (8.55) where ETc act and Kc act represent ET and associated Kc under actual field conditions that may include effects of environmental stresses. 8.5.1 Mean Kc and Dual Kc Methods Two approaches to Kc are described in this section. The first approach uses a mean or “single” Kc where time-averaged effects of evaporation from the soil surface are averaged into the Kc value. The mean Kc represents, on any particular day, average evaporation fluxes from the soil and plant surfaces. The second Kc approach is the

228

Chapter 8 Water Requirements

dual Kc method, where the Kc value is divided into a basal crop coefficient, Kcb, and a separate component, Ke, representing evaporation from the soil surface. The basal crop coefficient represents ET conditions when the soil surface is dry, but sufficient root zone moisture is present to support full transpiration. Generally, a daily calculation timestep is required to apply the dual Kc method, whereas the mean Kc method can be applied on daily, weekly, or monthly timesteps. The mean (single) Kc approach is used for planning studies and irrigation systems design where averaged effects of soil wetting are appropriate. For typical irrigation management, this use is valid. The dual Kc approach, which requires more numerical calculations, is suited for irrigation scheduling, for soil water balance computations, and for research studies where specific effects of day-to-day variation in soil surface wetness and the resulting impacts on daily ETc, the soil water profile, and deep percolation fluxes are important. The form of the equation for Kc act in the dual Kc approach is: (8.56) Kc act = Ks Kcb + Ke where Kact = actual crop coefficient that considers effects of moisture stressa stress reduction coefficient, 0 to 1 Ks = a stress reduction coefficient, 0 to 1 Kcb = basal crop coefficient, 0 to −1.4 for Kco and 0 to ~1.0 for Kcr Ke = soil water evaporation coefficient, 0 to −1.4 for Kco and 0 to ~1.0 for Kc r. All three terms are dimensionless. Kcb is defined as the ratio of ETc to ETref when the soil surface layer is dry, but where the average soil water content of the root zone is adequate to sustain full plant transpiration. Ks reduces the value of Kcb when the average soil water content of the root zone is not adequate to sustain full plant transpiration and is described later. Ke quantifies the evaporation component from wet soil in addition to the evapotranspiration represented in Kcb. The mean (single) Kc is a simplified, single coefficient that includes effects of timeaveraged Ke: K c = K cb + K e (8.57) and

K c act = K s K c

(8.58)

where Ke represents the time-averaged (i.e., multi-day) Ke. 8.5.2 The Crop Coefficient Curve The crop coefficient curve represents the changes in Kc over the course of the growing season, depending on changes in vegetation cover and maturation. During the initial period, shortly after planting of annuals or shortly after the initiation of new leaves for perennials, the value of Kc is small, often less than 0.4. Figure 8.1 illustrates two general shapes of Kc curves as used by FAO (Doorenbos and Pruitt, 1977; Allen et al., 1998) and as used by Wright (1982). The curves by Wright exhibit a smoothed change in Kc with time, whereas Kc curves by FAO are constructed using linear line segments. The simple, linear FAO Kc curve is widely used and generally provides sufficiently accurate description of linear crop growth and development and subsequently Kc, and is recommended for most applications. Definitions for three benchmark Kc values required to construct the curve and associated definitions for growth stage periods and relative ground cover are illustrated in Figure 8.2.

Design and Operation of Farm Irrigation Systems

229

1.4

K c Wright

1.2 K c FAO 1.0 0.8

K c 0.6 0.4 0.2 0.0

May

June

July

August

Figure 8.1. Typical crop coefficient curves from Wright (1982) and from FAO-24 and FAO-56.

K c mid

Kc 1.4 1.2 1 0.8 0.6 0.4

Kc ini

K c end

0.2 0

time (days) initial

crop development

mid-season

late season

Figure 8.2. FAO Kc curve with four crop stages and three Kc values relative to typical ground cover (from FAO-56).

Kc

Kc= K cb+ K e

1.4

Ke

1.2 1 0.8 0.6 0.4 0.2

Kcb

0

time (days) initial

crop development

mid-season

late season

Figure 8.3. Generalized FAO Kc curve definitions showing the basal Kcb, soil evaporation, Ke, and time-averaged mean Kc values.

230

Chapter 8 Water Requirements

Typical shapes for the Kcb, Ke, and mean Kc curves are shown in Figure 8.3. The Kcb curve represents the minimum Kc for conditions of adequate soil moisture and dry soil surface. The Ke spikes represent increased evaporation following wetting of the soil surface by precipitation or irrigation. The spikes reach a maximum of about 1.2 for Kco and about 1.0 for Kcr. When summed, the values for Kcb and for Ke represent the total crop coefficient, Kc. The mean Kc curve shown as the dashed line illustrates the effect of averaging Ke over time and is displayed as a smoothed curve. The mean Kc lies above the Kcb curve, with potentially large differences during the initial and development stages, depending on the frequency of soil wetting. During the midseason period when the crop is likely to be near full cover, the effects of soil wetness are less pronounced. 8.5.3 Construction of the FAO Kc Curve The linear FAO style Kc curve is constructed by the following steps. 1. Divide the growing period into four general growth stages describing crop phenology and development (Figure 8.2). The four stages are: (1) initial period from planting or green-up until about 10% ground cover; (2) crop development period; (3) mid-season period from effective full cover to the beginning of the late season period; and (4) late-season period from the beginning of senescence until harvest, crop death, or full senescence. 2. Specify the three Kc values corresponding to Kc ini, Kc mid, and Kc end, where Kc ini represents the average Kc during the initial period, Kc mid represents the average Kc during the midseason period, and Kc end represents the Kc at the end of the late-season period. 3. Connect straight line segments through each of the four growth stage periods, with horizontal lines drawn through Kc ini during the initial period and through Kc mid during the midseason period. Diagonal lines are drawn from Kc ini to Kc mid within the domain of the development period and from Kc mid to Kc end within the domain of the late-season period. In the case of basal Kcb, the same procedure for curve construction is followed. Kc mid represents the mean maximum Kc expected during the midseason period, and does not necessarily represent the absolute peak Kc reached by the crop. This is illustrated in Figure 8.1, where the smoothed Kc curve, in this case for dry, edible beans developed by Wright (1982; from Jensen et al. 1990), rises above the linearized curve during part of the midseason. As illustrated, the linearized FAO curve approximates the trends in Kc, and accurately estimates seasonal ET when the lengths of the growth stages and the value for Kc mid are properly selected. Values for the grass-based Kco ini, Kco mid, and Kco end, and for Kcbo ini, Kcbo mid, and Kcbo end are listed in Table 8.2 for various agricultural crops (the o subscript indicates an ETo basis). The three Kco columns represent typical irrigation management and precipitation frequencies. The Kc values are taken from FAO-56 (Allen et al., 1998) and are largely based on Doorenbos and Pruitt (1977) and Doorenbos and Kassam (1979). A primary departure made here is for trees and grapes where, rather than listing values for Kc and Kcb, values are instead given for Kcb full, Kc min and Kcb cover to be used in Equation 8.85 (with a density coefficient from Equation 8.80) introduced later. This modification provides flexibility in adjusting the value for Kc and Kcb according to the fraction of ground shaded by canopy, which is highly variable among orchards and cultures. Parameters are listed in footnotes for applying Equation 8.80 that reproduce

Design and Operation of Farm Irrigation Systems

231

various Kc values reported in the literature. The Kco in Table 8.2 are applicable with grass reference ETo as defined by the standardized FAO/ASCE PM Equation (8.3) and are generally valid for use with other grass reference equations, provided these agree relatively closely with the standardized PM definition (ASCE-EWRI, 2005). For several group types, only one value for Kc ini is listed for the group, since tabularized Kc ini are only approximate. Figures 8.5a-8.5c (in Section 8.5.3.4), from FAO-56, improve estimates for Kc ini by accounting for frequency of soil wetting and soil type. 8.5.3.1 The climatic basis of Table 8.2. The Kc mid values in Table 8.2 are typical values expected under a standard climatic condition defined in FAO-56 (Allen et al., 1998) as a subhumid climate having average daytime minimum relative humidity (RHmin) of 45 % and having calm to moderate wind speeds averaging 2 m s-1. More arid climates and conditions of greater wind speed have higher values for Kc and Kcb, especially for tall crops and more humid climates, and conditions of lower wind speed have lower values according to the relationship: ⎛h⎞ K c = K c table + [0.04 (u 2 − 2) − 0.004 (RH min − 45)] ⎜ ⎟ ⎝3⎠

0.3

(8.59)

where Kc table is the Kc value (or Kcb value) from Table 8.2 for Kc mid (and for Kc end when Kc end > 0.45) and h is mean crop height in m. The Kc for crops 2 to 3 m in height can increase by as much as 40% when going from a calm, humid climate (for example, u2 = 1 m s-1 and RHmin = 70%) to an extremely windy, arid climate (for example, u2 = 5 m s-1 and RHmin = 15%). The increase in Kc is due to the influence of the larger aerodynamic roughness of tall crops relative to grass on the transport of water vapor from the surface. The adjustments to Kc for climate are generally made using mean values for u2 and RHmin for the entire midseason period. 8.5.3.2 Lengths of crop growth stages. The four crop growth stages for the FAO Kco curves are characterized in terms of crop growth benchmarks as illustrated in Figure 8.2. The crop development period is presumed to begin when approximately 10% of the ground is covered by vegetation and ends at attainment of effective full cover. Effective cover can be defined for row crops such as beans, sugar beets, potatoes, and corn as the time when some leaves of plants in adjacent rows begin to intermingle so that soil shading becomes nearly complete, or when plants reach nearly full size, if no intermingling occurs. For crops taller than 0.5 m, the average fraction of the ground surface covered by vegetation at the time of effective full cover is about 0.7 to 0.8 (Neale et al., 1989; Grattan et al., 1998). Effective full cover for many crops begins at the initiation of flowering. It is understood that the crop or plant can continue to grow in both height and leaf area after the attainment of effective full cover. Another way to estimate the occurrence of full cover is when the leaf area index (LAI) reaches about 3 (Ritchie, 1972; Wright, 1982; Ritchie and NeSmith, 1991). The lengths of the initial and development periods are relatively short for deciduous trees and shrubs, depending on the amount of pruning, which develop new leaves in the spring at relatively fast rates. These two periods may thus be only a few days in length for trees and may include the flowering period. The Kc ini selected for trees and shrubs should reflect the ground condition prior to leaf emergence or initiation, since Kc ini is affected by the amount of grass or weed cover, soil wetness, tree density, and mulch density.

232

Chapter 8 Water Requirements

Table 8.2. Mean crop coefficients, Kco, and basal crop coefficients, Kcbo, for well-managed crops in a subhumid climate, for use with ETos (following FAO-56*). Kc mid Kc end Kcb ini Kcb mid Kcb end Crop Kc ini[1] a. Small Vegetables 0.7 1.05 0.95 0.15 0.95 0.85 Broccoli 1.05 0.95 0.95 0.85 Brussels sprouts 1.05 0.95 0.95 0.85 Cabbage 1.05 0.95 0.95 0.85 Carrots 1.05 0.95 0.95 0.85 Cauliflower 1.05 0.95 0.95 0.85 Celery 1.05 1.00 0.95 0.90 Garlic 1.00 0.70 0.90 0.60 Lettuce 1.00 0.95 0.90 0.90 Onions, dry 1.05 0.75 0.95 0.65 green 1.00 1.00 0.90 0.90 seed 1.05 0.80 1.05 0.70 Spinach 1.00 0.95 0.90 0.85 Radish 0.90 0.85 0.85 0.75 b. Vegetables, Solanum Family (Solanaceae) 0.6 1.15 0.80 0.15 1.10 0.70 Eggplant 1.05 0.90 1.00 0.80 Sweet peppers (bell) 1.05[2] 0.90 1.00[2] 0.80 1.10[2] 0.60-0.80 Tomato 1.15[2] 0.70-0.90 c. Vegetables, Cucumber Family (Cucurbitaceae) 0.5 1.00 0.80 0.15 0.95 0.70 Cantaloupe 0.5 0.85 0.60 0.75 0.50 0.75 0.95[2] 0.70 Cucumber, fresh market 0.6 1.00[2] machine harvest 0.5 1.00 0.90 0.95 0.80 Pumpkin, winter squash 1.00 0.80 0.95 0.70 Squash, zucchini 0.95 0.75 0.90 0.70 Sweet melons 1.05 0.75 1.00 0.70 Watermelon 0.4 1.00 0.75 0.95 0.70 d. Roots and Tubers 0.5 1.10 0.95 0.15 1.00 0.85 Beets, table 1.05 0.95 0.95 0.85 Cassava, year 1 0.3 0.80[3] 0.30 0.70[3] 0.20 year 2 0.3 1.10 0.50 1.00 0.45 Parsnip 0.5 1.05 0.95 0.95 0.85 1.10 0.65[4] Potato 1.15 0.75[4] Sweet potato 1.15 0.65 1.10 0.55 Turnip (and rutabaga) 1.10 0.95 1.00 0.85 Sugar beet 0.35 1.20 0.70[5] 1.15 0.50[5] * Primary sources of Table 8.2: FAO-56 (Allen et al., 1998), with Kc ini traceable to Doorenbos and Kassam (1979) and [1]

[2]

[3] [4] [5]

Kc mid and Kc end traceable to Doorenbos and Pruitt (1977), Pruitt (1986), Wright (1981, 1982), and Snyder et al. (1989a,b). These are general values for Kc ini under typical irrigation management and soil wetting. For frequent wettings such as with high frequency sprinkle irrigation or daily rainfall, these values may increase substantially and may approach 1.0 to 1.2. Kc ini is a function of wetting interval and potential evaporation rate during the initial and development periods and is more accurately estimated using Figure 8.5 or using the dual Kcb ini + Ke. Beans, peas, legumes, tomatoes, peppers and cucumbers are sometimes grown on stalks reaching 1.5 to 2 meters in height. In such cases, increased Kc values need to be taken. For green beans, peppers and cucumbers, 1.15 can be taken, and for tomatoes, dry beans and peas, 1.20. Under these conditions h should be increased also. The midseason values for cassava assume non-stressed conditions during or following the rainy season. The Kc end and Kcb end values account for dormancy during the dry season. The Kc end and Kcb end values for potatoes are about 0.40 and 0.35 for long season potatoes with vine kill. This Kc end and Kcb end values are for no irrigation during the last month of the growing season. The Kc end and Kcb end values for sugar beets are higher, up to 1.0 and 0.9, when irrigation or significant rain occurs during the last month.

(continued)

Design and Operation of Farm Irrigation Systems

233

Table 8.2 continued. Crop Kc ini[1] Kc mid Kc end Kcb ini Kcb mid Kcb end e. Legumes (Leguminosae) 0.4 1.15 0.55 0.15 1.10 0.50 Beans, green 0.5 1.05[2] 0.90 1.00[2] 0.80 0.35 1.10[2] 0.25 Beans, dry, and pulses 0.4 1.15[2] Chick pea 1.00 0.35 0.95 0.25 1.10 1.10[2] 1.05 Faba bean (broad bean), fresh 0.5 1.15[2] dry/seed 0.5 1.15[2] 0.30 1.10[2] 0.20 Garbanzo 0.4 1.15 0.35 1.05 0.25 Green gram and cowpeas 1.05 0.60-0.35[6] 1.00 0.55-0.25[6] Groundnut (peanut) 1.15 0.60 1.10 0.50 Lentil 1.10 0.30 1.05 0.20 Peas, fresh 0.5 1.15[2] 1.10 1.10[2] 1.05 dry/seed 1.15 0.30 1.10 0.20 Soybeans 1.15 0.50 1.10 0.30 f. Perennial Vegetables (with winter dormancy and initially bare or mulched) 0.5 1.00 0.80 Artichokes 0.5 1.00 0.95 0.15 0.95 0.90 Asparagus 0.5 0.95[7] 0.30 0.15 0.90[7] 0.20 Mint 0.60 1.15 1.10 0.40 1.10 1.05 Strawberries 0.40 0.85 0.75 0.30 0.80 0.70 g. Fiber Crops 0.35 0.15 Cotton 1.15-1.20 0.70-0.50 1.10-1.15 0.50-0.40 Flax 1.10 0.25 1.05 0.20 Sisal[8] 0.4-0.7 0.4-0.7 0.4-0.7 0.4-0.7 h. Oil Crops 0.35 1.15 0.35 0.15 1.10 0.25 Castor bean (Ricinus) 1.15 0.55 1.10 0.45 Rapeseed, canola 1.0-1.15[9] 0.35 0.95-1.10[9] 0.25 0.95-1.10[9] 0.20 Safflower 1.0-1.15[9] 0.25 Sesame 1.10 0.25 1.05 0.20 Sunflower 1.0-1.15[9] 0.35 0.95-1.10[9] 0.25 i. Cereals 0.3 1.15 0.4 0.15 1.10 0.25 Barley, oats 1.15 0.25 1.10 0.15 Spring wheat 1.15 0.25-0.4[10] 1.10 0.15-0.3[10] [10] [11] Winter wheat, frozen soils 0.4 1.15 0.25-0.4 0.15-0.5 1.10 0.15-0.3[10] with non-frozen soils 0.7 1.15 0.25-0.4[10] 1.20 0.60,0.35 0.15 1.15 0.50,0.15 Maize, field (grain corn) [12] Maize, sweet (sweet corn) [12] 1.15 1.05[13] 1.10 1.00[13] Millet 1.00 0.30 0.95 0.20 Sorghum, grain 1.00-1.10 0.55 0.95-1.05 0.35 sweet 1.20 1.05 1.15 1.00 Rice 1.05 1.20 0.90-0.60 1.00 1.15 0.70-0.45 [6] [7] [8] [9] [10] [11] [12]

[13]

The first Kc end is for harvested fresh. The second value is for harvested dry. The Kc for asparagus usually remains at Kc ini during harvest of the spears, due to sparse ground cover. The Kc mid value is for following regrowth of plant vegetation following termination of harvest of spears. Kc for sisal depends on the planting density and water management (e.g., intentional moisture stress). The lower values are for rainfed crops having less dense plant populations. The higher value is for hand-harvested crops. The two Kcb ini values for winter wheat are for less than 10% ground cover and for during the dormant, winter period, if the vegetation fully covers the ground, but conditions are nonfrozen. The Kc mid and Kcb mid values are for populations > 50,000 plants ha-1. For less dense populations or less uniform growth, Kc mid and Kcb mid can be reduced by 0.10 to 0.2. The first Kc end value is for harvest at high grain moisture. The second Kc end value is for harvest after complete field drying of the grain (to about 18% moisture, wet mass basis). If harvested fresh for human consumption. Use Kc end for field maize if the sweet maize is allowed to mature and dry in the field.

(continued)

234

Table 8.2 continued. Crop j. Forages Alfalfa hay, ave.cutting effects individual cutting periods for seed Bermuda hay, ave.cutting effects spring crop for seed Clover hay, berseem, averaged cutting effects individual cutting periods Ryegrass hay, ave. cuttings Sudan grass hay (annual), averaged cutting effects individual cutting periods Grazing pasture, rotated grazing extensive grazing Turf grass, cool season[16] warm season[16] k. Sugar Cane l. Tropical Fruits and Trees Banana, 1st year 2nd year Cacao Coffee, bare ground cover with weeds Date palms Palm trees Pineapple,[17] bare soil with grass cover Rubber trees Tea, non-shaded shaded[18] m. Berries and Hops Berries (bushes) Hops [13] [14] [15] [16]

[17]

[18]

Chapter 8 Water Requirements

Kc ini[1]

Kc mid

Kc end

Kcb ini

Kcb mid

Kcb end

0.40 0.40[14] 0.40 0.55 0.35

0.95[13] 1.20[14] 0.50 1.00[13] 0.90

0.90 1.15[14] 0.50 0.85 0.65

0.30[14] 0.30 0.50 0.15

1.15[14] 0.45 0.95[15] 0.85

1.10[14] 0.45 0.80 0.60

0.40 0.40[14] 0.95

0.90[13] 1.15[14] 1.05

0.85 1.10[14] 1.00

0.30[14] 0.85

1.10[14] 1.00[15]

1.05[14] 0.95

0.50 0.90[13] [14] 0.50 1.15[14] 0.40 0.85-1.05 0.30 0.75 0.90 0.90 0.85 0.90 0.40 1.25

0.85 1.10[14] 0.85 0.75 0.90 0.90 0.75

0.30[14] 0.30 0.30 0.80 0.75 0.15

1.10[14] 0.80-1.00 0.70 0.85 0.80 1.20

1.05[14] 0.80 0.70 0.85 0.80 0.70

0.50 1.00 1.00 0.90 1.05 0.90 0.95 0.50 0.50 0.95 0.95 1.10

1.10 1.20 1.05 0.95 1.10 0.95 1.00 0.30 0.50 1.00 1.00 1.15

1.00 1.10 1.05 0.95 1.10 0.95 1.00 0.30 0.50 1.00 1.00 1.15

0.15 0.60 0.90 0.80 1.00 0.80 0.85 0.15 0.30 0.85 0.90 1.00

1.05 1.10 1.00 0.90 1.05 0.85 0.90 0.25 0.45 0.90 0.95 1.10

0.90 1.05 1.00 0.90 1.05 0.85 0.90 0.25 0.45 0.90 0.90 1.05

0.30 0.30

1.05 1.05

0.50 0.85

0.20 0.15

1.00 1.00

0.40 0.80

This Kc mid for hay crops represents an averaged Kc for before and following cuttings. It is applied to the period following the first development period until the beginning of the last late season period of the growing season. These Kc coefficients for hay crops represent immediately following cutting; at full cover; and immediately before cutting, respectively. The growing season is described as a series of individual cutting periods (Figure 8.4). This Kcb mid for bermuda and ryegrass hay crops represents an averaged Kcb mid for before and following cuttings. It is applied to the period following the first development period until the beginning of the last late season period. Cool season grass varieties include dense stands of bluegrass, ryegrass, and fescue. Warm season varieties include bermuda and St. Augustine. Values given are for potential conditions representing a 0.06- to 0.08-m mowing height. Turf, especially warm season varieties, can be stressed at moderate levels and still maintain appearance (see Section 8.6 and Table 8.13). Generally a value for the stress coefficient Ks of 0.9 for cool-season and 0.7 for warm-season varieties can be employed where careful water management is practiced and rapid growth is not required. Incorporating these Ks values into an “actual Kc” will yield Kc act values of about 0.8 for cool-season and 0.65 for warm-season turf. The pineapple plant has very low transpiration because it closes its stomates during the day and opens them during the night. Therefore, the majority of ETc from pineapple is evaporation from the soil. The Kc mid < Kc ini since Kc mid occurs during full ground cover so that soil evaporation is less. Values assume that 50% of the ground surface is covered by black plastic mulch and that irrigation is by sprinkler. For drip irrigation beneath the plastic mulch, Kc values given can be reduced by 0.10. Includes the water requirements of the shade trees.

(continued)

Design and Operation of Farm Irrigation Systems

Table 8.2 continued. Crop n. Fruit Trees and Grapes Conifer trees[19] Kiwi

Kc ini[1]

Kc mid

1.00 0.40

1.00 1.05

235

Kc end 1.00 1.05

Kcb ini

Kcb mid

Kcb end

0.95 0.20

0.95 1.00

0.95 1.00

The following, shaded, section for trees and grapes gives values for Kcb full used in Eq. 8.85. Kcb full[20] Kcb full[20] Kcb full[20] Kc min[20] Kcb cover[20] Kcb cover[20] initial mid end initial mid, end Almonds, no ground cover[21] 0.20 1.00 0.70[22] 0.15 ground cover 0.20 1.00 0.70[22] 0.15 0.75 0.80 Apples, cherries, pears; 0.30 1.15 0.80[22] 0.15 0.40 0.80 killing frost[23] no killing frost[23] 0.30 1.15 0.80[22] 0.15 0.75 0.80 Apricots, peaches, pears, plums, pecans; killing frost[24] 0.30 1.20 0.80[22] 0.15 0.40 0.80 [25] no killing frost 0.30 1.20 0.80[22] 0.15 0.70 0.80 Avocado, no grnd cover[26] 0.30 1.00 0.90 0.15 ground cover 0.30 1.00 0.90 0.15 0.75 0.80 Citrus[27] 0.80 0.80 0.80 0.15 0.75 0.80 Mango, no ground cover[28] 0.25 0.85 0.70 0.15 Olives[29] 0.60 0.70 0.60 0.15 0.70 0.70 Pistachios 0.30 1.00 0.70 0.15 0.70 0.70 Walnut[30] 0.40 1.10 0.65 0.15 0.75 0.80 Grapes, table or raisin[31] 0.20 1.15 0.90 0.15 0.70 0.70 wine[32] 0.20 0.80 0.60 0.15 0.70 0.70 [19] [20]

[21] [22]

[23] [24] [25]

[26] [27] [28] [29]

[30] [31] [32]

Conifers exhibit substantial stomatal control due to soil water deficit. The Kc can easily reduce below the values presented, which represent well-watered conditions for large forests. The first three columns for the orchard crops in the following rows are values for Kcb full for initial, mid- and end-season periods to be used in Eq. 8.85 along with the Kc min of column 4 to calculate Kcb for the initial, midseason and lateseason periods, where the density coefficient, Kd, is calculated from Eq. 8.80 using effective fraction of cover fc eff and plant height, h, as noted in the following footnotes. The last two columns are Kcb cover for initial and for mid- and endseason periods to be used in Eq. 8.85 for when there is active ground cover. Generally the value for Kc ini is estimated as 0.10 + Kcb ini from Eq. 8.85 and Kc mid and Kc end are estimated as 0.05 + Kcb mid or Kcb end from Eq. 8.85. Apply Eq. 8.80 with fc eff = 0.4, ML = 1.5 and h = 4 m for Kd in Eq. 8.85 to derive Kcb values similar to FAO-56. These Kc end values represent Kc prior to leaf drop. After leaf drop, Kc end ≈ 0.20 for bare, dry soil or dead ground cover and Kc end ≈ 0.50 to 0.80 for actively growing ground cover. Apply Eq. 8.80 with fc eff = 0.5, ML = 2 and h = 3 m for Kd in Eq. 8.85 to derive Kcb values similar to FAO-56. Apply Eq. 8.80 with fc eff = 0.45, ML = 1.5 and h = 3 m for Kd in Eq. 8.85 to derive Kcb values similar to FAO-56. Apply Eq. 8.80 with fc eff = 0.8, ML = 1.5 and h = 3 m for Kd in Eq. 8.85 to derive Kc values similar to Johnson et al. (2005), with fc eff = 0.6, ML = 1.5, h = 3 m for Kc values similar to those derived from Girona et al. (2005), with fc eff = 0.45, ML = 1.5, h = 3 m for Kcb values similar to FAO-56, with fc eff = 0.29, ML = 1.5, h = 2.5 m for Kc values similar to Paço et al. (2006), and with any fc eff ≤ 0.7, ML = 1.5, h = 3 m to approximate Kc estimates by Ayars et al. (2003). Apply Eq. 8.80 with fc eff = 0.4, ML = 2 and h = 4 m for Kd in Eq. 8.85 to derive Kcb values similar to FAO-56. Apply Eq. 8.80 with fc eff = 0.2, 0.5 and 0.7, ML = 1.5 and h = 2, 2.5 and 3 m for Kd in Eq. 8.85 to derive recommended Kcb values that are about 15% higher than the values entered in FAO-56 for these same three levels of fc eff . Apply Eq. 8.80 with fc eff = 0.7 to 0.85, ML = 1.5 and h = 5 m for Kd in Eq. 8.85 for Kcb values similar to those derived from de Azevedo et al. (2003). Apply Eq. 8.80 with fc eff = 0.7, ML = 1.5 and h = 4 m for Kd in Eq. 8.85 to derive Kcb values similar to FAO-56. Apply with fc eff = 0.6, ML = 1.5, h = 4 m to derive Kc values similar to Pastor and Orgaz (1994), but with Kcb mid = 0.45. Apply with fc eff = 0.3 to 0.4, ML = 1.5, h = 4 m to derive Kcb values similar to Villalobos et al. (2000). Apply with fc eff = 0.05 and 0.25, ML = 1.5, h = 2 and 3 m to derive Kc values similar to Testi et al. (2004). Apply Eq. 8.80 with fc eff = 0.7, ML = 1.5 and h = 5 m for Kd in Eq. 8.85 to derive Kcb values similar to FAO-56. Apply Eq. 8.80 with fc eff = 0.65, ML = 1.5 and h = 2 m for Kd in Eq. 8.85 to derive Kc values similar to Johnson.et al. (2005). Apply with fc eff = 0.45, ML = 1.5, h = 2 m for Kcb values similar to FAO-56. Apply Eq. 8.80 with fc eff = 0.5, ML = 1.5 and h = 2 m for Kd in Eq. 8.85 to derive Kcb values similar to FAO-56.

(continued)

236

Table 8.2 continued. Crop Kc ini[1] o. Wetlands, Temperate Climate Cattails, bulrushes, 0.30 killing frost Cattails, bulrushes, no frost 0.60 Short veg., no frost 1.05 Reed swamp, standing water 1.00 Reed swamp, moist soil 0.90 p. Special Open water, < 2 m depth or in subhumid climates or tropics Open water, > 5 m depth, clear of turbidity, temperate climate [33]

Chapter 8 Water Requirements

Kc mid

Kc end

1.20 1.20 1.10 1.20 1.20

0.30 0.60 1.10 1.00 0.70

1.05

1.05

0.65[33]

1.25[33]

These Kc values are for deep water in temperate latitudes where large temperature changes in the water body occur during the year, and initial and peak period evaporation is low as radiation energy is absorbed into the deep water body. During fall and winter periods (Kc end), heat is released from the water body that increases the evaporation above that for grass. Therefore, Kc mid corresponds to the period when the water body is gaining thermal energy and Kc end when releasing thermal energy. These Kc values should be used with caution.

The start of maturity and beginning of decline in Kc is often signaled by the beginning of aging, yellowing or senescence of leaves, leaf drop, or the browning of fruit so that ETc is reduced relative to ETo. Calculations for Kc and ETc are sometimes presumed to end when the crop is harvested, dries out naturally, reaches full senescence, or experiences leaf drop. For some perennial vegetation in frost-free climates, crops may grow year round so that the date of termination is the same as the date of “planting.” The length of the late season period may be relatively short (less than 10 days) for vegetation killed by frost (for example, maize at high elevations in latitudes > 40°N) or for agricultural crops that are harvested fresh (for example, table beets and small vegetables). The value for Kc end should reflect the condition of the soil surface (average water content and any mulch cover) and the condition of the vegetation following harvest or after full senescence. Kc during nongrowing periods having little or no green ground cover can be estimated using the equation for Kc ini as described later. FAO-56 provides general lengths of growth (development) stages for a wide variety of crops under different climates and locations. This information is reproduced in Table 8.3. The lengths in Table 8.3 serve only to indicate typical proportions of growing season lengths under a variety of climates. In all applications, local observations of the specific plant stage development should be made to account for local effects of plant variety, climate, and cultural practices. Local information can be obtained by interviewing farmers, ranchers, agricultural extension agents, and local researchers, by conducting local surveys, or by remote sensing (Neale et al., 1989; Tasumi et al., 2005).

Design and Operation of Farm Irrigation Systems

237

Table 8.3. Lengths of crop development stages* for various planting periods and climatic regions (days) (from FAO-56). Dev. Mid Late Init. Region Crop (Lini) (Ldev) (Lmid) (Llate) Total Plant Date a. Small Vegetables Broccoli 35 45 40 15 135 Sept Calif. desert, U.S. Cabbage 40 60 50 15 165 Sept Calif. desert, U.S. 20 30 50/30 20 100 Oct/Jan arid climate Carrots 30 40 60 20 150 Feb/Mar Mediterranean 30 50 90 30 200 Oct Calif. desert, U.S. Cauliflower 35 50 40 15 140 Sept Calif. desert, U.S. 25 40 95 20 180 Oct (semi)arid 25 40 45 15 125 April Mediterranean Celery 30 55 105 20 210 Jan (semi)arid 20 30 20 10 80 April Mediterranean 25 35 25 10 95 February Mediterranean Crucifers[1] 30 35 90 40 195 Oct/Nov Mediterranean 20 30 15 10 75 April Mediterranean 30 40 25 10 105 Nov/Jan Mediterranean Lettuce 25 35 30 10 100 Oct/Nov arid region 35 50 45 10 140 Feb Mediterranean 15 25 70 40 150 April Mediterranean Onion, dry 20 35 110 45 210 Oct; Jan. arid region; Calif. 25 30 10 5 70 April/May Mediterranean 20 45 20 10 95 October arid region Onion, green 30 55 55 40 180 March Calif., U.S. Onion, seed 20 45 165 45 275 Sept Calif. desert, U.S. 20 20 15/25 5 60/70 Apr; Sep/Oct Mediterranean Spinach 20 30 40 10 100 November arid region 5 10 15 5 35 Mar/Apr Medit.; Europe Radish 10 10 15 5 40 Winter arid region b. Vegetables, Solanum Family (Solanaceae) 30 40 40 20 130 October arid region Egg plant 30 45 40 25 140 May/June Mediterranean Sweet peppers 25/30 35 40 20 125 April/June Europe and Medit. (bell) 30 40 110 30 210 October arid region 30 40 40 25 135 January arid region 35 40 50 30 155 Apr/May Calif., U.S. Tomato 25 40 60 30 155 Jan Calif. desert, U.S. 35 45 70 30 180 Oct/Nov arid region 30 40 45 30 145 April/May Mediterranean c. Vegetables, Cucumber Family (Cucurbitaceae) 30 45 35 10 120 Jan Calif., U.S. Cantaloupe 10 60 25 25 120 Aug Calif., U.S. 20 30 40 15 105 June/Aug arid region Cucumber 25 35 50 20 130 Nov; Feb arid region * Lengths of crop development stages provided in this table are indicative of general conditions, but may vary substantially from region to region, with climate and cropping conditions, and with crop variety. The user is strongly encouraged to obtain appropriate local information. [1] Crucifers include cabbage, cauliflower, broccoli, and Brussels sprouts.

(continued)

238

Table 8.3 continued. Dev. Init. Crop (Lini) (Ldev) Pumpkin, 20 30 winter squash 25 35 Squash, 25 35 zucchini 20 30 25 35 30 30 Sweet melons 15 40 30 45 20 30 Watermelons 10 20 d. Roots and Tubers 15 25 Beets, table 25 30 Cassava, yr 1 20 40 yr 2 150 40 25 30 25 30 Potato 30 35 45 30 30 35 20 30 Sweet potato 15 30 30 45 25 30 25 65 40 50 Sugar beet 25 35 45 75 35 60 e. Legumes (Leguminosae) 20 30 Beans, green 15 25 20 30 Beans, dry 15 25 25 25 Faba bean 15 25 Broad bean 20 30 Dry bean 90 45 Green bean 90 45 Green gram, 20 30 cowpeas 25 35 Groundnut 35 35 35 45 20 30 Lentil 25 35 15 25 Peas 20 30 35 25 (continued)

Chapter 8 Water Requirements

Mid (Lmid) 30 35 25 25 40 50 65 65 30 20

Late (Llate) 20 25 15 15 20 30 15 20 30 30

20 25 90 110 30/45 45 50 70 50 60 50 90 90 100 50 50 80 70

10 10 60 60 30 30 30 20 25 40 30 15 10 65 40 50 30 40

30 25 40 35 30 35 35 40 40

10 10 20 20 20 15 15 60 0

90 75 110 95 100 90 100 235 175

30

20

110

March

Mediterranean

45 35 35 60 70 35 35 30

25 35 25 40 40 15 15 20

130 140 140 150 170 90 100 110

Dry season May May/June April Oct/Nov May Mar/Apr April

West Africa high latitudes Mediterranean Europe arid region Europe Mediterranean Idaho, U.S.

Total 100 120 100 90 120 140 135 160 110 80

Plant Date Mar, Aug June Apr; Dec. May/June May March Aug Dec/Jan April Mat/Aug

Region Mediterranean Europe Medit.; arid reg. Medit.; Europe Mediterranean Calif., U.S. Calif. desert, U.S. arid region Italy Near East (desert)

70 90 210 360

Apr/May Mediterranean Feb/Mar Mediterranean, arid Rainy season tropical regions

115/130

Jan/Nov (semi)arid climate May continental climate April Europe Apr/May Idaho, U.S. Dec Calif. Desert, U.S. April Mediterranean Rainy seas. tropical regions March Calif., U.S. June Calif., U.S. Calif. desert, U.S. Sept April Idaho, U.S. Mediterranean May November Mediterranean November arid regions

130 145 165 140 150 125 180 155 255 180 160 230 205

Feb/Mar Calif.,Mediterranean Aug/Sep Calif.,Egypt,Lebanon May/June continental climate June Pakistan, Calif. June Idaho, U.S. May Europe Mar/Apr Mediterranean Nov Europe Nov Europe

Design and Operation of Farm Irrigation Systems

239

Table 8.3 continued. Init. Dev. Mid Late Region Crop (Lini) (Ldev) (Lmid) (Llate) Total Plant Date 15 15 40 15 85 Dec tropics 20 30/35 60 25 140 May central U.S. Soybeans 20 25 75 30 150 June Japan f. Perennial Vegetables (with winter dormancy and initially bare or mulched soil) 40 40 250 30 360 Apr (1st yr) California Artichoke 20 25 250 30 325 May (2nd yr) (cut in May) 50 30 100 50 230 Feb warm winter Asparagus 90 30 200 45 365 Feb Mediterranean g. Fiber Crops 30 50 60 55 195 Mar-May Egypt,Pakistan,Cal. 45 90 45 45 225 Mar Calif. desert,U.S. Cotton 30 50 60 55 195 Sept Yemen 30 50 55 45 180 April Texas 25 35 50 40 150 April Europe Flax 30 40 100 50 220 October Arizona h. Oil Crops 25 40 65 50 180 March (semi)arid climate Castorbeans 20 40 50 25 135 Nov. Indonesia 20 35 45 25 125 April California, U.S. 25 35 55 30 145 Mar high latitudes Safflower 35 55 60 40 190 Oct/Nov arid region Sesame 20 30 40 20 100 June China Sunflower 25 35 45 25 130 April/May Medit.; Calif. i. Cereals central India 120 November 15 25 50 30 March/Apr 35-45°L 20 25 60 30 135 Barley, July East Africa 15 30 65 40 150 oats, Apr 40 20 130 40 30 wheat Nov 40 60 60 40 200 Dec Calif. desert,U.S. 60 30 160 20 50 60[2] 20[2] 70 30 180 December California, U.S Winter 30 140 40 30 240 November Mediterranean wheat 160 75 75 25 335 October Idaho, U.S. Grains 20 30 60 40 150 April Mediterranean (small) 25 35 65 40 165 Oct/Nov Pakistan; arid reg. 180 April East Africa (alt.) 30 50 60 40 25 40 45 30 140 Dec/Jan arid climate Nigeria (humid) 20 35 40 30 125 June Maize 35 40 30 125 October India (dry, cool) 20 (grain) 30 40 50 30 150 April Spain(spr,sum);Cal. 30 40 50 50 170 April Idaho, U.S. [2]

These periods for winter wheat will lengthen in frozen climates according to days having zero growth potential and wheat dormancy. Under general conditions and in the absence of local data, fall planting of winter wheat can be presumed to occur in northern temperate climates when the 10-day running average of mean daily air temperature decreases to 17°C or December 1, whichever comes first. Planting of spring wheat can be presumed to occur when the 10-day running average of mean daily air temperature increases to 5°C. Spring planting of maize-grain can be presumed to occur when the 10-day running average of mean daily air temperature increases to 13°C.

(continued)

240

Chapter 8 Water Requirements

Table 8.3 continued. Init. Crop (Lini) 20 20 Maize 20 (sweet) 30 20 15 Millet 20 20 Sorghum 20 30 Rice 30 j. Forages Alfalfa, total 10 season[4] 10 Alfalfa, 1st 10 cutting cycle[4] Alfalfa, other 5 cutting 5 [4] cycles Bermuda 10 for seed Bermuda for hay 10 (several cuttings) Grass pasture[4]

Dev. (Ldev) 20 25 30 30 40 25 30 35 35 30 30

Mid (Lmid) 30 25 50/30 30 70 40 55 40 45 60 80

Late (Llate) Total 10 80 10 80 10 90 10[3] 110 10 140 25 105 35 140 30 130 30 140 30 150 40 180

30

var.

var.

var.

20 30

20 25

10 10

60 75

Jan Apr (last –4°C)

10 20

10 10

5 10

30 45

Mar Jun

Calif., U.S. Idaho, U.S.

25

35

35

105

March

Calif. desert, U.S.

15

75

35

135

–

Calif. desert, U.S.

Plant Date March May/June Oct/Dec April Jan June April May/June Mar/April Dec; May May

Region Philippines Mediterranean arid climate Idaho, U.S. Calif. desert, U.S. Pakistan central U.S. U.S., Pakis., Med. arid region tropics; Mediterr. tropics last –4°C in spring until first -4°C in fall Calif., U.S. Idaho, U.S.

7 days before last –4°C in spring until 7 days after first –4°C in fall

10

20

–

–

–

25

25

15

10

75

Apr

Calif. desert, U.S.

3

15

12

7

37

June

Calif. desert, U.S.

35 60 50 70 75 105 25 70 Sugar cane, 30 50 ratoon 35 105 l. Tropical Fruits and Trees Banana, 1st yr 120 90 Banana, 2nd yr 120 60 Pineapple 60 120

190 220 330 135 180 210

120 140 210 50 60 70

405 480 720 280 320 420

120 180 600

60 5 10

390 365 790

Sudan, 1st cutting cycle Sudan, other cutting cycles k. Sugar Cane Sugar cane, virgin

[3] [4]

low latitudes tropics Hawaii, U.S. low latitudes tropics Hawaii, U.S. Mar Feb

Mediterranean Mediterranean Hawaii, U.S.

The late season for sweet maize will be about 35 days if the grain is allowed to mature and dry. In climates having killing frosts, growing seasons can be estimated for alfalfa and grass as: alfalfa: last –4°C in spring until first –4°C in fall (Everson et al., 1978). grass: 7 days before last –4°C in spring and 7 days after last –4°C in fall (Kruse and Haise, 1974).

(continued)

Design and Operation of Farm Irrigation Systems

Table 8.3 continued. Init. Dev. Crop (Lini) (Ldev) m. Grapes and Berries 20 40 20 50 Grapes 20 50 30 60 Hops 25 40 n. Fruit Trees Citrus 60 90 Deciduous 20 70 orchard, 20 70 little pruning 30 50 Olives 30 90 Pistachios 20 60 Walnuts 20 10 o. Wetlands, Temperate Climate 10 30 Wetlands (cattails, 180 60 bulrush) Wetlands 180 60 (short veg.) [5]

Mid (Lmid)

Late (Llate)

241

Total

Plant Date

Region

120 75 90 40 80

60 60 20 80 10

240 205 180 210 155

April Mar May April April

low latitudes Calif., U.S. high latitudes mid-latitudes (wine) Idaho, U.S.

120 90 120 130 60 30 130

95 30 60 30 90 40 30

365 210 270 240 270[5] 150 190

Jan March March March March Feb April

Mediterranean high latitudes low latitudes Calif., U.S. Mediterranean Mediterranean Utah, U.S.

80 90

20 35

140 365

May Utah, U.S.(killing frost) November Florida, U.S.

90

35

365

November

frost-free climate

Olive trees gain new leaves in March and often have transpiration during winter, where the Kc continues outside of the “growing period” and total season length may be set to 365 days.

8.5.3.3 Kc curves for forage crops. Many crops grown for forage or hay receive multiple harvests during the growing season. Each harvest essentially terminates a sub-growing season and associated Kc curve and initiates a new sub-growing season and associated Kc curve. The resulting Kc curve for the entire growing season is the aggregation of a series of Kc curves associated with each sub-cycle. A Kc curve constructed for alfalfa grown for hay in southern Idaho is illustrated in Figure 8.4. Cuttings may create a ground surface with less than 10% vegetation cover. Cutting cycle 1 may have longer duration than cycles 2, 3, and 4 if low air and soil temperatures or short daylength during this period moderate the crop growth rate. Frosts terminate the growing season in southern Idaho sometime in the fall, usually in early to mid-October (day of year 280 to 300, see footnote 4 of Table 8.3). Magnitudes of Kc values during midseason periods for each cutting cycle change from cycle to cycle as a result of adjusting Kc mid and Kc end for each period using Equation 8.59. Basal Kcb curves for forage or hay crops can be constructed similar to Figure 8.4. 8.5.3.4 Mean Kc for the initial stage (annual crops). ET during the initial stage for annual crops is predominantly in the form of evaporation from soil. Accurate estimates for the time-averaged Kc ini for the mean Kc must consider the frequency that the soil surface is wetted. Kc mid and Kc end are less affected by wetting frequency since vegetation during these periods is generally near full ground cover so that effects of surface evaporation are smaller. Figures 8.5a-8.5c from FAO-56 estimate Kc ini as a function of ETo, soil type, and wetting frequency. Equations for these curves are given in Allen et al., (1998, 2005b). Figure 8.5a is used for all soil types when wetting events (precipitation and irrigation) are light (i.e., infiltrated depths average about 10 mm per wetting event), Figure 8.5b is used for “heavy” wetting events, where infiltrated depths are greater than

242

Chapter 8 Water Requirements

1.5

Alfalfa Hay

Kimberly, Idaho

First Cutting Cycle

1.0

Kc

Second Cutting Cycle

0.5 L ini L dev

0.0

L mid

L late

Third Cutting Cycle

Fourth Cutting Cycle

Frost

Cuttings

90 120 150 180 210 240 270 300 Day of the Year

Figure 8.4. Crop coefficient curve for alfalfa hay crop in southern Idaho having four cuttings.

30 to 40 mm, on coarse-textured soils3 and Figure 8.5c is used for heavy wetting events on fine and medium-textured soils. In general, the mean time interval is estimated by counting all rainfall and irrigation events occurring during the initial period that are greater than a few mm. Wetting events occurring on adjacent days are typically counted as one event. When average infiltration depths are between 10 and 40 mm, the value for Kc ini can be interpolated between Figure 8.5a and Figure 8.5b or Figure 8.5c. 8.5.4 The Dual Kc Method: Incorporating Specific Wet Soil Effects The dual Kc method introduced in this section, based on FAO-56, is applicable to both Kc based on ETo (Kco) and Kc based on ETr (Kcr), with differences only in the value for Kc max. The Ke component describes the evaporation component of ETc. Because the dual Kc method incorporates the effects of specific wetting patterns and frequencies that may be unique to a single field, this method can provide more accurate estimates of evaporation components and total ET on an individual field basis. When the soil surface layer is wet, following rain or irrigation, Ke is at a maximum, and when the soil surface layer is dry, Ke is small, even zero. When the soil is wet, evaporation occurs at some maximum rate and Kcb + Ke is limited by a maximum value Kc max: K e = K r ( K c max − K cb ) ≤ f ew K c max (8.60) where Kc max is the maximum value of Kc following rain or irrigation, Kr is a dimensionless evaporation reduction coefficient (defined later) and is dependent on the cumulative depth of water depleted (evaporated), and few is the fraction of the soil that is both exposed to solar radiation and that is wetted. The evaporation rate is restricted by the estimated amount of energy available at the exposed soil fraction, i.e., Ke cannot exceed few Kc max. Kc max for the ETo basis (Kc max o) ranges from about 1.05 to 1.30: 0.3 ⎛⎧ ⎪ ⎛ h ⎞ ⎫⎪ K c max o = max⎜ ⎨1.2 + [0.04(u2 − 2) − 0.004 ( RH min − 45) ]⎜ ⎟ ⎬, K cb o + 0.05 ⎜ ⎝ 3 ⎠ ⎪⎭ ⎝ ⎪⎩

{

3

⎞

}⎟⎟

(8.61a)

⎠

Coarse-textured soils include sands and loamy sand textured soils. Medium-textured soils include sandy loam, loam, silt loam, and silt textured soils. Fine-textured soils include silty clay loam, silty clay, and clay textured soils.

Design and Operation of Farm Irrigation Systems

243

Table 8.4. Typical crop height, ranges of maximum effective rooting depth (Zr), and soil water depletion fraction for no stress (p), for common crops (from FAO-56). Depletion Maximum Crop Maximum Root Fraction[2] (p) for Crop Height (h) (m) Depth[1] (m) ETc = 5 mm d-1 a. Small Vegetables Broccoli 0.3 0.4-0.6 0.45 Brussels sprouts 0.4 0.4-0.6 0.45 Cabbage 0.4 0.5-0.8 0.45 Carrots 0.3 0.5-1.0 0.35 Cauliflower 0.4 0.4-0.7 0.45 Celery 0.6 0.3-0.5 0.20 Garlic 0.3 0.3-0.5 0.30 Lettuce 0.3 0.3-0.5 0.30 Onions, dry 0.4 0.3-0.6 0.30 green 0.3 0.3-0.6 0.30 seed 0.5 0.3-0.6 0.35 Spinach 0.3 0.3-0.5 0.20 Radishes 0.3 0.3-0.5 0.30 b. Vegetables, Solanum Family (Solanaceae) Eggplant 0.8 0.7-1.2 0.45 Sweet peppers (bell) 0.7 0.5-1.0 0.30 Tomato 0.6 0.7-1.5 0.40 c. Vegetables, Cucumber Family (Cucurbitaceae) Cantaloupe 0.3 0.9-1.5 0.45 Cucumber, fresh market 0.3 0.7-1.2 0.50 machine harvest 0.3 0.7-1.2 0.50 Pumpkin, winter squash 0.4 1.0-1.5 0.35 Squash, zucchini 0.3 0.6-1.0 0.50 Sweet melons 0.4 0.8-1.5 0.40 Watermelon 0.4 0.8-1.5 0.40 d. Roots and Tubers Beets, table 0.4 0.6-1.0 0.50 Cassava, year 1 1.0 0.5-0.8 0.35 year 2 1.5 0.7-1.0 0.40 Parsnip 0.4 0.5-1.0 0.40 Potato 0.6 0.4-0.6 0.35 Sweet potato 0.4 1.0-1.5 0.65 Turnip and rutabaga 0.6 0.5-1.0 0.50 Sugar beet 0.5 0.7-1.2 0.55[3] [1]

[2]

[3]

The larger values for Zr are for soils having no significant layering or other characteristics that can restrict rooting depth. The smaller values for Zr may be used for irrigation scheduling and the larger values for modeling soil water stress or for rainfed conditions. The tabled values for p apply for ETc ≈ 5 mm/day. The value for p can be adjusted for different ETc according to p = pTable 8.4 + 0.04 (5 – ETc) where p is expressed as a fraction and ETc as mm/day. Sugar beets often experience late afternoon wilting in arid climates even at p < 0.55, with usually only minor impact on sugar yield.

(continued)

244

Chapter 8 Water Requirements

Table 8.4 continued. Maximum Crop Height (h) (m)

Depletion Maximum Root Fraction[2] (p) Depth[1] (m) for ETc = 5 mm d-1

Crop e. Legumes (Leguminosae) Beans, green 0.4 0.5-0.7 0.45 Beans, dry and pulses 0.4 0.6-0.9 0.45 Beans, lima, large vines 0.4 0.8-1.2 0.45 Chick pea 0.4 0.6-1.0 0.50 Faba bean (broad bean), fresh 0.8 0.5-0.7 0.45 dry/seed 0.8 0.5-0.7 0.45 Garbanzo 0.8 0.6-1.0 0.45 Green gram and cowpeas 0.4 0.6-1.0 0.45 Groundnut (peanut) 0.4 0.5-1.0 0.50 Lentil 0.5 0.6-0.8 0.50 Peas, fresh 0.5 0.6-1.0 0.35 dry/seed 0.5 0.6-1.0 0.40 Soybeans 0.5-1.0 0.6-1.3 0.50 f. Perennial Vegetables (with winter dormancy and initially bare or mulched soil) Artichokes 0.7 0.6-0.9 0.45 Asparagus 0.2-0.8 1.2-1.8 0.45 Mint 0.6-0.8 0.4-0.8 0.40 Strawberries 0.2 0.2-0.3 0.20 g. Fiber Crops Cotton 1.2-1.5 1.0-1.7 0.65 Flax 1.2 1.0-1.5 0.50 Sisal 1.5 0.5-1.0 0.80 h. Oil Crops Castor bean (Ricinus) 0.3 1.0-2.0 0.50 Rapeseed, canola 0.6 1.0-1.5 0.60 Safflower 0.8 1.0-2.0 0.60 Sesame 1.0 1.0-1.5 0.60 Sunflower 2.0 0.8-1.5 0.45 i. Cereals Barley 1 1.0-1.5 0.55 Oats 1 1.0-1.5 0.55 Spring wheat 1 1.0-1.5 0.55 Winter wheat 1 1.5-1.8 0.55 Maize, field (grain) (field corn) 2 1.0-1.7 0.55 Maize, sweet (sweet corn) 1.5 0.8-1.2 0.50 Millet 1.5 1.0-2.0 0.55 Sorghum, grain 1-2 1.0-2.0 0.55 sweet 2-4 1.0-2.0 0.50 Rice 1 0.5-1.0 0.20[4] [4]

The value for p for rice is 0.20 of saturation.

(continued)

Design and Operation of Farm Irrigation Systems

245

Table 8.4 continued. Crop j. Forages Alfalfa, for hay for seed Bermuda, for hay spring crop for seed Clover hay, berseem Ryegrass hay Sudan grass hay (annual) Grazing pasture - rotated grazing extensive grazing Turf grass, cool season[5] warm season[5] k. Sugar Cane l. Tropical Fruits and Trees Banana, 1st year 2nd year Cacao Coffee Date palms Palm trees Pineapple Rubber trees Tea, non-shaded shaded m. Grapes and Berries Berries (bushes) Grapes, table or raisin wine Hops n. Fruit Trees Almonds Apples, cherries, pears Apricots, peaches, stone fruit Avocado Citrus 70% canopy 50% canopy 20% canopy Conifer trees Kiwi Olives (40% to 60% ground coverage by canopy) Pistachios Walnut orchard [5]

Maximum Crop Height (h) (m)

Depletion Maximum Root Fraction[2] (p) Depth[1] (m) for ETc = 5 mm d-1

0.7 0.7 0.35 0.4 0.6 0.3 1.2 0.15-0.30 0.10 0.10 0.10 3

1.0-2.0 1.0-3.0 1.0-1.5 1.0-1.5 0.6-0.9 0.6-1.0 1.0-1.5 0.5-1.5 0.5-1.5 0.5-1.0 0.5-1.0 1.2-2.0

0.55 0.60 0.55 0.60 0.50 0.60 0.55 0.60 0.60 0.40 0.50 0.65

3 4 3 2-3 8 8 0.6-1.2 10 1.5 2

0.5-0.9 0.5-0.9 0.7-1.0 0.9-1.5 1.5-2.5 0.7-1.1 0.3-0.6 1.0-1.5 0.9-1.5 0.9-1.5

0.35 0.35 0.30 0.40 0.50 0.65 0.50 0.40 0.40 0.45

1.5 2 1.5-2 5

0.6-1.2 1.0-2.0 1.0-2.0 1.0-1.2

0.50 0.35 0.45 0.50

5 4 3 3

1.0-2.0 1.0-2.0 1.0-2.0 0.5-1.0

0.40 0.50 0.50 0.70

4 3 2 10 3 3-5

1.2-1.5 1.1-1.5 0.8-1.1 1.0-1.5 0.7-1.3 1.2-1.7

0.50 0.50 0.50 0.70 0.35 0.65

1.0-1.5 1.7-2.4

0.40 0.50

3-5 4-5

Cool season grass varieties include bluegrass, ryegrass, and fescue. Warm season varieties include bermuda grass, buffalo grass, and St. Augustine grass. Grasses are variable in rooting depth. Some root below 1.2 m while others have shallow rooting depths. The deeper rooting depths for grasses represent conditions where careful water management is practiced with higher depletion between irrigations to encourage the deeper root exploration.

246

Chapter 8 Water Requirements 1.2

small infiltration depths ~ 10 mm

1.0

1 da y

K c ini

0.8 0.6

2 day

0.4 10 day 20 da y

0.2

(a)

0.0

0

1

2

3

4

4 day

7 day

5

6

7

8

9

10 11 12

ETo mm/day low

moderate

high

very high

1.2

large infiltration depths > 40 mm

1.0

1 day

K c ini

0.8 2 day

0.6 4 day

0.4 10 da y

0.2

(b)

0.0

0

1

2

3

7 day

20 day

a. coarse textures

4

5

6

7

8

9

10 11 12

ETo mm/day low

moderate

high

very high

1.2

1 day

large infiltration depths > 40 mm

K c ini

1.0 0.8

2 day

0.6

4 day

7 day 10 day

0.4 0.2

20 day b. fine and medium textures

(c)

0.0

0

1

2

3

4

5

6

7

8

9

10 11 12

ETo mm/day low

moderate

high

very high

Figure 8.5. Average Kc ini for the initial crop development stage as related to the level of ETo and the interval between irrigations and/or significant rain during the initial period for (a) all soil types when wetting events are light (about 10 mm per event); (b) coarsetextured soils when wetting events are greater than about 40 mm; and (c) medium and fine-textured soils when wetting events are greater than about 40 mm (from FAO-56); Equations for these curves are given in Allen et al. (2005b).

Design and Operation of Farm Irrigation Systems

247

Table 8.5. Typical soil water characteristics for different soil types (from FAO-56). Depth of Water That Can Be Depleted by Evaporation Soil Type (USDA Soil Stages 1 and 2 Stages 1 and 2 Soil Water Characteristics Texture TEW[1] Stage 1 TEW[1] ClassificaθFC θWP (θFC – θWP) REW (Ze = 0.10 m) (Ze = 0.15 m) tion) mm mm mm m3 m-3 m3 m-3 m3 m-3 Sand 0.07-0.17 0.02-0.07 0.05-0.11 2-7 612 9-13 Loamy sand 0.11-0.19 0.03-0.10 0.06-0.12 4-8 9-14 13-21 Sandy loam 0.18-0.28 0.06-0.16 0.11-0.15 6-10 15-20 22-30 Loam 0.20-0.30 0.07-0.17 0.13-0.18 8-10 16-22 24-33 Silt loam 0.22-0.36 0.09-0.21 0.13-0.19 8-11 18-25 27-37 Silt 0.28-0.36 0.12-0.22 0.16-0.20 8-11 22-26 33-39 Silt clay loam 0.30-0.37 0.17-0.24 0.13-0.18 8-11 22-27 33-40 Silty clay 0.30-0.42 0.17-0.29 0.13-0.19 8-12 22-28 33-42 33-43 Clay 0.32-0.40 0.20-0.24 0.12-0.20 8-12 22-29 [1]

TEW = (θFC – 0.5 θWP) Ze

where h is the mean plant height (in m) during the period of calculation (initial, development, midseason, or late-season), and the max ( ) function indicates the selection of the maximum of values separated by the comma. Kc max for the tall reference ETr (Kc max r) does not require adjustment for climate, due to the greater roughness of the reference basis: (8.61b) Kc max r = max (1.0, {Kcbr + 0.05}) Kcbo and Kcbr are the basal Kc for the ETo and ETr bases. Equation 8.61 ensures that Kc max is always greater than or equal to the sum Kcb + 0.05, suggesting that wet soil increases the Kc value above Kcb by 0.05 following complete wetting of the soil surface, even during periods of full ground cover. Equation 8.56 and Equations 8.60 to 8.69 can be applied with both the straight-line Kcb curve style of FAO and with curvilinear Kcb curves such as by Wright (1982), as illustrated later in this section. The surface soil layer is presumed to dry to an air-dry water content approximated as halfway between wilting point, θ WP, and oven dry. The amount of water that can be removed by evaporation during a complete drying cycle is estimated as: TEW = 1000 ( θ FC − 0.5 θWP ) Z e (8.62) where TEW (total evaporable water) is the maximum depth of water that can be evaporated from the surface soil layer when the layer has been initially completely wetted (in mm). Field capacity, θ FC, and θ WP are expressed in m3 m-3 and Ze is the effective depth (m) of the surface soil subject to drying to 0.5 θ WP by way of evaporation. Typical values for θ FC, θ WP, and TEW are given in Table 8.5 for a range in soil types. Ze is an empirical value based on observation. Some evaporation or soil drying will be observed to occur below the Ze depth. FAO-56 recommended values for Ze of 0.10 to 0.15 m, with 0.1 m recommended for coarse soils and 0.15 m recommended for finetextured soils. The Kr coefficient of Equation (8.60) is calculated, assuming a 2-stage drying process: (8.63a) K r = 1.0 for De, j -1 ≤ REW Kr =

TEW − De, j -1 for De, j -1 > REW TEW − REW

(8.63b)

248

Chapter 8 Water Requirements

where De,j-1 is cumulative depletion from the soil surface layer at the end of day j-1 (the previous day) (in mm), REW is the readily evaporable water that is evaporated during stage 1, and TEW and REW are in mm (REW < TEW). Evaporation from the soil beneath the crop canopy, occurring at a slower rate, is assumed included in the basal Kcb coefficient. In the FAO-56 model, the term fw is defined as the fraction of the surface wetted by irrigation and/or precipitation. This term defines the potential spatial extent of evaporation. Common values for fw are listed in Table 8.6 and illustrated in Figure 8.6. When the soil surface is completely wetted, as by precipitation or sprinkler, few of Equation 8.60 is set equal to (1 – fc), where fc is the fraction of soil surface effectively covered by vegetation. For irrigation systems where only a fraction of the ground surface is wetted, few is limited to fw:

f ew = min(1 − f c , f w )

(8.64)

Both 1 – fc and fw, for numerical stability, have limits of 0.01 to 1. In the case of drip irrigation, Allen et al. (1998) suggest that where the majority of soil wetted by irrigation is beneath the crop canopy and is shaded, fw be reduced to about one-half to one-third of that given in Table 8.6. Their general recommendation for drip irrigation is to multiply fw by [1 – (2/3) fc]. Pruitt et al. (1984) and Bonachela et al. (2001) have described evaporation patterns and extent under drip irrigation. The value for fc is limited to < 0.99 for numerical stability and is generally determined by visual observation. For estimating few, fc can be estimated from Kcb as: ⎛ K cb − K c min fc = ⎜ ⎜ K c max − K c min ⎝

⎞ ⎟ ⎟ ⎠

(1 + 0.5 h )

(8.65)

where fc is limited to 0 to 0.99 and Kc min is the minimum Kc for dry bare soil with no ground cover and h is crop height in meters. The difference Kcb – Kc min is limited to > 0.01 for numerical stability. The value for fc will change daily as Kcb changes. Kc min ordinarily has the same value as Kcb ini used for annual crops under nearly bare soil conditions (i.e., Kc min ~ 0.15). However, Kc min is set to 0 or nearly zero under conditions with large time periods between wetting events, for example in applications with natural vegetation in deserts. The value for fc decreases during the late-season period in proportion to Kcb to account for local transport of sensible heat from senescing leaves to the soil surface. Table 8.6. Common values for the fraction of soil surface wetted by irrigation or precipitation (after FAO-56). Wetting Event fw Precipitation 1.0 1.0 Sprinkler irrigation, field crops Sprinkler irrigation, orchards 0.7 to 1.0 Basin irrigation 1.0 Border irrigation 1.0 Furrow irrigation (every furrow), narrow bed 0.6 to 1.0 Furrow irrigation (every furrow), wide bed 0.4 to 0.6 Furrow irrigation (alternated furrows) 0.3 to 0.5 Microspray irrigation, orchards 0.5 to 0.8 Trickle (drip) irrigation 0.3 to 0.4

Design and Operation of Farm Irrigation Systems

249

initial crop development

Rain Basin Border Sprinkler Irrigation

1 - fc

fc

mid and late season

1 - fc

fc

fw = 1

fw = 1

f

fc ew

1 - fc

= (1 - f c1)- fc

fc

Furrow Irrigation fw

fw

few = f w 1 - fc

Drip Irrigation

fc

fw

few = (1 - f c) 1 - fc

fc

fw

few = 1.0 ... 0.3 f w

Figure 8.6. Crop Determination of few (grayed areas) as a function of the fraction of ground surface coverage (fc) and the fraction of the surface wetted (fw) (from FAO-56).

8.5.4.1 Water balance of the soil surface layer. Estimation of Ke requires a daily water balance for the few fraction of the surface soil layer. The daily soil water balance equation is: Ij Ej De, j = De, j -1 − ( Pj − RO j ) − + + Te, j + DPe, j (8.66) fw f ew

where De,j-1 and De,j = cumulative depletion depth at the ends of days j–1 and j, mm Pj = precipitation on day j, mm ROj = precipitation runoff from the soil surface on day j, mm Ij = irrigation depth on day j that infiltrates the soil, mm Ej = evaporation on day j (i.e., Ej = Ke ETref), mm Te, j = depth of transpiration from the exposed and wetted fraction of the soil surface layer on day j, mm DPe j = deep percolation from the few fraction of the soil surface layer on day j if soil water content exceeds field capacity, mm. Assuming that the surface layer is at field capacity following heavy rain or irrigation,

250

Chapter 8 Water Requirements

the minimum value for De,j is zero. The limits imposed on De,j are consequently 0 < De,j < TEW. It is recognized that water content of the soil surface layer can exceed TEW for short periods of time while drainage is occurring. However, because the length of time that this occurs varies with soil texture, wetting depth, and tillage, De,j > 0 is assumed. Additionally, it is recognized that some drainage in soil occurs at very small rates at water contents below field capacity. To some extent, impacts of these simple assumptions can be compensated for, if needed, in setting the value for Ze or TEW. The irrigation depth Ij is divided by fw to approximate the infiltration depth to the fw portion of the soil surface. Similarly, Ej is divided by few because it is assumed that all Ej (other than residual evaporation implicit to the Kcb coefficient) is taken from the few fraction of the surface layer. 8.5.4.2 Transpiration from the surface layer. The amount of transpiration extracted from the few fraction of the evaporating soil layer is generally a small fraction of total transpiration. However, for shallow-rooted annual crops where the depth of the maximum rooting is less than about 0.5 m, Te may have significant effect on the water balance of the surface layer and therefore on estimation of the evaporation component, especially for the period midway through the development period. The following extension to FAO-56 by Allen et al. (2005a) estimates Te from the few fraction of the evaporation layer in proportion to the water content of that layer: Te = K t K s K cb ETref

(8.67)

where Kt, having a range of 0 to 1, is the proportion of basal ET (= Kcb ETref) extracted as transpiration from the few fraction of the surface soil layer. Ks is the soil water stress factor computed for the root zone (range of 0 to 1). Kt is determined by comparing relative water availability in the Ze and Zr layers along with the presumed rooting distribution. For the few fraction: D ⎞ ⎛ ⎜ 1 − e ⎟ ⎛ Z ⎞ 0.6 TEW ⎟ ⎜ e ⎟ (8.68) Kt = ⎜ ⎜ 1 − Dr ⎟ ⎜⎝ Z r ⎟⎠ ⎜ ⎟ TAW ⎠ ⎝ where the numerator and denominator of the first expression are limited to > 0.001 and the value for Kt is limited to < 1.0. TAW is the total available water (mm) in the root zone and Dr is the current depletion from the effective root zone. In the simple water balance procedure of FAO-56, it is assumed that the soil water content is limited to < θ FC on the day of a complete wetting event. This is a reasonable assumption considering the shallowness of the surface layer. Downward drainage (percolation) of water from the surface layer is calculated as: Ij DPe, j = ( Pj − RO j ) + (8.69) − De, j -1 ≥0 fw 8.5.4.3 Initialization of the water balance and order of calculation. To initiate the water balance for the evaporating layer, the user can assume that the soil surface layer is near θ FC following a heavy rain or irrigation so that De,j-1 = 0. Where a long period of time has elapsed since the last wetting, the user can assume that all evaporable water has been depleted from the evaporation layer at the beginning of calculations so that De, j-1 = TEW = 1000 (θ FC – 0.5 θWP) Ze. Calculations for the dual Kcb +

Design and Operation of Farm Irrigation Systems

251

Ke procedure, for example when using a spreadsheet, proceed in the following order: Kcb, h, Kc max, fc, fw, few, Kr, Ke, E, DPe, De, I, Kc, and ETc. 8.5.4.4 Impacts of water stress. The Ks component in Equation 8.56 and 8.58 is the water stress coefficient used to reduce Kc or Kcb under conditions of water stress or salinity stress. FAO-56 describes the application of the salinity stress function for reducing ET. The water stress function is described here. Mean water content of the root zone is expressed as root zone depletion, Dr, defined as water shortage relative to field capacity. At field capacity, Dr = 0. The degree of stress is presumed to progressively increase as Dr increases past RAW, the depth of readily available water in the root zone. For Dr > RAW, Ks is: TAW − Dr TAW − Dr (8.70) Ks = = TAW − RAW (1 − p ) TAW where TAW is total available soil water in the root zone (mm), and p is the fraction of TAW that a crop can extract from the root zone without suffering water stress. When Dr < RAW, Ks = 1. Values for p are listed in Table 8.4. The total available water in the root zone is estimated as the difference between the water content at field capacity and wilting point: TAW = 1000 (θ FC – θ WP) Zr (8.71) where Zr is the effective rooting depth (m) and Zr contains Ze. RAW is estimated as: RAW = p TAW (8.72) where RAW has units of TAW (mm). Table 8.4 contains typical maximum values for Zr. FAO-56 describes several means to estimate the development (increase) in Zr with time for annual crops including in proportion to development of Kcb and in proportion to time. Other methods for Zr development include a sine function of time (Borg and Grimes, 1986), an exponential function of time dampened by soil temperature and soil moisture (Danuso et al., 1995) and a full root-growth simulation model by Jones et al. (1991). 8.5.5 Conditions for Maximum Transpiration The Kc values in Table 8.2 represent potential water consumption by healthy, relatively disease-free, and densely planted stands of vegetation having adequate levels of soil water. When stand density, height, or leaf area are significantly less than that attained under ideal or normal (pristine) conditions, the value for Kc may be reduced. Low stand density, height, and leaf area are caused by disease, low soil fertility, high soil salinity, waterlogging or water shortage (moisture stress), or by poor stand establishment. The reduction in the value for Kc during the midseason for poor crop stands can be as much as 0.3 to 0.5 for extremely poor crop stands and can be approximated according to the amount of effective (green) leaf area relative to that for healthy vegetation having normal planting densities. Procedures for reducing Kc according to the reduction in leaf area and the fraction of ground cover are given in Chapter 9 of FAO-56. 8.5.6 Application of the Basal Kcb Procedure over a Growing Season The first step in applying the basal Kcb approach is to construct the Kcb curve using Kcb ini, Kcb mid, and Kcb end similar to constructing the Kc curve. Equations for computing Ke (and Ks if necessary) are applied on a daily calculation timestep where daily Kcb is interpolated from the constructed Kcb curve. Figure 8.7 is an illustration of applying the Kcb + Ke procedure for a snap bean crop harvested for dry seed. The measured ETc

Chapter 8 Water Requirements

1.6 1.4

300

Snap Beans

FAO-56 PM ETo FAO-56 Kcbo FAO-56 Ke

250

Kcb, Kc

1.2

200

1

150

0.8 0.6

100

0.4

I

0.2 0

PP

135

I

P P 150

I

165

I

I

I

P 180

195

I

50 PP

210

225

240

0

Precipitation, net irrigation, mm

252

255

Day of Year, 1974 Basal Kcb

Ks Kcb + Ke

Kc from Lys. P = precipitation event I = irrigation

Figure 8.7. Measured (circles) and estimated (thin line) daily crop coefficients for a snap bean crop at Kimberly, Idaho. The basal crop curve (Kcb) was derived from Kcb values given in Table 8.2. (Data from Wright, 1990.)

data are from a precision lysimeter system at Kimberly, Idaho (Wright, 1990; Vanderkimpen, 1991). The soil at Kimberly had a silt loam texture. Soil evaporation parameters were Ze = 0.15 m, TEW = 34 mm, and REW = 8 mm. Nearly all wetting events were from alternate-row furrow irrigation so that the value for fw was set to 0.5. Irrigation events occurred at about midday or during early afternoon. The agreement between the estimated values for daily Kc from Equation 8.56 (thin, continuous line) and actual 24-hour measurements (symbols) is relatively good. 8.5.7 Alfalfa-Based Crop Coefficients Wright (1982, 1995) defined crop coefficients for crops common to higher elevations and latitudes of the western U.S. These coefficients, derived at Kimberly, Idaho, were initially based on the alfalfa reference ETr represented by the 1982 Kimberly Penman Equation (Equations 8.4–8.6). Wright developed both mean and basal Kc, denoted here as Kcr and Kcbr to distinguish them from the grass based Kc described earlier. These coefficients, if converted to an ETo basis by multiplying by the ratio of ETr to ETo and simplified to linear segment forms, agree closely with the Kco from Table 8.2 for similar crops. The Kimberly coefficients were converted from the 1982 Kimberly Penman ETr basis to the ASCE standardized Penman-Monteith ETrs basis by Allen and Wright (2006) and are listed in Tables 8.7 and 8.8. The Kc curves of Wright (1981, 1982) were changed little by the conversion due to the similarity in the two ETr references. Coefficient curves by Wright are based on the relative time from planting to effective full cover and on days after effective full cover. Effective full cover was defined by Wright (1982) as the time when leaves between row crops begin to interlock, at heading of grain, and flowering of peas. The Kc curves are constructed from the data in Tables 8.7 and 8.8 by using linear or curvilinear interpolation between adjacent columns or by fitting regression equations to the first half (planting to effective full cover) and to the second half (days after effective full cover) for each curve. Table 8.9 lists strategic dates of crop development recorded by Wright (1982).

Design and Operation of Farm Irrigation Systems

253

Table 8.7. Alfalfa-based mean (single) crop coefficients (Kcr) by Wright (1981, 1995) converted for use with the ASCE standardized PM equation (Allen and Wright, 2006). Percent Time from Planting to Effective Full Cover Date Crop 0 10 20 30 40 50 60 70 80 90 100 Spring grain[1] 0.20[2] 0.20 0.20 0.25 0.37 0.50 0.63 0.76 1.00 1.03 1.03 Peas 0.15 0.17 0.19 0.21 0.32 0.42 0.52 0.63 0.73 0.83 0.93 Sugar beets 0.26 0.26 0.26 0.26 0.26 0.28 0.30 0.38 0.55 0.74 1.03 Potatoes 0.20 0.20 0.20 0.22 0.30 0.41 0.53 0.67 0.73 0.77 0.8 Corn 0.20 0.20 0.20 0.20 0.24 0.34 0.44 0.58 0.72 0.90 1.00 Beans 0.20 0.20 0.22 0.26 0.35 0.45 0.55 0.68 0.83 0.95 0.97 Winter wheat 0.25 0.25 0.27 0.38 0.60 0.80 0.90 0.96 1.00 1.03 1.03 Days after Effective Full Cover Date Crop 0 10 20 30 40 50 60 70 80 90 100 Spring grain[1] 1.03 1.03 1.03 1.03 0.94 0.50 0.30 0.15 0.10 Peas 0.93 0.93 0.70 0.54 0.38 0.22 0.12 0.10 Sugar beets 1.03 1.03 1.03 1.00 0.97 0.92 0.82 0.74 0.65 0.61 0.56 Potatoes 0.80 0.80 0.76 0.72 0.68 0.63 0.58 0.50 0.38 0.20 0.15 Field corn 1.00 0.99 0.98 0.95 0.88 0.80 0.72 0.63 0.35 0.18 Sweet corn 1.00 0.97 0.94 0.90 0.84 0.70 0.55 0.35 0.20 0.10 Beans 0.97 0.97 0.94 0.64 0.32 0.15 0.10 0.05 Winter wheat 1.03 1.03 1.03 1.03 1.00 0.55 0.25 0.15 0.10 Alfalfa Hay: Percent Time from New Growth to Harvest and from Harvest to Harvest Cycle 0 10 20 30 40 50 60 70 80 90 100 First[3] 0.50 0.62 0.73 0.83 0.88 0.94 1.00 1.00 1.00 0.98 0.96 Intermediate 0.30 0.40 0.55 0.80 0.94 0.97 1.00 1.00 1.00 0.97 0.94 Last 0.30 0.35 0.45 0.53 0.58 0.58 0.54 0.48 0.46 0.44 0.44 Alfalfa Hay and Ryegrass: Total Season (including all cutting effects) (days) Crop 0 20 40 60 80 100 120 140 160 180 200 Alfalfa[4] 0.45 0.69 0.87 0.88 0.70 0.75 0.88 0.81 0.88 0.71 0.65 Alfalfa[5] 0.44 0.77 0.82 0.86 0.90 0.88 0.85 0.82 0.78 0.66 0.50 Alfalfa, overall seasonal mean 0.85 Perennial ryegrass 0.55 0.66 0.77 0.8 0.8 0.8 0.78 0.76 0.72 0.68 0.55 [1] [2]

[3]

[4]

[5]

Spring grain includes wheat and barley. The initial values 0.15 to 0.26 list for all crops are appropriate for relatively dry surface soil conditions from planting until significant crop development. For moderately wet surface soil, as with preemergence irrigation(s) or some precipitation, use 0.35, and for very wet conditions use 0.50. First denotes first growth cycle. Intermediate growth cycles (following first cutting) may number one or more, depending on length of season. The last harvest terminates when crop becomes dormant in freezing weather. Alfalfa cultivar was Ranger. Each value represents the average Kc over the 20-day period at Kimberly, Idaho. The first value represents from 0 to 10 days and the last value from 190 to 200 days. This is a seasonal curve for alfalfa cut periodically for hay where the values are from a smoothed, running average of Kc.

254

Chapter 8 Water Requirements

Table 8.8. Alfalfa-based basal crop coefficients (Kcbr) by Wright (1982, 1995), converted for application with the ASCE standardized PM equation (Allen and Wright, 2006). Percent Time from Planting to Effective Full Cover Date Crop 0 10 20 30 40 50 60 70 80 90 100 Spring grain[1] 0.15 0.15 0.15 0.19 0.24 0.36 0.48 0.62 0.92 0.98 1.03 Peas 0.12 0.13 0.14 0.15 0.18 0.27 0.36 0.50 0.65 0.78 0.92 Sugar beets 0.15 0.15 0.15 0.15 0.15 0.16 0.17 0.21 0.40 0.66 1.03 Potatoes 0.15 0.15 0.15 0.15 0.15 0.20 0.34 0.49 0.64 0.72 0.77 Corn 0.15 0.15 0.15 0.16 0.17 0.20 0.27 0.41 0.55 0.80 0.96 Beans 0.15 0.15 0.17 0.19 0.23 0.35 0.46 0.60 0.78 0.93 0.95 Winter wheat 0.12 0.12 0.14 0.22 0.45 0.70 0.84 0.96 1.00 1.03 1.03 Days after Effective Full Cover Date Crop 0 10 20 30 40 50 60 70 80 90 100 Spring grain[1] 1.03 1.03 1.03 1.03 0.94 0.40 0.15 0.07 0.05 Peas 0.92 0.92 0.72 0.52 0.32 0.16 0.07 0.05 Sugar beets 1.03 1.03 1.02 0.98 0.93 0.86 0.78 0.72 0.66 0.60 0.54 Potatoes 0.77 0.77 0.73 0.68 0.64 0.59 0.54 0.47 0.20 0.08 0.08 Field corn 0.96 0.96 0.96 0.92 0.85 0.79 0.72 0.62 0.28 0.16 0.12 Sweet corn 0.96 0.95 0.93 0.88 0.80 0.65 0.47 0.23 0.12 Beans 0.95 0.95 0.88 0.64 0.30 0.09 0.05 Winter wheat 1.03 1.03 1.03 1.03 1.00 0.50 0.20 0.10 0.05 Alfalfa Hay: Percent Time from New Growth to Harvest and from Harvest to Harvest Cycle 0 10 20 30 40 50 60 70 80 90 100 First[2] 0.35 0.45 0.56 0.72 0.82 0.90 1.00 1.00 1.00 0.98 0.96 Intermediate 0.25 0.30 0.42 0.72 0.90 0.95 1.00 1.00 0.98 0.96 0.94 Last 0.25 0.27 0.36 0.42 0.50 0.45 0.35 0.30 0.25 0.22 0.22 [1] [2]

Spring grain includes wheat and barley. First denotes first growth cycle. Intermediate growth cycles (following first cutting) may number one or more, depending on length of season. The last harvest terminates when crop becomes dormant in freezing weather. Alfalfa cultivar was Ranger.

Table 8.9. Dates (month/day) for crop growth stages at Kimberly, Idaho used in development of crop coefficient curves (after Wright, 1982). Rapid Full Head/ Crop Planting Emergence Growth Cover Bloom Ripening Harvest Spring grain 4/01 4/15 5/10 6/10 6/10 7/20 8/10 Peas 4/05 4/25 5/10 6/05 6/15 7/05 7/25 Sugar beets 4/15 5/10 6/01 7/10 10/15 Potatoes 4/25 5/25 6/10 7/10 7/01 9/20 10/10 Field corn 5/05 5/25 6/10 7/15 7/30 9/10 9/20 Sweet corn 5/05 5/25 6/10 7/15 7/20 8/15 Beans 5/22 6/05 6/15 7/15 7/05 8/15 8/30 (3/01) [1] 3/20 6/05 6/05 7/15 8/10 Winter wheat (2/15)[1] Alfalfa First cutting 4/01 4/20 6/15 Second cutting 6/15 6/25 7/31 Third cutting 7/31 8/10 9/15 Fourth cutting 9/15 10/01 10/30 [1]

For winter wheat, the first two dates are “effective dates” for green-up in the spring. Actual dates were 10/10 for planting and 10/25 for emergence, the previous year.

Design and Operation of Farm Irrigation Systems

1.2

300

Snap Beans

ASCE-PM ETr Wright (1982) Kcbr FAO-56 Ke

Precipitation, net irrigation, mm

1.4

255

250

1

200

Kcb, Kc

0.8

150

0.6

100

0.4

I

0.2 0

135

I

PP

PP

150

I

165

I

I

I

P 180

195

I

50 PP

210

225

240

0

Day of Year, 1974 Basal Kcb

Ks Kcb + Ke

Kc from Lys. P = precipitation event I = irrigation

Figure 8.8. Measured (circles) and estimated (thin line) daily crop coefficients for a snap bean crop at Kimberly, Idaho. The basal crop curve (Kcb) was derived from Kcbr values given in Table 8.8. (Data from Wright, 1990.)

8.5.8 Comparisons between FAO and Wright Kc Methods. Estimates of daily ETc by the linearized FAO-56 basal Kcb + Ke method and ETos basis for a snap bean crop at Kimberly, Idaho, were shown in Figure 8.7 along with measured daily Kc. The same measurement data are plotted in Figure 8.8, but using the Wright (1982) curvilinear basal Kcb data from Table 8.8, instead, for snap beans, and ETrs by the ASCE PM ETr equation instead of ETos. Ke was calculated in both applications using the FAO-56 method (Equations 8.60 to 8.68), with the only difference being in the calculation for Kc max (Equation 8.61). The measured Kc values were calculated for each day by dividing ET measured by the precision lysimeter system by ETo or ETr calculations. Stage lengths for the 1974 bean crop were 25/25/30/20 days for the FAO Kcb application and stage dates for constructing the curvilinear Kcb curve of Wright were taken from Table 8.9. Both methods for constructing the daily Kcb + Ke curves fit observed Kc relatively well (Figures 8.7 and 8.8), where wet soil evaporation following precipitation and irrigation temporarily increased values for Kc act. The water-balance based FAO-56 Ke method estimated Kc act to remain above the Kcb curve throughout the midseason period for the bean crop due to the frequent irrigation and low, sustained rate of evaporation caused by the nearly full ground cover. The estimated Kc act followed measured Kc relatively closely during the initial period when reduction in Kcb and Es was simulated due to dryness of the surface soil layer, and during the late season period when soil water content of the root zone was estimated to be insufficient to sustain ET at potential levels. These estimates were borne out by the lysimeter record. Lysimeter-computed Kc act on about five days around day 190 (beginning of midseason period) lay beneath both the FAO and Wright (1982) Kcb curves. The Wright Kcb curve (Figure 8.8) was constructed by Wright based on two years of data (only one is shown). The FAO-56 Kcb curve (Figure 8.7) was constructed to follow the same general shape as the Wright curve via the selection of stage lengths to provide for con-

256

Chapter 8 Water Requirements

sistent comparisons between methods. The maximum estimated and measured Kc act for the FAO-56 ETo-based method averaged about 1.35 during the midseason period (Figure 8.7), whereas the estimated and measured Kc act for the Wright method averaged about 1.0 during the midseason period (Figure 8.8) due to the use of the alfalfa reference basis. 8.5.9 Soil Water Balance The calculation of Ks requires a daily water balance computation for the root zone. Often, for purposes of estimating Kc, the water content of the root zone is expressed in terms of net depletion. This makes the adding and subtracting of losses and gains straightforward and various components of the soil water budget can be expressed in terms of water depth. Rainfall, irrigation, and capillary rise of groundwater add water to the root zone and decrease the root zone depletion. Soil evaporation, crop transpiration, and percolation losses remove water from the root zone and increase the depletion. A daily water balance, expressed in terms of depletion at the end of the day, is: Dr , i = Dr , i -1 − ( P − RO) i − I i − CRi + ETc act , i + DPi

(8.73)

where Dr,i = root zone depletion at the end of day i, mm Dr,i-1 = water content in the root zone at the end of the previous day, i-1, mm Pi = precipitation on day i, mm ROi = runoff from the soil surface on day i, mm Ii = net irrigation depth on day i that infiltrates the soil, mm CRi = capillary rise from the groundwater table on day i, mm ETc act,i = actual crop evapotranspiration on day i, mm DPi = water flux out of the root zone by deep percolation on day i, mm. Although following heavy rain or irrigation, soil water content might temporally exceed field capacity, in the above equation, the total amount of water exceeding field capacity is assumed to be lost the same day via deep percolation, following any ET for that day. This does permit the extraction of one day’s ET from this excess before percolation. The root zone depletion will gradually increase as a result of ET. In the absence of a wetting event, the root zone depletion will reach the value TAW defined from rooting depth, θ FC and θWP from Equation 8.71. At that moment no water is left for ET, and Ks becomes zero, from Equation 8.70. Limits imposed on Dr,i are: 0 ≤ Dr , i ≤ TAW (8.74) 8.5.9.1 Initial depletion. To initiate the water balance for the root zone, the initial depletion Dr,i-1 can be derived from measured soil water content by: Dr , i-1 = 1000(θ FC − θ i -1 ) Z r (8.75)

where θ i-1 is the average soil water content at the end of day i – 1 for the effective root zone. Following heavy rain or irrigation, the user can assume that the root zone is near field capacity, i.e., Dr,i-1 ≈ 0. Pi is equivalent to daily precipitation. Daily precipitation in amounts less than about 0.2 ETref is normally entirely evaporated and can generally be ignored in depletion calculations. Ii is equivalent to the mean infiltrated irrigation depth expressed for the entire field surface. Runoff from the surface during precipitation can be estimated using standard procedures from hydrological texts. 8.5.9.2 Capillary rise (CR). The amount of water transported upwards by capillary rise from the water table to the root zone depends on the soil type, the depth of the water table and the wetness of the root zone. Capillary rise can normally be assumed

Design and Operation of Farm Irrigation Systems

257

to be zero when the water table is more than about 1 m below the bottom of the root zone. CR can be computed from parametric equations as a function of the water table depth, water content of the root zone, root density and soil characteristics (Liu et al., 2006). Some information on capillary rise is presented in Chapter 9. 8.5.9.3 Deep percolation from the root zone (DP). Following heavy rain or irrigation, the soil water content in the root zone may exceed field capacity. In applications of Equation 8.73, DP is assumed to occur within the same day of a wetting event, so that the depletion Dr, i in Equation 8.73 becomes zero. Therefore, DPi = ( Pi − ROi ) + I i − ETc, i − Dr , i -1

≥0

(8.76)

As long as the soil water content in the root zone is below field capacity (i.e., Dr, i > 0), the soil will not drain and DPi = 0. The DPi term in Equations 8.73 and 8.76 is not to be confused with the DPe,i term used in Equations 8.66 and 8.69 for the evaporation layer. Both terms can be calculated at the same time, but are independent of one another. 8.5.9.4 Depth of root zone. The depth of the effective root zone can be estimated as: DoY i - DoY ini (8.77) z r i = z r min + ( z r max - z r min ) Lroot growth for DoYi ≤ DoYini + Lroot growth where zr i = effective depth of the root zone on day i, mm zr min = initial effective depth of the root zone (generally at DoY = DoYini) zr max = maximum effective depth of the root zone, m, reached Lroot growth days following DoYini DoYi = day of the year (1–365) corresponding to day i DoYini = day of the year corresponding to the date of planting or initiation of growth (or January 1 if a perennial is growing all of the year). When DoYi > DoYini + Lroot growth, zr = zr max. Other nonlinear schemes for root growth such as that by Borg and Grimes (1986) can be applied, although these are often not very different from linear estimates. 8.5.10 ET During the Nongrowing Season During nongrowing periods, ET is dominated by evaporation, rather than transpiration, especially if the nongrowing season is caused by killing frosts. Nongrowing season ET is therefore generally best estimated using techniques that accurately estimate evaporation from the soil surface. Evaporation is a strong function of wetting frequency and reference ET rate and therefore, the Kc ini estimated from Figure 8.4 can be used as an estimate of Kc during the nongrowing season (Martin and Gilley 1993) with some adjustment for impacts of surface cover by dead vegetation, as described in Section 8.9.2. Alternatively, Kc during the nongrowing season can be estimated using the dual Kcb + Ke method with some adjustment to TEW and REW during cold periods (Allen et al., 2005a,b). The dual procedure was applied by Allen and Robison (2007) for complete calendar years including winter, with adjustments made during periods of snow cover. Snyder and Eching (2004, 2005) suggested a procedure for combining the Kc during the nongrowing season, estimated using a procedure similar to Figure 8.4, with the Kc curve for the growing season to create a continuous Kc for the entire calendar year. This was done by taking the maximum, for each day, of Kc ini as estimated from Figure 8.4 or similar curve and the Kc obtained from the growing season Kc ini curve.

258

Chapter 8 Water Requirements

8.6 EVAPOTRANSPIRATION COEFFICIENTS FOR LANDSCAPES Over the past several decades, the water requirements and water consumption by residential and urban landscapes have become increasingly important in terms of quantity and value of water consumed. Procedures similar to those from agriculture have been adapted to estimate ET from landscapes. However, two distinctions between agriculture and landscapes exist regarding ET quantification: landscape systems are nearly always comprised of a mixture of multiple types and species of vegetation, thereby complicating the estimation of ET, and, typically, the objective of landscape irrigation is to promote appearance rather than biomass production, as is the case in agriculture. Therefore, target ET for landscapes may include an intentional “stress” factor in the baseline value for ETc act, where landscape plants are watered less than they would be if they were irrigated like a crop. They are watered enough to look good and to survive, but the plants are stressed and will not be at maximum productivity. This adjustment can result in significant water conservation. The magnitude of the stress factor depends on physiological and morphological requirements of the plants; the goal is to sustain health and appearance with minimal irrigation. For example, water conservation studies on turfgrasses have demonstrated that water savings of 30% for cool-season turfgrasses and 40% for warm-season turfgrasses may be attainable without significant loss of quality (Pittenger and Shaw, 2001). Many shrubs and groundcovers can be managed for even more stress-induced reduction in ET. Additionally, few landscape sites meet the “extensive surface” requirement. Because of the frequent inclusion of water stress in target ET values for landscape design and management, distinction must be made between these target ET values and actual ET values. Actual ET values may exceed target ET values if the landscape receives more water than required by the target that includes intentional stress. Under these conditions, landscape vegetation may exploit the additional available water, subject to some limit constrained by environmental energy for evaporation and leaf area. This limit, which follows behavior and principles used for agricultural crops, may exceed the targeted ET rate for the particular landscape cover. Conversely, actual ET may be less than target ET values if actual stress levels to the landscape are greater than targeted. Therefore, two ET values for landscape are distinguished here. The first is the target landscape ET, referred to as ETL, that is based on minimum ET levels, relative to climate, necessary to sustain a healthy, attractive landscape. The second ET value is the actual landscape ET, ETL act, that is based on landscape type and on actual water availability. The target ET for a landscape is calculated as (8.78) ETL = KL ETo -1 -1 -1 where ETL is the target landscape ET (in mm d , mm month , or mm year ), and ETo is the grass reference ET in the same units. KL is the target landscape coefficient, similar to the crop coefficient used in agricultural applications. There has been relatively limited experimental research on quantifying water needs of the diverse array of landscape plant types (Pittenger and Henry 2005). Much of the existing information is based largely on observation rather than on scientifically obtained data. Some of the leading work on landscape ET has been done in California, where water applied to landscapes in southern California is estimated to be 25% to

Design and Operation of Farm Irrigation Systems

259

30% of all water used in the region (Pittenger and Shaw 2001). Pittenger and Shaw (2007, unpublished work from University of California Cooperative Extension) have produced a table of KL values for 35 landscape groundcovers and shrubs that provide acceptable landscape performance after establishment, but that cause a managed amount of moisture stress via limited water application. Costello et al. (2000) and IA (2005) described a recent procedure termed WUCOLS (water use classification of landscape species), where the KL has been decoupled into reproducible and visually apparent components representing the effects of three or fourfour factors that control the value for KL. The decoupling was done to provide for application to the wide diversity of vegetation types and environments of landscape systems. Snyder and Eching (2004, 2005) have proposed a similar decoupled procedure for estimating a formulated KL, but which uses different ranges for the compoents. The Snyder and Eching (2004) procedure is described here, where (8.79) KL = Kv Kd Kmc Ksm where Kv is a vegetation species factor, Kd is a vegetation density factor, Kmc is a microclimate factor, and Ksm is a managed stress factor. Kv can be considered to be the ratio of ETL to ETo for a specific single or mixture of plant species under full or nearly full ground cover and full soil water supply. Factors Kd, Kmc and Ksm modify Kv for less than effective full ground-cover, for impacts of shading or exposure to advective or reflective sources, and for intentional water stress. Each of these factors can be estimated separate from the other, based on visual observation of the landscape (Kd and Kmc) and based on visual observation and grower experience (Ksm). Following the estimation of the individual factors, KL is calculated using Eq. 8.79 and represents a relatively accurate and reproducible estimate of relative landscape ET. The procedure of Snyder and Eching (2004, 2005), used in the University of California-Davis LIMP software, differs from that of WUCOLS (Costello et al. 2000) in the ranges used to define Kd. In addition, the procedure of WUCOLS combines the values for Kv and Ksm into a “species” coefficient which can make the combined product difficult to estimate. The procedure and factor ranges of LIMP and Pittenger et al. (2001) may be more likely to produce more accurate and reproducible estimates of landscape ET. 8.6.1 Vegetation Coefficient The value for Kv for landscape vegetation represents the ratio of ETL to ETo for full or nearly full cover (shading) of the ground and with full soil water supply and is used to estimate the maximum, potential rate of ratio of KL for the vegetation under ideal conditions. Costello and Jones (2000) have suggested values for a “species coefficient,” Ksp, for hundreds of landscape species in California, however their values include the Ksm factor and therefore should not be used with the LIMP procedure. Based on the definition for Kv used here, where Kv is fraction of ETo when the foliage is at near maximum density (Kd = 1) and full water availability (Ksm = 1), large numbers of landscape vegetation tend to have similar values for Kv due to similarity in total leaf area and stomatal response. Therefore, a condensed table of typical values for general species types can be used to provide general estimates for Kv for used in Equation 8.79, where Kv ranges from about 0.8 to 1.2. Historically the grass reference ETo has been used to estimate ET from landscapes, thus the upper limit for Kv can exceed 1.0 for some tall, leafy vegetation. Table 8.10 contains general values for Kv for general types of landscape vegetation.

260

Chapter 8 Water Requirements

Table 8.10. Vegetation (species) factors (Kv) for general plant types. Vegetation Category[1] Kv Trees 1.15 Shrubs, desert species 0.7 nondesert 0.8 Groundcover 1.0 Annuals (flowers) 0.9 Mixture of trees, shrubs, 1.20 and groundcover[2] Cool season turfgrass[3] 0.9 Warm season turfgrass[4] 0.9 [1]

[2]

[3] [4]

The tree, shrub, and groundcover categories listed are for landscapes that are composed solely or predominantly of one of these vegetation types. Mixed plantings are composed of two or three vegetation types (i.e., where a single vegetation type does not predominate). Cool season grasses include Kentucky Blue grass, fescues, perennial ryegrass Warm season grasses include Bermuda grass, St. Augustine grass, buffalo grass, and blue grama.

The typical Kv values in Table 8.10 represent nearly full effective ground cover (fc > ~0.7) (see Section 8.5.3), and no water stress. The values in Table 8.10 are general and under most conditions, these values are not met due to the density factor being less than 1.0 and managed, intentional moisture stress. The Kv value for warm season grass is equal to that for cool season grass in Table 8.10 since both of these values represent ETL/ETo for nonstress conditions. Typically, warm season grasses can tolerate more moisture stress than cool season grasses so that a lower managed stress factor can be applied to warm season grasses with less visual effect, as is illustrated later in Table 8.13. The Kv values for both cool season and warm season grasses are less than 1.0 in Table 8.10 due to the tendency for their mean height to be less than that of the standardized 0.12 m grass reference. 8.6.2 Density Factor Landscapes can vary considerably in vegetation density, due to wide variations in plant spacing and maturity. Plant density factors, Kd, are listed in Table 8.11 for the primary landscape vegetation types. Vegetation density refers to the collective leaf area of all plants in the unit landscape area. More densely growing vegetation has a higher Kd and will transpire and require more water. Immature and sparsely planted landscapes typically have less total leaf area per unit landscape area than mature landscapes and are assigned a Kd value in the low category. Often, landscapes have two and three tiers of vegetation including turf or groundcover, shrubs, and trees. Overlapping tiers are capable of more radiative and other energy exchange and tend to increase ET. Allen et al. (1998) introduced a general equation for Kd that is based on the fraction of ground covered (or shaded at noon) by vegetation and mean plant height. This relationship has the form: ⎛ 1 ⎞⎞ ⎛ ⎜ ⎟ ⎜ 1+h ⎠ ⎟ (8.80) K d = min⎜1, M L f c eff , f c⎝eff ⎟ ⎜ ⎟ ⎝ ⎠ where fc eff is the effective fraction of ground covered or shaded by vegetation (0 to 1.0) near solar noon, h is the mean height of the vegetation in m, and ML is a multiplier on fc eff to impose an upper limit on relative transpiration per unit ground area as

Design and Operation of Farm Irrigation Systems

261

Table 8.11. Density factor (Kd) for different vegetation heights, h, over a range of effective fraction of ground covered or shaded by vegetation, fc eff, from Equation 8.80, where ML = 1.5. h = 0.1 m h = 0.4 m h=1m h=4m fc eff 0.0 0.00 0.00 0.00 0.00 0.1 0.12 0.15 0.15 0.15 0.2 0.23 0.30 0.30 0.30 0.3 0.33 0.42 0.45 0.45 0.4 0.43 0.52 0.60 0.60 0.5 0.53 0.61 0.71 0.87 0.6 0.63 0.69 0.77 0.90 0.7 0.72 0.78 0.84 0.93 0.8 0.82 0.85 0.89 0.96 0.9 0.91 0.93 0.95 0.98 1.0 1.00 1.00 1.00 1.00

represented by fc eff. Equation 8.80 estimates larger values for Kd as vegetation height increases. This accounts for the impact of larger aerodynamic roughness and generally more leaf area with taller vegetation, given the same fraction of ground covered or shaded. Allen et al. (1998) provided means to estimate values for fc eff as a function of time of year, latitude, vegetation height, canopy shape and, in the case of row vegetation, row orientation. These relationships are provided in Equations 8.81 and 8.82 for randomly placed vegetation. For practical purposes, and due to the uncertainties associated with estimating fc for landscape vegetation, the value for fc eff can often be assumed to be the same as fc. The “min” function takes the smallest of the three values separated by commas. Parameter ML is expected to range from 1.5 to 2.0, depending on the canopy density and thickness and the value can be modified to fit the specific vegetation. The ML fc eff limit usually becomes invoked only for h greater than about 1 to 2 m. For canopies such as trees or randomly planted vegetation, fc eff can be estimated as fc f c eff = ≤1 (8.81) sin(β ) where β is the mean angle of the sun above the horizon during the period of maximum ET (generally between 1100 and 1500 hours). β is calculated from Equation 8.43. Generally, fc eff can be calculated at solar noon when ω = 0 and Equation 8.43 becomes β = arcsin [sin(ϕ )sin(δ ) + cos(ϕ )cos(δ )] (8.82) Figure 8.9 shows Kd estimated by Equation 8.80 over a range of fraction of (effective) ground cover and various mean heights for vegetation and compared to estimates for Kd from other sources for specific vegetation types. Hernandez-Suarez (1998) produced estimates for Kd for low-growing vegetation and Fereres (1991) produced estimates for tree orchards. Snyder and Eching (2005) introduced an equation fitted to the Fereres (1991) data (Sammis et al., 2004). The Kd estimates by Hernandez-Suarez are similar to those by Equation 8.80 for h of 0.1 to 0.3 m and those by Fereres (1991) are simlar to those by Equation 8.80 for h = 5 m. These comparisons suggest that Equation 8.80 works well as a general estimate for Kd over a range of fc eff and h.

262

Chapter 8 Water Requirements

1.2 1.0

Fereres and Snyder and Eching (2005) (Orchards)

h=5m

0.8

K

d

h=3m h=1m

0.6

h = 0.4 m

0.4

Hernandez-Suarez (Vegetables) Eq. 8.82 , h = 0.1 m

0.2 0.0

0

0.2

0.4

0.6

0.8

1

Fraction of Ground Cover Figure 8.9. Density coefficient, Kd, estimated from Equation 8.80 with ML = 2 over a range of fraction of ground cover and various plant heights, and compared with estimates by Fereres (1981) for orchards and Hernandez-Suarez (1988) for vegetables.

When two substantial tiers of vegetation are present, for example trees shading grass or flowers, the value for h can be approximated as a geometric mean of the heights of the vegetation or in proportion to the fc for each tier. The value for fc should be the total fc for the two vegetation tiers combined, accounting for any overlapped shading, i.e. using a weighted average. It is important to note that the Kd factor estimated in this section from fc presumes that the fraction of ground surface not covered by vegetation (i.e., bare soil), is relatively dry and therefore does not contribute substantially to the ETL. In situations where the exposed soil surface is wet a majority of the time, the value for Kd should be increased by 10 to 20%, subject to Kd ≤ 1, to account for soil surface evaporation, especially for trees and shrubs. Alternatively, a daily dual Kcb + Ke procedure, described previously, can be applied where Kcb is set equal to KL from Eq. 8.79 and Ke is estimated from wetting frequency. 8.6.3 Microclimate Factor Structures and paved areas typical of urban landscapes have pronounced effect on the local energy balance and thus ET demand of adjacent vegetated areas due to the transfer of additional energy for evaporation. The environmental conditions of a landscape may vary significantly across the landscape, for example, areas on the south side of a building vs. on the north side. The microclimate factor, Kmc, accounts for impacts of sun, shading, protected areas, hot or cool areas, reflected and emitted radiation from structures, wind, and transfer of heat energy from low ET surroundings on ET. Plantings adjacent to paved, open areas may have 50% higher ET demand than similar plantings among other vegetation due to the transfer of energy to the vegetation. Conversely, plantings in shaded areas from sun and wind may have ET rates only one-half as high as those for open settings. Values for Kmc are listed in Table 8.12 for the general classes of vegetation. In general, the “high” category (Kmc > 1) in Table 8.12 reflects harsh microclimate conditions such as planting in direct sunlight near a paved or other nonvegetated surface, planting near a reflecting window or heat-absorbing surfaces, or in exposed, windy conditions. The “low” category (Kmc < 1) reflects an environment where the planting is

Design and Operation of Farm Irrigation Systems

263

Table 8.12. Microclimate factor, Kmc, for different plant types (from Irrigation Association, 2003). Average (reference Vegetation High condition) Low Trees 1.4 1.0 0.5 Shrubs 1.3 1.0 0.5 Groundcover, flowers 1.2 1.0 0.5 Mixture of trees, shrubs, and groundcover 1.4 1.0 0.5 Turfgrass 1.2 1.0 0.8

located in shade, shielded from wind, and away from dry, hot surfaces. The average or medium category (Kmc = 1) represents a reference condition similar to an open park setting, where conditions caused by buildings, pavement, shade, and reflection do not influence the ET by the landscape. The values given for Kmc are only approximate and local measurements are recommended to confirm or to derive local values. The water manager should select the appropriate microclimate adjustment factor for each sector of a landscape or for each irrigation zone. For example, a turfgrass zone that thrives in sun may have an average Kmc of 1.0. The same zone in full shade of a building during midday may be assigned a Kmc of 0.8 or lower to better reflect the actual plant water needs. 8.6.4 Managed Stress Factor As stated previously, the typical objective of landscape irrigation is to promote appearance rather than biomass production, as is the case in agriculture. Therefore, target ET for landscapes may include an intentional and managed “stress” factor in the baseline value for ETc act, where landscape plants are watered less than they would be if they were irrigated like a crop. This management is done by adjusting irrigation water schedules to apply less water than the vegetation will potentially transpire. The magnitude of the stress factor depends on physiological and morphological requirements of the plants. For example, water conservation studies on turfgrasses have demonstrated that water savings of 30% for cool-season turfgrasses and 40% for warm-season turfgrasses may be attainable without significant loss of quality (Pittenger and Shaw, 2001). Many woody shrubs and groundcovers can be managed for even more stressinduced reduction in ET (Kjelgren et al., 2000). Pittenger et al. (2001), Shaw and Pittenger (2004), and Pittenger and Shaw (2007) defined water needs of non-turf landscape plants as a percentage of ETo needed to maintain their appearance and intended function (e.g. shade, green foliage, screening element). In Equation 8.79, the landscape coefficient KL (equivalent to the crop coefficient) is decoupled into components that describe the impacts of vegetation type, density of vegetation, microclimatic effects, and managed stress factor. The managed stress factor, Ksm, represents the fraction of full ET rate targeted to obtain the functional and visual characteristics of the landscape vegetation. Parameter Ksm has a range of 0 to 1.0 where 1.0 represents conditions of no moisture stress (and no real water conservation) and 0 represents complete lapse of plant transpiration (i.e., probable plant death). High values for Ksm will sustain relatively more lush, high leaf-area vegetation stands that tend to maximize ET and that may be necessary to sustain long-term plant health or appearance. Low values for Ksm represent substantial managed plant

264

Chapter 8 Water Requirements

water stress and reduction in ET, generally at the cost of biomass accumulation and potentially visual effects (Richie and Pittenger, 2000). Many landscape species can exercise significant stomatal control and can be “forced” to support relatively lower levels of ET. For instance, the low range for groundcover is 0.2, which is appropriate for a selected group of drought-tolerant groundcover species. This value may not be appropriate for some ornamental groundcovers, however, that require higher amounts of water (and less water stress) to maintain health and appearance. It is recommended that one consult with local or regional sources to determine the appropriate managed stress factor (Ksm) to apply. Pettinger and Shaw (2007) suggested KL for more than 30 groundcovers and shrubs grown in southern California that contain low Ksm components and thus provide good water conservation. Many of the vegetation types listed by Pettinger and Shaw are native desert vegetation types that tolerate water stress. Other sources of KL information for specific species where the KL includes an implied Ksm < 1 include the WUCOLS publication by Costello and Jones (1999). 8.6.5 Actual ET from Landscapes As discussed in the previous section, the vegetation coefficient KV represents a sufficient water supply to support full ET from relatively dense vegetation having near maximum ground cover and open environmental exposure. However, the KL coefficient may contain an implicit amount of managed stress for purposes of water conservation.. The degree of implied managed stress is quantified in Eq. 8.79 by the Ksm term. Therefore, the KL derived from Equation 8.79 using a recommended Ksm term may not represent actual conditions where actual stress deviates from the managed or target stress. Under these conditions, for purposes of water balance and determination of consumptive use from a landscaped or larger area, the “managed stress” coefficient in Equation 8.79 must be replaced by an “actual” stress coefficient, Ks, where Ks is computed from Equation 8.70 based on soil water depletion determined from a daily balance of root-zone soil water, for example usng Equation 8.73. Equation 8.79 therefore assumes the form: KL act = Kv Kd Kmc Ks (8.83) where Ksm has been replaced by an actual stress coefficient Ks and where KL act is the actual ET anticipated from the landscape under actual water availability. Following Table 8.13. Managed stress factors (Ksm) for general plant types and the general values of depletion fraction for no stress. Average Depletion managed fraction, p, Vegetation Category High stress stress Low stress for no stress Trees 0.4 0.6 0.8 0.6 Shrubs, desert species 0.4 0.6 0.3 0.6 nondesert 0.6 0.8 0.4 0.6 Groundcover 0.3 0.5 0.8 0.5 0.5 0.7 0.8 0.4 Annuals (flowers) Mixture of trees, shrubs, and groundcover[1] 0.6 0.8 0.4 0.6 0.7 0.8 0.9 0.4 Cool season turfgrass Warm season turfgrass 0.6 0.7 0.8 0.5 [1]

Mixed plantings are composed of two or three vegetation types and where a single vegetation type does not predominate.

Design and Operation of Farm Irrigation Systems

265

calculation of KL act, actual ET from the landscape under actual watering conditions is ETL act = K L act ETo (8.84) When Ks is estimated from Equation 8.70, the depletion fraction parameter p used to estimate RAW can be set to values listed in Table 8.13. However, it is recommended that one use specific values determined for specific species if these are available. The effective depth of the root zone used to estimate TAW can be species or variety specific and therefore information specific to the variety can be important.

8.7 ESTIMATING Kc FROM FRACTION OF COVER

When a Kc value is needed for vegetation that is not similar to that listed in Tables 8.2, 8.3, or 8.5, or where a more specific value for KL (or the product KLKd ) is desired than provided by Table 8.10, the function of FAO-56 used to create the Kd density function and factor of Equation 8.80 can be applied. The estimation of Kc for the initial growth stage of annuals, where the soil surface is mostly bare, can be determined from following Section 8.5.3.4 for the mean Kc (Kc ini) where the crop coefficient in this stage is primarily determined by the frequency with which the soil is wetted. In the dual Kc approach, the Kcb for the initial period can be estimated as 0.1 to 0.15 for bare soil. The Kc during the midseason and late season periods, if for a low fraction of ground cover, will be affected to a large extent by the frequency of precipitation and/or irrigation. Therefore, the basal Kcb + Ke approach is recommended, with Kcb estimated using the following Equation 8.85 based on the fraction of ground covered by vegetation during the particular period. For landscape crops, the impact of less than full ground shading on the ETL is incorporated by the Kd parameter in Equation 8.79 or 8.83 for KL. In the case of agricultural crops, Kd is used in a similar way: (8.85a) K cb = K c min + K d K cb full − K c min

(

)

where Kcb = estimated basal Kcb (for example, during the midseason) when plant density and/or leaf area are at or lower than full cover Kc min = the minimum Kcb representing bare soil Kcb full = the basal Kcb anticipated for the vegetation under full cover conditions and corrected for climate Kd = the density factor from Equation 8.80. Kc min ≈ 0.0 during long periods of no rain or irrigation and Kc min ≈ 0.15 to 0.20 during periods of rain or irrigation. For tree crops having grass or other ground cover, Equation 8.85a can be re-expressed as: (8.85b) K cb = K cb cover + K d (max K cb full − K cb cover , 0 )

[

]

where Kcb cover is the Kcb of the ground cover in the absence of tree foliage. Kcb full for use with ETo can be approximated as a function of mean plant height and adjusted for climate similar to the Kcb mid: 0 .3

⎛h⎞ (8.86a) full = min (1.0 + 0.1 h, 1.20 ) + [0.04 (u 2 − 2) − 0.004 ( RH min − 45) ]⎜ ⎟ ⎝3⎠ For use with alfalfa reference ETr, the climatic correction is not required for Kcb full because of the characteristics of the alfalfa reference crop. The equation is then: K cb full = min (0.8 + 0.1 h, 1.0 ) (8.86b)

K cb

266

Chapter 8 Water Requirements

8.8 EFFECT OF IRRIGATION METHOD ON KC

The type of irrigation system can affect the magnitude of ETc, principally in the fraction of the field surface that is wetted during irrigation and in the irrigation interval. These two characteristics affect the average rate of evaporation from the soil surface and the frequency with which the soil surface is wetted. Other factors influencing ETc are the amounts of evaporation losses from droplets during sprinkle irrigation and evaporation of intercepted water from wet canopies following sprinkle irrigation. All water evaporated from sprinkle droplets and from wet canopies consumes energy, as liquid is changed to vapor. This energy is supplied from sensible heat contained in the lower atmosphere and from solar radiation. The evaporation process cools and humidifies the air. As a result, the ET demand placed on the vegetation canopy itself (inside the leaf stomates) is generally moderated during the free water evaporation of water droplets. This effectively reduces the ETc demand from the soil profile estimated to occur in the absence of the free water evaporation. Therefore, the evaporation losses computed from free water surfaces during sprinkle irrigation cannot be directly added to the potential ETc demand as computed with Equation 8.56. As implied by Equation 8.61, an upper limit on the sum of free water evaporation from sprinkler irrigation and ETc exists. This sum approximately equals Kc max ETref. Therefore, Kc max ETref can be used to represent the maximum amount of evaporation and transpiration losses to be expected from an entire field during and immediately following irrigation for most types of irrigation systems.

8.9 EFFECTS OF SURFACE MULCHING ON KC

Mulches are frequently used in vegetable production to reduce evaporation losses from the soil surface, to accelerate crop development in cool climates by increasing soil temperature, to reduce erosion, or to assist in weed control. Mulches may be composed of organic plant materials or they may be synthetic mulches comprised of plastic sheets. Plastic mulches are the most common type of mulch used in vegetable production. 8.9.1 Plastic Mulches Plastic mulches (or covers) are generally comprised of thin sheets of polyethylene or similar material placed over the ground surface and generally along plant rows. Holes are cut through the film at plant spacings to allow emergence of vegetation. Polyethylene covers are usually either clear or black. Effects on ETc by the two colors are generally similar (Haddadin and Ghawi, 1983; Battikhi and Hill, 1988a,b; and Safadi, 1991). Plastic mulches substantially reduce the evaporation of water from the soil surface, especially under trickle irrigation systems. Associated with the reduction in evaporation is a general increase in transpiration from vegetation caused by transfer of both sensible and radiative heat from the surface of the plastic cover to adjacent vegetation. Usually, the ETc from mulched vegetables is about 5% to 30% lower than for vegetable production without a plastic mulch. A summary of observed reductions in Kc, evaporation, and increases in transpiration over growing seasons is given in Table 8.14 for five vegetable crops. Even though the transpiration rates under mulch may increase by an average of 10% to 30% over the season as compared to using no mulch, the Kc’s decrease by an average of 10% to 30% due to the 50% to 80% reduction in evaporation from wet soil. Generally, crop growth rates and vegetable yields are increased with the use of plastic mulches. To consider the effects of plastic mulch on ETc, the values for mean Kc mid and Kc end for vegetables listed in tables can be reduced by 10% to 30%, depending on the fre-

Design and Operation of Farm Irrigation Systems

267

Table 8.14. Approximate reductions in Kc, surface evaporation, and increases in transpiration for various vegetable crops under plastic mulch as compared with no mulch using trickle irrigation (from Allen et al., 1998). Reduction in Increase in Reduction Evaporation[1] Transpiration[1] in Kc[1] (%) (%) (%) Crop Squash 5-15 40-70 10-30 Cucumber 15-20 40-60 15-30 Cantaloupe 5-10 80 35 Watermelon 25-30 90 –10[2] Tomato 35 n/a n/a Average 10-30 50-80 10-30 [1] [2]

Relative to using no mulch. A negative increase, i.e., a decrease in transpiration.

quency of irrigation; use the higher value for frequent trickle irrigation. The value for Kc ini is often as low as 0.10. When adjusting the basal Kcb for mulched production, less adjustment is needed, as compared to the mean Kc curve, being on the order of perhaps 5% to 15% reduction in Kcb, since the basal evaporation of water from the soil surface is less with a plastic mulch, but the transpiration is relatively more. Local calibration of Kcb (and Ke) is encouraged. When applying a basal approach with plastic mulch, fw should represent the relative equivalent fraction of the ground surface that contributes to evaporation through the vent holes in the plastic cover. This fraction can be substantially (at least 2 to 3 times) larger than the area of the vent holes to account for vapor transfer from under the sheet. 8.9.2 Organic Mulches Organic mulches are sometimes used with orchard production and row crops under reduced-tillage operations. Organic mulches may be comprised of unincorporated plant residue or foreign material imported to the field. The depth of the organic mulch and the fraction of the soil surface covered can vary widely. These two parameters affect the amount of reduction in evaporation from the soil surface. A general rule of thumb with a mulched surface is to reduce the amount of soil water evaporation by about 5% for each 10% of soil surface that is covered by the organic mulch. For example, if 50% of the soil surface were covered by an organic cropresidue mulch, then the soil evaporation would be reduced by about 25%. To apply this to the Kc values in tables, one would reduce Kc ini values by about 25% and would reduce Kc mid values by 25% of the difference between Kc mid and Kcb mid. When applying the basal approach with separate water balance of the surface soil layer, the magnitude of Esp can be reduced by about 5% for each 10% of soil surface covered by the organic mulch. These recommendations are approximate and attempt to account for the effects of partial reflection of solar radiation from residue, microadvection of heat from residue into the soil, lateral movement of soil water from below residue to exposed soil, and the insulating effect of the organic cover. These parameters can vary widely, so that local research and measurement are encouraged.

8.10 PRECIPITATION RUNOFF Runoff during precipitation events is strongly influenced by land cover, soil texture, soil structure, sealing and crusting of the soil surface, land slope, local land forming (tillage and furrowing), antecedent moisture, and precipitation intensity and dura-

268

Chapter 8 Water Requirements

tion. Therefore, estimation of runoff during precipitation is generally fraught with uncertainty. For general purposes, runoff can be estimated using the USDA-NRCS curve number approach. The NRCS curve number is simple to apply and is widely used within the hydrologic, soils, and water resources communities. Required data are daily precipitation depth and computation of a daily soil water balance by which to select the antecedent soil water condition. 8.10.1 The NRCS Curve Number The curve number (CN) represents the relative imperviousness of a soil-vegetation complex and ranges from 0 for infinite perviousness and infiltration to 100 for complete imperviousness and total runoff (beyond abstraction). Generally the value for CN is selected from standard tables based on general crop and soil type and is adjusted for the soil water content prior to the wetting event. Examples of values for CN for various crop and soil combinations are given in Table 8.15. Parameter S in the CN procedure is the maximum depth of water that can be retained as infiltration and canopy interception during a single precipitation event (in mm). S is calculated as ⎛ 100 ⎞ S = 250 ⎜ − 1⎟ ⎝ CN ⎠

(8.87)

and surface runoff is then calculated for P > 0.2 S as

(P − 0.2S )2

(8.88) P + 0 .8 S where RO is the depth of surface runoff during the event (mm), and P is the depth of rainfall during the event (mm). The 0.2S term is abstracted precipitation, i.e., that intercepted by canopy and soil surface before any runoff occurs. If P < 0.2 S, then RO = 0.0. In addition, RO ≤ P applies. RO =

Table 8.15. Typical curve numbers for general crops for antecedent soil water condition (AWC) II from USDA-SCS (1972) and Allen (1988). Soil Texture Crop Coarse Medium Fine Spring wheat 63 75 85 Winter wheat 65 75 85 Field corn 67 75 85 Potatoes 70 76 88 Sugar beets 67 74 86 Peas 63 70 82 Dry edible beans 67 75 85 Sorghum 67 73 82 Cotton 67 75 83 Paddy rice 50 60 70 Sugar cane, virgin 60 69 75 Sugar cane, ratoon 60 68 76 Fruit trees, bare soil 65 72 82 Fruit trees, ground cover 60 68 70 Small garden vegetables 72 80 88 Tomatoes 65 72 82 Alfalfa hay 60 68 77 Suggested defaults 65 72 82

Design and Operation of Farm Irrigation Systems

269

The curve number is affected by the soil water content prior to the rainfall event, since soil water content affects the soil infiltration rate. Therefore, the CN is adjusted according to estimated soil water content prior to the rainfall event. This soil water content is termed the antecedent soil water condition or AWC. Adjustment ranges for CN were defined by SCS (1972) for dry (AWC I) and wet (AWC III) conditions. SCS defined AWC I as occurring when “watershed soils are dry enough for satisfactory plowing or cultivation to take place” and AWC III as when the “watershed is practically saturated from antecedent rains” (USDA-SCS, 1972, p. 4.10). AWC II is defined as the “average condition” and represents values in Table 8.14. Hawkins et al. (1985) expressed tabular relationships in USDA-SCS (1972) in the form of equations relating CN for AWC I and AWC III to CN for AWC II: CN II (8.89) CN I = 2.281 − 0.01281 CN II CN III =

CN II 0.427 + 0.00573 CN II

(8.90)

where CNI is the curve number associated with AWC I (dry) [0 - 100], CNII is the curve number associated with AWC II (average condition) [0 - 100], and CNIII is the curve number associated with AWC III (wet) [0 - 100]. The soil surface layer water balance associated with the dual Kc procedure (Equation 8.66) can be used to estimate the AWC condition. An approximation for the depletion of the soil surface layer at AWC III (wet) is when De = 0.5 REW, i.e., when the evaporation process is halfway through stage 1 drying. This point will normally be when approximately 5 mm or less have evaporated from the top 150 mm of soil since the time it was last completely wetted. Thus, the relationship: De- AWC

III

= 0 .5 REW

(8.91)

where De-AWC III is the depletion of the evaporative layer at AWC III. AWC I can be estimated to occur when 15 to 20 mm of water have evaporated from the top 150 mm of soil from the time it was last completely wetted. This is equivalent to when the evaporation layer has dried to the point at which De exceeds 30% of the total evaporable water in the surface layer beyond REW. This depletion amount is expressed as De = REW + 0.3(TEW – REW), where TEW is the total evaporable water in the surface layer. Therefore, De - AWC I = 0 .7 REW + 0.3 TEW (8.92) where TEW is the cumulative evaporation from the surface soil layer at the end of stage 2 drying. When De is in between these two extremes, i.e., 0.5 REW < De < 0.7 REW + 0.3 TEW, then the AWC is in the AWC II condition and the CN value is linearly interpolated between CNI and CNIII. In equation form: CN = CN III for De ≤ 0.5 REW CN = CN I

for De ≥ 0.7 REW + 0.3 TEW

(8.93) (8.94)

and, for 0.5 REW < De < REW + 0.3 (TEW – REW):

CN =

(De − 0.5 REW )CN I + (0.7 REW + 0.3 TEW − De )CN III 0.2 REW + 0.3 TEW

(8.95)

270

Chapter 8 Water Requirements

Equation 8.95 produces CNII when De is halfway between the endpoints of CNI and CNIII due to the symmetry of CNI and CNIII relative to CNII. 8.10.2 Infiltrated Precipitation Once the surface runoff depth is estimated using the curve number procedure, the depth of rainfall infiltrated is calculated as Pinf = P – RO (8.96) where Pinf = depth of infiltrated precipitation, mm P = measured precipitation depth, mm RO = depth of surface runoff, mm. If the soil will not hold the amount infiltrated, the remainder goes to deep percolation.

8.11 OTHER WATER REQUIREMENTS 8.11.1 Germination of Seeds Seeds for many crops are planted within only a few cm of the soil surface where the soil dries quickly, as estimated in Section 8.5.4. Therefore, frequent irrigation during the week following planting may be required in dry environments to increase plant viability and health. Sprinkler irrigation is especially suited to seed germination because small depths of water can be applied. Soil wetting by furrow irrigation is practiced in many areas, but more water is required than with sprinklers since water requires more time to move horizontally from the furrow to the ridge where the seed may be planted. Furthermore, salt tends to concentrate in the ridge by evaporation when saline water is used. Sprinkler systems for germination are normally solid set or center pivot systems that allow for daily or even more frequent wetting. Evapotranspiration requirements under conditions of high frequency irrigation are estimated using the Kc = Kcb + Ke procedure, where Kc typically approaches 1.2 to 1.3 for grass reference ETo and 1.0 for alfalfa reference ETr during the germination period. Evaporation from soil following irrigation reduces the warming of the soil, which may benefit seedling growth. 8.11.2 Climate Modification Irrigation can be used to both warm and cool the crop environment to increase yields. Evaporation during irrigation and from wet soil following irrigation cools the air boundary layer and improves growing conditions during periods of high air temperature for sensitive crops of fresh vegetables and fruits including peas, tomatoes, cucumbers, muskmelons, strawberries, apples, and grapes (Burman et al., 1980). Cooling is accomplished by conversion of sensible heat of the air and soil into latent heat of vaporization. All crops generally have foliar temperature thresholds above which photosynthesis is reduced due to stomatal closure or where a component of the carbon fixation process moves outside its kinetic window (Hatfield et al., 1987). For cool-season crops this occurs at temperatures in the neighborhood of 23° to 28° C and for warm-season crops at about 32° C (Hatfield et al., 1987; Keller and Bliesner, 1990). Foliar cooling requires high-intensity irrigation, where from two to six short applications are needed each hour. These intensities can only be practiced with automated fixed (solid set) sprinkler systems. Application rates are light, typically averaging 1 mm h-1 (Keller and Bliesner, 1990). Foliar cooling is of course more effective in arid climates where the low vapor pressure of the air and corresponding large vapor pressure deficit (es – ea) facilitates conversion of sensible heat into heat of vaporization. Misting to improve

Design and Operation of Farm Irrigation Systems

271

greenhouse environments is a common practice. Wolfe et al. (1976) and Griffin (1976) explored the use of sprinkler irrigation systems to cool orchard crops during early season warm periods in order to delay fruit bloom until later periods so as to reduce the risk of bloom freezing. Griffin (1976) and Griffin and Richardson (1979) developed mathematical models to estimate when irrigation bloom delay is effective. Their procedure was reviewed by Keller and Bliesner (1990). 8.11.3 Freeze Protection Both overhead and under-tree sprinkler systems are used for freeze protection. The goal of irrigation for freeze protection is to convert enough liquid irrigation water to solid (ice) that the latent heat of fusion (approximately 335 kJ kg-1) released during the freezing of irrigation water is sufficient to maintain air temperature at 0°C. Normally fruit and vegetables will not freeze at 0°C due to the presence of sugars and other molecules in the biomass. Because irrigation for freeze protection is practiced when air temperatures and wind speeds are low and generally during nighttime, evaporation rates are low. Therefore, most irrigation water added infiltrates the soil or runs off following melting. Further information is provided by Snyder et al. (2005). Design of sprinkler systems for frost control is covered in Keller and Bliesner (1990), Martin and Gilley (1993), and in Chapter 16. Minimum application rates are typically 2.5 to 3 mm h-1 (Keller and Bliesner, 1990). Larger applications are needed as the atmospheric dewpoint decreases below 0°C (Blanc et al., 1963; Keller and Bliesner, 1990). 8.11.4 Fertilizer Application Application of fertilizer by irrigation is often more economical than by other means. Generally, irrigation for fertilizer, herbicide, or pesticide application is only initiated when there is sufficient water deficit in the root zone to retain the applied water. Otherwise, some of the applied chemical will leach below the root zone. Therefore, the evapotranspiration requirement for fertilizer application is considered to be essentially zero, unless it requires more frequent irrigation than required for soil water replacement only. This topic is described in detail in Chapter 19. 8.11.5 Soil Temperature Soil temperatures can be markedly affected by irrigation water, through cooling by the water itself and by cooling during evaporation of water from wet soil following irrigation. Low irrigation water temperatures may depress soil temperatures and impede plant development. Soil cooling may be desirable under certain conditions, such as establishing seedling stands for head lettuce and during germination of grasses. 8.11.6 Dust Suppression Dust suppression, though not limited to agricultural fields, can be achieved using sprinkler systems. Feedlot dust can be generated in hot, dry climates when cattle become active in early evening and the air boundary layer is stable, impeding mixing. Dust is common around construction sites and on unpaved roads. In all cases, dust can be suppressed by sprinkling. The evaporation component can be estimated using the dual Kc procedure in Section 8.5.4. A reasonable approximation can be made using Kc ini from Figure 8.4a (light wetting events) where the wetting frequency and ETo rate are used to estimate the average Kc during each wetting period. 8.11.7 Leaching Requirements Leaching of salts from the crop root zone is an essential component of irrigation when the irrigation water carries salts. Generally, leaching of salts is accomplished

272

Chapter 8 Water Requirements

during irrigation to replace ET losses so that the ET requirement is not increased by leaching. However, in some arid climates having heavy soils, special leaching irrigations are necessary. In these situations, the annual consumptive irrigation water requirement is increased by the evaporation from wet soil during and following each leaching irrigation. Chapter 7 describes the estimation of leaching requirements.

8.12 EFFECTIVE RAINFALL Estimation of effective rainfall is necessary in calculating net and gross irrigation water requirements (Equations 8.1 and 8.2). In very arid climates, the majority of infiltrating rainfall is retained in the root zone and is consumed at some point as evaporation or transpiration. In more humid climates where rainfall totals are often greater than ET, portions of rainfall percolate below the root zone or runoff, thereby reducing the fraction that is effective in supplying crop ET. Evaporation of rainfall from wet soil is estimated using the Kcb + Ke procedure described in Section 8.5.4. The depth of effective rainfall is estimated by subtracting runoff estimated using the curve number (Section 8.10.1) or other infiltration estimation approach and subtracting deep percolation from the root zone computed by Equation 8.76. The soil water balance for the root zone is updated each day using Equation 8.73 or similar equation. Patwardhan et al. (1990) found that using a daily soil water balance equation (Equation 8.73) to estimate effectiveness of precipitation to be significantly more accurate than more simple and vague procedures such as the SCS monthly effective precipitation method (NRCS, 1993; Dastane, 1974), especially for poorly drained soils. The effectiveness of rainfall is influenced by irrigation, since irrigation prior to a rainfall event reduces the capacity of the soil to both infiltrate and retain the precipitation in the profile. The impact of irrigation on rainfall effectiveness is best determined by daily soil water balance. Application of the Kcb + Ke procedure provides an estimate of fraction of rainfall that evaporates with and without irrigation. The difference in DP and RO over the season with and without irrigation provides an indication of the relative impact of irrigation on retained precipitation. The daily soil water balance is the best method for estimating the carryover of winter precipitation into the growing period. However, the application is complicated by the impact of frozen soils on infiltration, by the estimation of crop coefficients and ETc during the dormant season, and by the challenge of estimating the energy and evaporation balance from any snow cover (Allen, 1996b; Wright, 1996). Chapter 11 of FAO56 provides guidance on estimating ET during winter and dormant seasons. There is some question on whether the evaporating compoenent of P (E from the soil surface) should be considered to be “effective.” This is especially true if estimating accumulation of nongrowing season P in the root zone. In this case, only the portion of P that infiltrates and does not evaporate from the surface evaporation layer will accrue over time. Similar arguments can be made in estimating effective Pduring the growing season. If the dual Kcb + Ke method is ued to estimate ET, then the evaporative component of P is accounted for in the daily soil water balance and the estimation for effective precipitation is more accurate.

8.13 DESIGN REQUIREMENTS Design and operation of irrigation systems is affected by both the peak irrigation water requirement and by the seasonal irrigation requirement. The seasonal irrigation water requirement dictates the annual operating time for the system and corresponding

Design and Operation of Farm Irrigation Systems

273

costs for labor, water, and energy. The peak water requirement dictates the minimum capacity for supply pipes, pumps, and canals to sustain potential crop growth. It also governs the maximum land area that can be adequately supplied by a sprinkler irrigation lateral. The peak water requirement decreases as the allowable interval between irrigations increases as discussed in the next section. 8.13.1 Estimating Peak ET Using Climatic Records Because evapotranspiration is driven by weather parameters, ET can vary significantly and randomly from day to day and from year to year. This is illustrated in Figure 8.10 where daily alfalfa reference ETr estimated using the ASCE-EWRI standardized Penman-Monteith equation (Equation 8.3) is shown for a 20-year period at Kimberly, Idaho. The variation in daily ETr is large. Overlain on the daily ETr are probability lines for various levels of nonexceedence, based on a normal probability distribution. Nonexceedence is defined as the value of ET that is not expected to be exceeded p% of the time, where p is the probability level. A normal probability distribution presumes that the coefficient of skewness (CS) = 0, where CS is the ratio of skew to the population mean. Generally, with evapotranspiration, CS approaches 0, so that frequency estimates based on the normal distribution are generally valid (Allen and Wright, 1983; Allen et al., 1983). In the case of the normal distribution, an estimate for ET for a specific probability of nonexceedance is estimated as: (8.97) ET Pn = ET mean + K Pn s where ETPn is the ET rate expected to be exceeded only 100% – p% of the time, KPn is a probability factor, and ETmean and s are the estimates for the mean and standard deviation of the underlying ET population. Generally, ETmean, s, and ETPn are computed for a specific time period during the growing season, for example, for the peak 30-day period. The ETmean is calculated as: ∑ ET (8.98) ET mean = n

Daily ETr, mm/day

14 12 98% 90%

10 8

95%

70% 50%

6

30%

4

5%

2%

10%

2 0

0

30

60

90

120

150

180

210

240

270

300

330

360

Day of year Figure 8.10. Distribution of daily calculated alfalfa reference ETr at Kimberly, Idaho, over a 20-year period and overlay of probability lines of nonexceedence.

274

Chapter 8 Water Requirements

Table 8.16. Probability factors (KPn) for a normal distribution. Probability of non2 5 10 25 50 75 80 90 95 98 exceedance, p (%) Probability of exceedance (%) 98 95 90 75 50 25 20 10 5 2 Return period 1.0 1.0 1.1 1.3 2 4 5 10 20 50 (years) KPn –2.054 –1.645 –1.28 –0.675 0.0 0.675 0.841 1.28 1.645 2.054

where ET is the ET for each day of the period, over all years of record available for analysis, and n is the number of observations. Usually, a minimum of five years of records are required to obtain viable estimations for ETmean, s, and ETPn. ET is generally estimated from weather records. The unbiased estimate for standard deviation is calculated as: 2⎤ ⎡ ∑(ET − ET mean ) s=⎢ ⎥ (n - 1 ) ⎥⎦ ⎢⎣

0.5

⎡ n ∑( ET 2 ) - (∑ ET )2 ⎤ =⎢ ⎥ n (n - 1 ) ⎣⎢ ⎦⎥

0.5

(8.99)

Table 8.16 lists KPn values for probabilities of interest in irrigation systems design and operation based on the normal distribution. When the population of ET can not be considered to follow a normal distribution, for example, when the coefficient of skewness is less than about –0.5 or greater than about 0.5, then a Pearson Type III distribution can be used to estimate ETPn: ETPn = ETmean + KPPIII s (8.100) where KPPIII is the probability factor for the Pearson Type III distribution. For cases where –1.0 < CS < 1.0, an expression from Wilson and Hilferty (1931) and USGS (1981) can be used for KPPIII : 3 ⎤ ⎤ CS ⎞ CS 2 ⎡⎢ ⎡⎛ + 1⎥ - 1⎥ ⎟ K PPIII = ⎢⎜ K Pn CS ⎢ ⎣⎝ 6 ⎠ 6 ⎥ ⎦ ⎣ ⎦

(8.101)

where KPn is from Table 8.16 for the normal distribution and CS is the coefficient of skewness: CS =

=

n ∑( ET - ET mean )3 (n - 1) (n - 2) s3 n

2

∑( ET 3) - 3 n ∑ ET ∑( ET 2) + 2 (∑ ET )3 n (n - 1) (n - 2) s3

(8.102)

When CS exceeds the limits for Equation 8.100, values for KPPIII can be taken from tables given in USGS (1981) or other texts on frequency analysis. Martin and Gilley (1993) described the application of the Weibull probability distribution. The length of the averaging period has substantial effect on the values for ETPn estimated for a specific probability of nonexceedence. The length of averaging period in irrigation systems design and operation should normally be based on the allowable number of days between irrigation events. This time length, known as the irrigation interval, Iint, is estimated as:

Design and Operation of Farm Irrigation Systems

I int =

275

RAW

(8.103)

ET c

where RAW is the depth of water that can be depleted with no stress to the plant, from Equation 8.72 and ET c is the average ETc during the period: i = d2

∑ ET ci

ET c =

i = d1

(8.104)

I int

where ETci is ETc on day i and d1 and d2 are beginning and ending days for the Iint averaging period, estimated as: ⎛ ⎞ I int (8.105) + 0.6 ⎟ d 1 = INT⎜ J 2 ⎝ ⎠ ⎛ ⎞ I int - 0.4 ⎟ d 2 = INT⎜ J + 2 ⎝ ⎠

(8.106)

Average ETr during period, mm/day

where INT( ) is the integer expression of the argument within the parentheses and J is the day of year number at the center of the period. Figure 8.11 shows values for ETPn for the peak June 21–July 20 period at Kimberly, Idaho between 1966 and 1985 as a function of the averaging period (i.e., irrigation interval). Alfalfa reference ETr is shown. ETPn is shown for both the normal and Pearson Type III frequency distributions, where differences between distributions are small. Coefficients of skewness for the data ranged from CS = –0.5 for Iint = 1 day to CS = –0.1 for Iint = 30 days. The comparison of ETPn computed by the two distributions indicates that the normal distribution is valid and representative for the Kimberly climate.

12 10

98%

8

75% 10%

6

95% 50% 5%

90% 25% 2%

4 2 0

0

10

20

30

Number of days in averaging period Figure 8.11. Expected alfalfa reference ET rate at specified levels of nonexceedence as a function of number of days in the averaging period for peak June 21–July 20 period, 1966–1985, at Kimberly, Idaho. Solid lines represent normal distributions and grey lines represent log Pearson type III distributions.

276

Chapter 8 Water Requirements

Figure 8.11 illustrates how ETPn decreases as the length of the averaging period (or irrigation interval) increases. For example, if an irrigation system design were based on probability of nonexceedance = 90% (10-year return period), the ETr used in design calculations at Kimberly would be 9.4 mm d-1 for crops having an irrigation interval of only 3 days. Design ETr would be 9.1 mm d-1 for crops having an irrigation interval of 7 days, and 8.6 mm d-1 for crops having an irrigation interval of 30 days. In other words, irrigation pipe, pump, and canal systems for the 30-day irrigation interval at Kimberly could be sized about 9% smaller than systems required for the 3-day irrigation interval (assuming that crop coefficients were the same). A 3-day interval would represent shallow rooted crops (Zr ~ 0.6 m) on coarse soils. The 30-day interval would represent deep-rooted crops (Zr ~ 2.5 m) on medium-textured soils. 8.13.2 General Design Curves Figure 8.11 illustrates the variation in ETPn with length of averaging period for the southern Idaho climate, which is characterized as semiarid with relatively cloudless days during the peak monthly period. When climates have more variable cloudiness, the variation in day-to-day ET increases since solar radiation is a primary driver of evapotranspiration. Under variable cloudiness conditions, the likelihood of experiencing a period with cloud-free weather and corresponding high ET increases as the length of averaging period decreases. Therefore, the ratio of peak ET during short averaging periods relative to mean monthly ET increases as mean cloudiness of a location increases. This effect is illustrated in Figure 8.12 (adapted from Doorenbos and Pruitt, 1977), where the ratio of peak ET relative to mean monthly ET is shown to increase with increasing humidity and cloudiness. The peak ET estimated by Figure 8.12 is for the 75% probability level of nonexceedence (i.e., the ET level that is exceeded in 2.5 years out of every 10). The curves in Figure 8.12 are expressed as a function of net application depth. Net application depth (Dn) is related to time length between irrigations, in days, as Dn = Iint ET c and where Dn is equivalent to the numerator in Equation 8.103.

Mean Monthly ET

Peak ET (75% Prob.)

1.4 1.3 1.2 1.1

4

4 Mid-continental subhumid to humid climates with highly variable cloudiness during peak month. 3 Mid-continental climates with variable cloudiness and peak monthly ET = 8 mm/day. 2 Mid-continental climates with variable cloudiness and peak monthly ET = 5 mm/day. 1 Arid and semiarid climates with predominantly clear weather during peak month.

3 2 1

1.0 0

40

80

120

160

200

Net Application Depth, mm Figure 8.12. Expected ratio of peak design ET to mean monthly ET of the peak month for a 75% level of nonexceedence for four types of climates (after Doorenbos and Pruitt, 1977).

Design and Operation of Farm Irrigation Systems

277

Figure 8.11 follows Curve 1 in Figure 8.12 relatively closely, where Curve 1 represents arid and semiarid climates with predominantly clear weather. The four curves in Figure 8.12 can be reproduced mathematically by: ET peak ET month

= (1.18 + 0.2 N cl ) D n( - 0.037 - 0.02 N cl )

(8.107)

where Ncl is the number of the curve in Figure 8.12 [1 - 4] and Dn is the net application depth, mm. The standard error of estimate for the ratio estimated by Equation 8.107 is 0.012 relative to the original curves of Doorenbos and Pruitt (1977). Equation 8.107 should be applied with limits of Dn > 20 mm. 8.13.3 Population Statistics from ET Estimation Methods ET estimation equations, such as the Penman-Monteith, Penman, and Hargreaves equations described in this chapter, may not reproduce the same variation in ET as is found in the actual, underlying ET population. In general, the more weather parameters contained in the ET equation, the larger will be the variation in ET estimates. Therefore, in applications where weather data are represented by long-term averages, for example wind speed, variation among ET estimates is reduced. Deviation in statistics from the underlying population, as represented by standard deviation, s, and coefficient of skewness, CS, will impact the estimation of ETPn or ETPPIII and subsequent irrigation design and operation. This behavior was discussed in more detail by Allen et al. (1983) and Allen and Wright (1983).

8.14 ANNUAL IRRIGATION WATER REQUIREMENTS Annual evapotranspiration requirements can be calculated by computing ETref on daily, weekly, 10-day, or monthly timesteps and applying a crop coefficient over each period. When calculations are performed on a daily basis, the Kcb + Ke procedure can be utilized to improve accuracy. When weekly or longer timesteps are used, a timeaveraged Kc is applied to ETref. Annual irrigation water requirements are estimated by summing the calculated ETc over the year and subtracting effective rainfall. Gross annual water requirements require division by a uniformity term (consumed fraction of applied water) and potentially a leaching factor (see Equation 8.2). Ordinarily, ET and irrigation water requirements are needed for the growing season only. The soil water stored in the soil profile at the start of the season is estimated by operating the soil water balance over part or all of the winter period. In winter climates with frozen soils, this can be challenging. Often, in irrigated agriculture, the soil profile is recharged to near field capacity by the start of the growing season. In climates where the growing season is year-round, continuity between ending soil water content from one crop to the beginning soil water content of a subsequent crop should be considered to adequately account for the total water requirement. Acknowledgements The ASCE-EWRI, with support by the Irrigation Association, is commended for the work on standardization of the calculation of reference evapotranspiration that provided the means for beginning the standardization of crop coefficients. Work by the Irrigation Association Water Management committee has helped shape the structure for the landscape ET coefficients. The authors appreciate and acknowledge input and advice on agricultural crop coefficients and landscape coefficients by Dr. R. L. Snyder of the University of California, Davis.

278

Chapter 8 Water Requirements

LIST OF SYMBOLS BR Cd Cn CF CN CRi CS De,j-1 DM Dr DoY DPei,j Ej ET ETc ETc act ETL ETmean ETmonth ETo ETpeak ETr ETref G Gsc GW Ij Iint J Jmonth IR Kc Kc act Kcb Kcb adj Kcb full Kd Ke Kc end Kc ini Kc m Kc max Kc mid

Bowen ratio denominator coefficient for the ASCE standardized PM equation, s m-1 numerator coefficient for the ASCE standardized PM equation; K mm s3 Mg-1 d-1, K mm s3 Mg-1 h-1 consumed fraction of applied water curve number capillary rise from the groundwater table on day i, mm coefficient of skewness cumulative depletion from the soil surface layer at the end of day j – 1, mm day of month (1–31) cumulative depletion from the root zone, including De, mm day of year deep percolation from the few fraction of the soil surface layer on day j, mm evaporation on day j, mm evapotranspiration rate; mm d-1, mm h-1 ET from a particular crop; mm d-1, mm h-1 actual evapotranspiration rate; mm d-1, mm h-1 target ET from a particular landscape vegetation; mm d-1, mm h-1 mean ET of the underlying population; mm d-1 mean ET during a month, mm d-1 ET from a well-watered grass reference crop; mm d-1, mm h-1 mean ET during peak ET period, mm d-1 ET from a well-watered alfalfa reference crop; mm d-1, mm h-1 reference evapotranspiration, general; mm d-1, mm h-1 heat flux density to the ground; MJ m-2 d-1, MJ m-2 h-1, W m-2 solar constant, MJ m-2 h-1 contributions of shallow ground water; mm d-1, mm h-1 irrigation depth on day j that infiltrates the soil, mm irrigation interval, d day of the year day of the year for the middle of the month irrigation water requirement; mm d-1, mm season-1 crop coefficient general crop coefficient under actual field conditions crop coefficient (basal), soil water not limiting transpiration, but the soil surface is visually dry basal Kcb (or Ksp pot) during the midseason when plant density and/or leaf area are at or lower than full cover basal Kcb anticipated for vegetation under full cover conditions vegetation density factor evaporation coefficient Kc at the end of the late season period average Kc during the initial period mean crop coefficient maximum value of Kc following rain or irrigation average Kc during the midseason period

Design and Operation of Farm Irrigation Systems

279

minimum Kc for dry bare soil with no ground cover Kc min Kco crop coefficient based on grass (ETo) reference Kcr crop coefficient based on alfalfa (ETr) reference Kd vegetation density coefficient KL target landscape coefficient KL act actual KL anticipated from the landscape under water availability Kmc microclimate factor Ko offset between mean daily Td and daily Tmin, °C KPn probability factor KPPIII probability factor for the Pearson Type III distribution Kr evaporation reduction coefficient Ks adjustment coefficient for water stress Ksm coefficient for managed water stress Kv vegetation species factor Ksp pot potential vegetation species factor under fully watered conditions Kt proportion of basal ET extracted as transpiration from surface soil layer Kw units parameter in the Penman equation; mm d-1kPa-1, mm h-1kPa-1 Ldev length of development period of crop growing season, d Lini length of initial period of crop growing season, d Lmid length of midseason period of crop growing season, d Llate length of late season period of crop growing season, d Lroot growth length of period of root growth, d Lm longitude of the solar radiation measurement site, degrees west of Greenwich Lz longitude of the center of the local time zone, degrees west of Greenwich LAI leaf area index LE latent heat energy (ET) per unit area; MJ m-2 d-1, MJ m-2 h-1, W m-2 LR leaching requirement M month number (1–12) P atmospheric pressure, kPa P precipitation, mm Pe effective precipitation (entering and remaining in the root zone), mm Pinf depth of infiltrated precipitation, mm Ra exoatmospheric solar radiation; MJ m-2 d-1, MJ m-2 h-1, W m-2 Rn net radiation; MJ m-2 d-1, MJ m-2 h-1, W m-2 Rnl net long-wave radiation; MJ m-2 d-1, MJ m-2 h-1, W m-2 Rns net solar radiation; MJ m-2 d-1, MJ m-2 h-1, W m-2 Rs solar radiation at the surface, MJ m-2 d-1, MJ m-2 h-1, W m-2 Rso solar radiation for cloudless sky; MJ m-2 d-1, MJ m-2 h-1, W m-2 RAW readily available water in the root zone, mm REW readily evaporable water from the surface soil layer, mm RH relative humidity, % RHmax maximum daily relative humidity, % RHmean mean daily or hourly relative humidity, % RHmin minimum daily relative humidity, % RO surface runoff, mm S maximum depth of water in CN procedure that can be retained as infiltration, mm

280

Sc T Td Tdry Tei,j TKmax TKmin Tmax Tmean Tmin Twet TAW TEW Y Ze Zr a, b apsy aw bw dr ea es eo(T) fc fcd fc eff few fw h n p s t tl uz u2 z zs zw

α

β

γ γpsy

Δ Δθ

δ

Chapter 8 Water Requirements

seasonal correction for solar time, h mean daily or hourly air temperature, °C dewpoint temperature of the air, °C dry bulb temperature of the air, °C depth of transpiration from the few fraction of the soil surface on day j, mm maximum daily air temperature, K minimum daily air temperature, K maximum daily air temperature, °C mean daily or hourly temperature, °C minimum daily temperature, °C wet bulb temperature of the air, °C total available water in the root zone, mm total evaporable water that can be evaporated from the surface soil layer, mm year (2007, for example) effective depth of the surface soil subject to drying by way of evaporation, m maximum effective rooting depth, m constants; see usage ventilation coefficient for psychrometric constant empirical coefficient in Penman wind function empirical coefficient in Penman wind function, s m-1 inverse relative distance factor (squared) for the earth-sun mean actual vapor pressure, kPa saturation vapor pressure, kPa saturation vapor pressure function at temperature T, kPa fraction of the soil surface covered by vegetation cloudiness function effective fraction of the soil surface effectively covered by vegetation fraction of the soil that is both exposed to solar radiation and that is wetted fraction of the soil surface wetted by irrigation and/or precipitation height of vegetation, m number of observations soil water depletion fraction for no stress standard deviation of the underlying ET population, mm d-1 time, s, h, d time length of the calculation period, h horizontal wind speed at height z, m s-1 horizontal wind speed at 2 m above the ground surface, m s-1 elevation, m depth of soil experiencing the change in water content, mm, m height of anemometer above ground surface, m shortwave reflectance coefficient or albedo angle of the sun above the horizon, radians γ = 0.000665 P, Pa °C-1 psychrometric constant for psychrometers, kPa °C-1 slope of the saturation vapor pressure-temperature curve, de/dT, kPa °C-1 change in soil water content in the root zone during period of calculation solar declination, radians

Design and Operation of Farm Irrigation Systems

θ θFC θWP ϕ Φ

ω ωs ω1 ω2

281

volumetric soil water content volumetric soil water content at field capacity volumetric soil water content at wilting point station latitude, radians Stefan-Boltzmann constant, kJ m-2 h-1 K-4, MJ m-2 d-1 K-4 solar time angle at midpoint of hourly or shorter period, radians sunset hour angle, radians solar hour angle at beginning of hourly or shorter period, radians solar hour angle at end of hourly or shorter period, radians

REFERENCES Aboukhaled, A., A. Alfaro, and M. Smith. 1982. Lysimeters. Irrigation and Drainage Paper 39. Rome, Italy: FAO. Allen, R. G. 1988. IRRiSKED: Irrigation scheduling program for demand and rotation scheduling: Logan, Utah: Dept. Biological and Irrigation Engineering, Utah State Univ. Allen, R. G. 1996a. Assessing integrity of weather data for use in reference evapotranspiration estimation. J. Irrig. Drain. Eng. 122(2):97-106. Allen, R.G. 1996b. Nongrowing season evaporation in Northern Utah. In Proc. North American Water and Environment Congress, ASCE. Reston, Va.: American Soc. Civil Engineers. Allen, R. G., C. E. Brockway, and J. L. Wright. 1983. Weather station siting and consumptive use estimates. J. Water Resources Plan. and Man. Div., ASCE 109(2): 134-146. Allen, R. G., A. J. Clemmens, C. M. Burt, K. Solomon, and T. O’Halloran. 2005. Prediction accuracy for project-wide evapotranspiration using crop coefficients and reference evapotranspiration. J. Irrig. Drain. Eng. 131(1): 24-36. Allen, R. G., and F. N. Gichuki. 1989. Effects of projected CO2-induced climate changes on irrigation water requirements in the Great Plains states (Texas, Oklahoma, Kansas and Nebraska). In The Potential Effects of Global Climate Change on the United States: Appendix C-Agriculture. EPA-230-05-89-053. Washington, D.C.: U.S. Environmental Protection Agency, Office of Policy, Planning and Eval. Allen, R. G., T. A. Howell, W. O. Pruitt, I. A. Walter, and M. E. Jensen, eds. 1991a. Lysimeters for Evapotranspiration and Environmental Measurements. New York, N.Y.: American Soc. Civil Engineers. Allen, R. G., M. E. Jensen, J. L. Wright and R. D. Burman. 1989. Operational estimates of reference evapotranspiration. Agronomy J. 81: 650-662. Allen, R. G., L. S. Pereira, D. Raes, and M. Smith. 1998.Crop Evapotranspiration: Guidelines for Computing Crop Water Requirements. United Nations FAO, Irrigation and Drainage Paper 56. Rome, Italy: FAO. Allen, R. G., L. S. Pereira, M. Smith, D. Raes, and J. L. Wright. 2005a. FAO-56 dual crop coefficient method for estimating evaporation from soil and application extensions. J. Irrig. Drain. Eng. 131(1): 2-13. Allen, R. G., W. O. Pruitt, J. A. Businger, L. J. Fritschen, M. E. Jensen, and F. H. Quinn 1996. Chapt. 4: Evaporation and transpiration. In ASCE Handbook of Hydrology. Wootton et al., eds. New York, N.Y.: American Soc. Civil Engineers. Allen, R. G., W. O. Pruitt, and M. E. Jensen. 1991b. Environmental requirements of lysimeters. In Lysimeters for Evapotranspiration and Environmental Measure-

282

Chapter 8 Water Requirements

ments, 170-181. R. G. Allen, T. A. Howell, W. O. Pruitt, I. A. Walter, and M. E. Jensen, eds. New York, N.Y.: American Soc. Civil Engineers. Allen, R. G., W. O. Pruitt, D. Raes, M. Smith, and L. S. Pereira. 2005b. Estimating evaporation from bare soil and the crop coefficient for the initial period using common soils information. J. Irrig. Drain. Eng. 131(1): 14-23. Allen, R. G., W. O. Pruitt, J. L. Wright, T. A. Howell, F. Ventura, R. Snyder, D. Itenfisu, P. Steduto, J. Berengena, J. Baselga Yrisarry, M. Smith, L. S. Pereira, D. Raes, A. Perrier, I. Alves, I. Walter, and R. Elliott. 2006. A recommendation on standardized surface resistance for hourly calculation of reference ETo by the FAO56 Penman-Monteith method. Agric. Water Mgmt. Allen, R.G. and C.W. Robison. 2007 Evapotranspiration and Irrigation Water Requirements for Crops in Idaho. Research Completion Report, Univ. Idaho submitted to Idaho Dept. Water Resources. 277 pp. Allen, R. G., and J. L. Wright. 1983. Variation within the measured and estimated consumptive use requirements. In Proc. of the 1983 Specialty Conference, Irrigation and Drainage Division, 1-12. Jackson Hole, Wyo.: American Soc. Civil Engineers. Allen, R. G., and J. L. Wright. 1997. Translating wind measurements from weather stations to agricultural crops. J. Hydrol. Eng. ASCE 2(1): 26-35. Allen, R. G., and J. L. Wright. 2007. Conversion of Kimberly, Idaho crop coefficients to the ASCE standardized Penman-Monteith reference evapotranspiration basis. Inhouse report, Univ. Idaho, Kimberly, Idaho. ASCE-EWRI. 2005. The ASCE Standardized Reference Evapotranspiration Equation. Report 0-7844-0805-X, ASCE Task Committee on Standardization of Reference Evapotranspiration. Reston, Va.: American Soc. Civil Engineers. Ayars, J. E., R. S. Johnson, C. J. Phene, T. J. Trout, D. A. Clark, and R. M. Mead. 2003. Water use by drip-irrigated late-season peaches. Irrig. Sci. 22: 187-194. Battikhi, A. M., and R. W. Hill. 1986a. Irrigation scheduling and watermelon yield model for the Jordan Valley. J. Agronomy Crop Sci. 157: 145-155. Battikhi, A. M., and R. W. Hill. 1986b. Irrigation scheduling and cantaloupe yield model for the Jordan Valley. Agric. Water Mgmt. 15: 177-187. Blad, B. L., and N. J. Rosenberg. 1974. Lysimetric calibration of the Bowen-ratio energy balance method for evapotranspiration estimation in the Central Great plains. J. App. Meteorol. 13(2): 227-236. Blanc, M. L., H. Geslin, I. A. Holzberg, and B. Mason. 1963. Protection against frost damage. Tech. Note No. 51. Geneva, Switzerland: World Meteorol. Org. Bonachela, S., F. Orgaz, F. J. Villalobos, and E. Fereres. 2001. Soil evaporation from drip-irrigated olive orchards. Irrig. Sci. 20(2): 65-71. Borg, H., and D. W. Grimes. 1986. Depth development of roots with time: An empirical description. Trans. ASAE 29: 194-197. Bosen, J. F. 1958. An approximation formula to compute relative humidity from dry bulb and dew point temperatures. Monthly Weather Rev. 86(12): 486. Brutsaert, W. H. 1982. Evaporation into the Atmosphere. Dordrecht, Holland: R. Deidel Publ. Co. Brunt, D. 1932. Notes on radiation In the atmosphere: I. Quart. J. Roy. Meteorol. Soc. 58: 389-420. Brunt, D. 1952. Physical and Dynamical Meteorology. 2nd ed. Cambridge, UK: Cambridge Univ. Press.

Design and Operation of Farm Irrigation Systems

283

Burman, R. D., P. R. Nixon, J. L. Wright, and W. O. Pruitt. 1980. Water requirements. In Design and Operation of Farm Irrigation Systems, 189-232. M. E. Jensen, ed. St. Joseph, Mich.: ASAE. Burman, R. D., M. E. Jensen, and R. G. Allen. 1987. Thermodynamic factors in evapotranspiration. In Proc. Irrig. and Drain. Spec. Conf., ASCE, 28-30. L. G. James, and M. J. English, eds. Reston, Va.: American Soc. Civil Engineers. Businger, J. A., M. Miyake, A. J. Dyer, and E. F. Bradley. 1967. On direct determination of turbulent heat flux near the ground. J. Appl. Meteorol. 6(6): 1025-1032. Campbell, G. S., and B. D. Tanner. 1985. A krypton hygrometer for measurement of atmospheric water vapor concentration. In Moisture and Humidity, 609-612. Research Triangle Park, N.C.: ISA. Costello, L. R., N. P. Matheny, and J. R. Clark. 2000. Estimating the irrigation water needs of landscape plantings in California: Part 1: The landscape coefficient method. Univ. Calif. Coop. Ext. and Calif. Dept. Water Resources. Costello, L. R., and K. S. Jones. 1999. Estimating the irrigation water needs of landscape plantings in California. Part 2:WUCOLS III-Water use classification of landscape species. Danuso F., M. Gani, and R. Giovanardi. 1995. Field water balance: BIdriCo 2. In Crop Water models in Practice, 49-73. L. S. Pereira, B. J. Van den Broek, P. Kabat, and R. G. Allen, eds. Wageningen, The Netherlands: Wageningen Press. Dastane, N. G. 1974. Effective rainfall in irrigated agriculture. FAO Irrig. and Drainage Paper No. 25. Rome, Italy: United Nations Food and Agr. Organization. de Azevedo, P. V., B. B. da Silva, and V. P. R. da Silva. 2003. Water requirements of irrigated mango orchards in northeast Brazil. Agric. Water Man. 58: 241-254. Dolman, A. J. 1993. A multiple-source land surface energy balance model for use in general circulation models. Agric. For. Meteorol. 65: 21-45. Doorenbos, J., and A. H. Kassam. 1979. Yield response to water. Irrig. and Drainage Paper No. 33, (rev.) Rome, Italy: FAO. Doorenbos, J., and W. O. Pruitt. 1977. Crop water requirements. Irrig. and Drainage Paper No. 24, (rev.) Rome, Italy: FAO. Droogers, P., and R. G. Allen. 2002. Estimating reference evapotranspiration under inaccurate data conditions. Irrig. Drain. Systems (16): 33-45 Dyer, A. J. 1974. A review of flux-profile relationships. Boundary Layer Meterol. 7: 363-372. Everson, D. O., M. Faubion, and D. E. Amos. 1978. Freezing temperatures and growing seasons in Idaho. Univ. Idaho Agric. Exp. Sta. Bull. 494. Fereres, E. 1981. Drip irrigation management. Leaflet No. 21259. Berkeley, Calif.: Cooperative Ext., Univ. California. Gavilan, P., I. J. Lorite, S. Tornero, and J. Berengena. 2005. Regional calibration of Hargreaves equation for estimating reference ET in a semiarid environment. Agric. Water Mgmt. 81(3): 257–281. Girona, J., M. Gelly, M. Mata, A. Arbones, J. Rufat, and J. Marsal. 2005. Peach tree response to single and combined deficit irrigation regimes in deep soils. Agric. Water Man. 72:97-108. Grattan, S. R., W. Bowers, A. Dong, R. L. Snyder, J. J. Carroll, and W. George. 1998. New crop coefficients estimate water use of vegetables, row crops. Calif. Agric. 52(1): 16-21.

284

Chapter 8 Water Requirements

Grebet, P., and R. H. Cuenca. 1991. History of lysimeter design and effects of environmental disturbances. Lysimeters for Evapotranspiration and Environmental Measurements, 10-18. R. G. Allen, T. A. Howell, W. O. Pruitt, I. A. Walter, and M. E. Jensen, eds. New York, N.Y.: American Soc. Civil Engineers. Griffin, R. E. 1976. Micro-climate control of deciduous fruit production with overhead sprinklers. In Reducing Fruit Losses Caused by Low Spring Temperatures, Appendix F. Document No. 10550101. Final report of Utah Agric. Exp. Sta. to Four Corners Regional Commission. Logan, Utah: Utah State Univ. Griffin, R. E., and E. A. Richardson. 1979. Sprinklers for micro climate cooling of bud development. In Modfication of the Aerial Environment of Plants, 441-455. B. J. Barfield, and J. Gerber, eds. St. Joseph, Mich.: ASAE. Haddadin, S. H., and I. Ghawi. 1983. Effect of plastic mulches on soil water conservation and soil temperature in field grown tomato in the Jordan Valley. Dirasat 13(8): 25-34. Hargreaves, G. L., G. H. Hargreaves, and J. P. Riley. 1985. Agricultural benefits for Senegal River Basin. J. Irrig. Drain. Eng. 111: 113-124. Hargreaves, G. H., and Z. A. Samani. 1982. Estimating potential evapotranspiration. Tech. Note. J. Irrig. Drain. Eng. 108(3): 225-230. Hargreaves, G. H., and Z. A. Samani. 1985. Reference crop evapotranspiration from temperature. Appl. Eng. Agric. 1(2): 96-99. Hargreaves, G. H., and R. G. Allen. 2003. History and evaluation of the Hargreaves evapotranspiration equation. J. Irrig. Drain. Eng. 129(1): 53-63. Hatfield, J. L., J. J. Burke, J. R. Mahan, and D. F. Wanjura. 1987. Foliage temperature measurements: A link between the biological and physical environment. In Proc. International Conf. on Measurement of Soil and Plant Water Status, 2: 99-102. Logan, Utah: Utah State Univ. Hatfield, J. L., and M. Fuchs. 1990. Chapt. 3: Evapotranspiration models. In Management of Farm Irrigation Systems. ASAE Monograph 33-59.St. Joseph, Mich.: ASAE. Hawkins, R. H., A. T. Hjelmfelt, and A. W. Zevenbergen. 1985. Runoff probability, storm depth, and curve numbers. J. Irrig. Drain. Eng. 111(4): 330-340. Hernandez-Suarez, M. 1988. Modeling irrigation scheduling and its components and optimization of water delivery scheduling with dynamic programming and stochastic ETo data. Ph.D. diss. Davis, Calif.: Univ. California, Davis. Howell, T. A., A. D. Schneider, and M. E. Jensen. 1991. History of lysimeter design and use for evapotranspiration measurements. In Lysimeters for Evapotranspiration and Environmental Measurements, 1-9. R. G. Allen, T. A. Howell, W. O. Pruitt, I. A. Walter, and M. E. Jensen, eds. New York, N.Y: American Soc. Civil Engineers. Huntingford, C., S. J. Allen, and R. J. Harding 1995, An intercomparsion of single and dual-source vegetation-atmosphere transfer models applied to transpiration from Sahelian savannah. Boundary-Layer Meteorology 74: 397-418. Irmak, S., T. A. Howell, R. G. Allen, J. O. Payero, and D. L. Martin. 2005. Standardized ASCE Penman-Monteith: Impact of sum-of-hourly vs. 24-hour timestep computations at reference weather station sites. Trans. ASAE 48(3): 1063-1077. Irrigation Association. 2003. Landscape Irrigation Scheduling and Water Management: Practices Guidelines. Report by Water Management Committee. J. McCabe, J. Ossa, R. G. Allen, B. Carleton, B. Carruthers, C. Corcos, T. A. Howell, R. Marlow, B. Mecham, and T. L. Spofford, eds. Itenfisu, D. R. L. Elliott, R. G. Allen, and I. A. Walter. 2003. Comparison of reference

Design and Operation of Farm Irrigation Systems

285

evapotranspiration calculations as a part of the ASCE standardization effort. J. Irrig. Drain. Eng. 129(6): 440-448. Jensen, M. E., ed. 1974. Consumptive use of water and irrigation water requirements. Irrig. Drain. Div. Report, American Soc. Civil Engineers. Jensen, M. E., R. D. Burman, R. G. Allen, eds. 1990. Evapotranspiration and Irrigation Water Requirements. ASCE Manual 70. Reston, Va.: American Soc. Civil Engineers. Johnson, R. S., L. E. Williams, J. E. Ayars, and T. J. Trout. 2005. Weighing lysimeters aid study of water relations in tree and vine crops. Calif. Agric. 59(2):133-136. Jones, J. W., K. J. Boote, S. S. Jagtap, G. G. Wilderson, G. Hoogenboom, and J. W. Mishoe. 1987. SOYGRO V5.4 Technical Documentation. Gainesville, Fla.: Agric. Eng. Dept. Res. Report, Univ. of Florida. Jones, C. A., W. L. Bland, J. T. Ritchie, and J. R. Williams. 1991. Simulation of root growth. In Modeling Plant and Soil Systems, 91-123. Monograph 31. J. Hanks and J. T. Ritchie, eds. Madison, Wis.: American Soc. Agronomy. Keller, J., and R. D. Bliesner. 1990. Sprinkle and Trickle Irrigation. New York, N.Y.: van Nostrand Reinhold. Kruse, E. G., and R. H. Haise. 1974. Water use by native grasses in high altitude Colorado meadows. Western Region report ARS-W-6-1974. Washington, D.C.: USDA-ARS. Ley, T. W., R. G. Allen, and R. W. Hill. 1996. Weather station siting effects on reference evapotranspiration. In Proc. ASAE International Conference on Evapotranspiration and Irrigation Scheduling, 727-734. St. Joseph, Mich.: ASAE. Liu Y., L. S. Pereira, and R. M. Fernando. 2006. Fluxes through the bottom boundary of the root zone in silty soils: Parametric approaches to estimate groundwater contribution and percolation. Agric. Water Man. 84: 27-40. Martin, D., and J. Gilley. 1993. Chapt. 2: Irrigation water requirements. In National Engineering Handbook, Part 623. Natural Resources Conservation Service. Martınez-Cob, A., and M. Tejero-Juste. 2004. A wind-based qualitative calibration of the Hargreaves ETo estimation equation in semiarid regions. Agric. Water Manage. 64: 251-264. McQueen, J. 2004. Air Quality Forecast Model Evaluation Study during ICARTT/NEAQS-2004, NOAA/NCEP: CMAQ-ETA: 301-763-8000 x7226. At: www.al.noaa.gov/ICARTT/StudyCoordination/ICARTTdocs/Plan2k4_AQFM.pdf. Meyer, W. S., and L. Mateos. 1990. Effects of soil type on soybean crop water use in weighing lysimeters. II: Effect of lysimeter canopy height discontinuity on evaporation. Irrig. Sci. 11: 233-237. Munger, J. W., and H. W. Loescher. 2004. Guidelines for making eddy covariance flux measurements. At: public.ornl.gov/ameriflux/measurement_standards_4.doc. National Resouce Conservation Service. 1993. Part 623.0207: Effective precipitation. In National Engineering Handbook, 2.142-2.154. Neale, C. M. U., W. C. Bausch, and D. E. Heermann. 1989. Development of reflectance based crop coefficients for corn. Trans. ASAE 32(6): 1891-1899. Pastor, M., and F. Orgaz. 1994. Los programas de recorte de riego en olivar. Agricultura 746: 768-776. Patwardhan, A. S., J. L. Nieber, and E. L. Johns. 1990. Effective rainfall estimation methods. J. Irrig. Drain. Eng. 116(2): 182-193. Paço, T. A., M.I. Ferreira, and N. Conceição. 2006. Peach orchard evapotranspiration in a sandy soil: Comparison between eddy covariance measurements and estimates

286

Chapter 8 Water Requirements

by the FAO 56 approach. Agric. Water Man. 85(3):305-313. Payero, J. O., C. M. U. Neale, J. L. Wright, and R. G. Allen. 2003. Guidelines for validating Bowen ratio data. Trans. ASAE 46(4): 1051-1060. Penman, H. L. 1948. Natural evaporation from open water, bare soil and grass. Proc. Roy.Soc. London A193: 120-146. Penman, H. L. 1963. Vegetation and Hydrology. Tech. Comm. No. 53. Harpenden, England: Commonwealth Bureau of Soils. Pereira L.S., A. Perrier, R. G. Allen, and I. Alves. 1999. Evapotranspiration: Review of concepts and future trends. J. Irrig. Drain. Eng. 125(2): 45-51. Pruitt, W. O. 1986. Traditional methods: Evapotranspiration research priorities for the next decade. ASAE Paper No. 86-2629. St. Joseph, Mich.: ASAE. Pruitt, W. O. 1991. Development of crop coefficients using lysimeters. Lysimeters for Evapotranspiration and Environmental Measurements, Proc. of the ASCE Int. Symp. on Lysimetry, 182-190. R. G. Allen, T. A. Howell, W. O. Pruitt, I. A. Walter, and M. E. Jensen, eds., New York, N.Y.: American Soc. Civil Engineers. Pruitt, W. O., and F. J. Lourence. 1985. Experiences in lysimetry for ET and surface drag measurements. In Advances in Evapotranspiration, 51-69. St. Joseph, Mich.: ASAE. Pruitt, W. O., E. Fereres, P. E. Martin, H. Singh, D. W. Henderson, R. M. Hagan, E. Tarantino, and B. Chandio. 1984. Microclimate, evapotranspiration, and water-use efficiency for drip- and furrow-irrigated tomatoes. In Proc. 12th Congress, International Commission on Irrigation and Drainage, 367-394. Pruitt, W. O., B. D. Swann, A. Held, B. Sutton, A. Matista, and T. C. Hsiao. 1987. Bowen ratio and Penman: Australian-California tests. In Irrigation Systems for the 21st Century, Proc. of the ASCE Nat’l Conf. Irrig. and Drainage Engineering, 149158. L. G. James and M. J. English, eds. American Soc. Civil Engineers Ritchie, J. T. 1972. Model for predicting evaporation from a row crop with incomplete cover. Water Resources Res. 8: 1204-1213. Ritchie, J. T., and D. S. NeSmith. 1991. Chapter 2: Temperature and crop development. In Modeling Plant and Soil Systems, 5-29. R. J. Hanks and J. T. Ritchie, eds. Agronomy Series No. 31. Madison, Wis.: American Soc. Agron. Ritchie, J. T., and S. Otter. 1985. Description and performance of CERES-Wheat: A user-oriented wheat yield model. In ARS Wheat Yield Project, 159-175. ARS-38. Springfield, Va.: National Tech. Info. Serv. Richie, W. E., and D. R. Pittenger. 2000. Mixed landscape irrigation research findings. Proc. UCR Turfgrass and Landscape Management Research Conf., 12-13. Rogers, J. S., L. H. Allen, and D. J. Calvert. 1983. Evapotranspiration for humid regions: Developing citrus grove, grass cover. Trans. ASAE 26(6): 1778-83, 92. Safadi, A. S. 1991. Squash and cucumber yield and water use models. Unpublished Ph.D. diss. Logan, Utah: Dept. Biological and Irrigation Engineering, Utah State Univ. Sammis, T. W., Andales, A., and L. Simmons. 2004. Adjustment of closed canopy crop coefficients of pecans for open canopy orchards. 38th Western Irrigation Pecan Conference Proceedings, 117-119. Sellers, W. D. 1965. Physical Climatology. Chigago, Ill.: Univ. Chicago Press. Shaw, D. A. and D. R. Pittenger. 2004. Performance of landscape ornamentals given irrigation treatments based on reference evapotranspiration. Acta Hort. 664: 607-613. Shuttleworth, W. J., and J. S. Wallace. 1985. Evaporation from sparse crops: An energy combination theory. Quart. J. Roy. Meteorol. Soc. 111: 839-853.

Design and Operation of Farm Irrigation Systems

287

Sinclair, T. R., L. H. Allen, and E. R. Lemon. 1975. An analysis of errors in the calculation of energy flux densities above vegetation by a Bowen ratio profile method. Boundary-Layer Meteorol. 8: 129-139. Smith, M., R. G. Allen, J. L. Monteith, A. Perrier, L. S. Pereira, and A. Segeren. 1991. Report of the expert consultation on procedures for revision of FAO guidelines for prediction of crop water requirements. Rome, Italy: UN-FAO. Smith, M., R. G. Allen, and L. S. Pereira. 1996. Revised FAO methodology for crop water requirements. In Proc. International Conf. on Evapotranspiration and Irrigation Scheduling, 116-123. St. Joseph, Mich.: ASAE. Snyder, R. L., J. P. De Melo-Abreu, and S. Matulich. 2005. Frost Protection: Fundamentals, Practice and Economics. Vol. 2, FAO Environment and Natural Resources Service Series 10. Rome: FAO. 240 pp and 64 pp +CD-ROM. Snyder, R. L,. and S. Eching. 2004. Landscape Irrigation Management Program— IS005 Quick Answer. Univ. California, Davis. biomet.ucdavis.edu/irrigation_ scheduling/LIMP/limp.pdf. Snyder, R. L., and S. Eching. 2005. Urban Landscape Evapotranspiration. California State Water Plan, vol. 4, 691-693. www.waterplan.water.ca.gov/reference/index. cfm#infrastructure Snyder, R. L., and W. O. Pruitt. 1992. Evapotranspiration data management in California. In Proc. of the Irrigation and Drainage sessions at ASCE Water Forum ’92, 128-133. T. Engman, ed. New York, N.Y.: American Soc. Civil Engineers. Snyder, R. L., B. J. Lanini, D. A. Shaw, and W. O. Pruitt. 1989a. Using reference evapotranspiration (ETo) and crop coefficients to estimate crop evapotranspiration (ETc) for agronomic crops, grasses, and vegetable crops. Leaflet No. 21427. Berkeley, Calif.: Cooperative Extension, Univ. California. Snyder, R. L., B. J. Lanini, D. A. Shaw, and W. O. Pruitt. 1989b. Using reference evapotranspiration (ETo) and crop coefficients to estimate crop evapotranspiration (ETc) for trees and vines. Leaflet No. 21428. Berkeley, Calif.: Cooperative Extension, Univ. California. Soil Conservation Service. 1982. National Engineering Handbook. U.S. Govt. Printing Office. Sun, J., S. K. Esbensen, and L. Mahrt. 1995. Estimation of surface heat flux. J. Atmos. Sci. 52: 3162-3171. Swinbank, W. C. 1951. The measurement of vertical transfer of heat and water vapor by eddies in the lower atmosphere. J. Meteorol. 8: 135-145. Tanner, C. B. 1960. Energy balance approach to evapotranspiration from crops. Soil Sci. Soc. America Proc. 24: 1-9. Tanner, C. B. 1967. Measurement of evapotranspiration. In Irrigation of Agricultural Lands, 534-574. R. M. Hagan, H. R. Haise, and T. W. Edminster, eds. Madison, Wis.: American Soc. of Agron. Tanner, B. D. 1988. Use requirements for Bowen ratio and eddy correlation determination of evapotranspiration. In Management of Irrigation and Drainage Systems: An Integrated Approach, Proc. National Conference on Irrigation and Drainage Engineering, 605-616 R. G. Allen, and C. M. Neale, eds. New York, N.Y.: American Soc. Civil Engineers. Tasumi, M., R. G. Allen, R. Trezza, and J. L. Wright. 2005. Satellite-based energy balance to assess within-population variance of crop coefficient curves. J. Irrig.

288

Chapter 8 Water Requirements

Drain. Eng. 131(1): 94-109. Testi, L., F. J. Villalobos, and F. Orgaza. 2004. Evapotranspiration of a young irrigated olive orchard in southern Spain. Agric. and For. Meteor. 121(1,2):1-18. Tetens, O. 1930. Uber einige meteorologische Begriffe. Z. Geophys. 6: 297-309. Thornton, P. E., and S. W. Running. 1999. An improved algorithm for estimating incident daily solar radiation from measurements of temperature, humidity, and precipitation. Agric. and For. Meteor. 93: 211-228. Twine, T. E., W.P. Kustas, J.M. Norman, D.R. Cook, P.R. Houser, T. P. Meyers, J. H. Prueger, P. J. Starks, and M.L. Wesely. 2000. Correcting eddy-covariance flux underestimates over a grassland. Agric. For. Meteor. 103: 279-300. USDA-SCS. 1972. National Engineering Handbook, Section 4, Table 10.1. USGS (U.S. Geological Survey). 1981. Guidelines for determining flood flow frequency. Bulletin 17B. Reston, Va.: Office of Water Data Coordination, Hydrology Subcommittee, Interagency Advisory Committee on Water Data. Vanderkimpen, P. J. 1991. Estimation of crop evapotranspiration by means of the Penman-Monteith equation. Unpublished Ph.D. diss. Logan, Utah: Dept. of Biol. and Irrig. Engineering, Utah State Univ. Vanderlinden, K., J. V. Giraldez, and M. Van Mervenne. 2004. Assessing reference evapotranspiration by the Hargreaves method in Southern Spain. J. Irrig. Drain. Eng. 129(1): 53-63. Villalobos, F. J., F. Orgaz, L. Testi, and E. Fereres. 2000. Measurement and modeling of evapotranspiration of olive orchards. European J. Agron. 13:155-163. Wilson, E. B., and M. M. Hilferty. 1931. The distribution of Chi-square. Proc. National Academy of Science 17(12): 684-688. Wilson, K. et al. 2002. Energy balance closure at FLUXNET sites. Agric. For. Meteol. 113: 223-243. Wolfe, J. W., P. B. Lombard, and M. Tabor. 1976. The effectiveness of a mist versus a low pressure sprinkler system for bloom delay. Trans. ASAE 19(3): 510-513. World Meteorological Organization. 1966. Measurement and estimation of evaporation and evapotranspiration. Tech. Note No. 83. Geneva, Switzerland: WMO. Wright, J. L. 1981. Crop coefficients for estimates of daily crop evapotranspiration. Irrig. Scheduling for Water and Energy Conserv. in the 80’s. St. Joseph, Mich.: ASAE. Wright, J. L. 1982. New evapotranspiration crop coefficients. J. Irrig. Drain. Div., ASCE 108: 57-74. Wright, J. L. 1985. Lysimeter measurements of alfalfa evapotranspiration. Unpublished data. Kimberly, Idaho: USDA-ARS. Wright, J. L. 1988. Daily and seasonal evapotranspiration and yield of irrigated alfalfa in southern Idaho. Agron. J. 80: 662-669. Wright, J. L. 1990. Evapotranspiration data for dry, edible beans, sugar beets, and sweet corn at Kimberly, Idaho. Unpublished data. Kimberly, Idaho: USDA-ARS. Wright, J. L. 1995. Calibrating an ET procedure and deriving ET crop coefficients. In Proc. of the 1995 Specialty Seminar. Sponsored by the Water Res. Comm., American Consulting Engin. Council of Colo. and the Colo. State Engineers Office. Wright, J. L. 1996. Dormant Season Evaporation in Southern Idaho. Presentation at the North American Water and Environment Congress, ASCE. Wright, J. L., and M. E. Jensen. 1972. Peak water requirements of crops in southern Idaho. J. Irrig. Drain. Div., ASCE 96(IR1): 193-201.

CHAPTER 9

DRAINAGE SYSTEMS Larry G. King (Washington State University, Pullman, Washington) Lyman S. Willardson (Utah State University, Logan, Utah) Abstract. This chapter describes the basics of design, operation, and maintenance of subsurface drainage systems as needed for sustained irrigated agriculture. Emphasis is on the control and management of the water table as needed for crop production in harmony with environmental concerns. Two approaches to design of subsurface drainage systems are based on steady state and transient flows. Steady state design of drains is based on Hooghoudt’s equations as modified by Moody. Two examples are presented in the form of spreadsheets. Transient flow design is based on procedures developed by Dumm and Glover and others in the U.S. Bureau of Reclamation (USBR). Spreadsheets are presented using the USBR transient approach for parallel relief drains. Spreadsheets are also presented for calculating drain spacing using the Hooghoudt equation and the USBR transient flow approach. Procedures are described for determining design variables. Drainage materials, construction methods, and equipment are described. The chapter concludes with brief summaries of system operation, maintenance, and performance evaluation. Keywords. Drainage systems, Drainage system design, Drainage system materials, Environmental considerations, Hooghoudt equation, Leaching and salinity control, Moody equations, Operation and maintenance, Steady state flow, Subsurface drainage systems, Transient flow, Water table control.

9.1 INTRODUCTION This chapter addresses agricultural drainage with emphasis on drainage under irrigated agriculture, in contrast to drainage of building foundations, dams, airport runways, highway embankments, etc. Of course the principles can be applied to these latter cases of drainage, but the focus will be on drainage of agricultural lands for crop production under irrigation. Thus, more attention will be given to control and management of the water table in the soil profile than to removal of excess water from the land surface. Proper application of principles to drainage of irrigated land should not be in direct conflict with the maintenance and/or restoration of wetlands, since naturally occurring wetlands are not removed. In some cases, as within the Columbia Basin Project in central Washington, large wetlands areas have been created as a result of an irrigation project. Drainage of irrigated fields has provided water to supply wetlands within the

290

Chapter 9 Drainage Systems

project. However, engineers must be very careful with uses of drainage effluent due to the quality of this water, as has been so dramatically demonstrated by the problems in California’s Central Valley and Kesterson Wildlife Reserve. It may appear paradoxical that after large sums are spent on infrastructure to deliver irrigation water to farms, even larger sums must be spent to remove water through drainage systems. It is true that improved irrigation systems and water management may in many situations reduce the costs of drainage systems. However, even after making practical improvements in irrigation water management at various scales, drainage system construction will be necessary. The goal of this chapter is to develop and discuss the basics of design, operation, and maintenance of subsurface drainage systems to serve irrigated farms. The material presented in this chapter provides scientifically based engineering methods for design of these systems. Because the principals are the same for both humid and arid regions, and because design is more critical in arid regions, the chapter will treat drainage design in arid regions in detail. Any important differences in design related to humid areas will be pointed out as appropriate. 9.1.1 Definition of Drainage Systems Agricultural drainage systems related to crop production are classified as surface drainage systems or subsurface drainage systems depending on where the water causing the drainage problem exists. That is, the classification is based on the function of the system not on the location of system components. Surface drainage systems remove excess water resulting from rainfall, snowmelt, and surplus irrigation from the land surface. Subsurface drainage systems remove soil water to keep the water table level below the bottom of the active root zone, to avoid waterlogging. Buried pipes may be part of a surface or subsurface drainage system depending upon the source of the water they are conveying. Likewise, an open ditch may function to control the water table by removing soil water and be part of a subsurface drainage system, or it may carry excess water from the soil surface as part of a surface drainage system. In arid regions, subsurface drains are often deeper than in humid regions because of the need for salinity control; deep subsurface drains make salinity control easier. Surface drains may be shallow or deep and are normally unlined open ditches. Subsurface drains are deeper, and may be open ditches, but are usually buried perforated pipes. Before the advent of corrugated plastic drain tubing, buried subsurface drains were constructed of rigid clay tiles or concrete pipes with open joints between short lengths of these pipes. The corrugated plastic drain tubing has perforations to admit water and usually is surrounded by a graded gravel envelope. Many buried pipe drainage systems discharge by gravity directly into waterways or constructed open channels. Others discharge to a sump from which the water is pumped into a suitable open channel or waterway. Typically, drainage systems consist of laterals, collectors, and outlets. The water to be removed by the system first enters the lateral. The water from several laterals enters a collector. If the system is large we may also have sublaterals and subcollectors where the water first enters the sublateral and sequentially goes to the lateral, subcollector, and collector. Depending on system size, the collector or collectors convey the water to the system outlet. The outlet is the discharge point or points of the system where the water enters existing channels or waterways.

Design and Operation of Farm Irrigation Systems

291

9.2 ENVIRONMENTAL CONSIDERATIONS The time has long since past (if it ever existed) when an engineer can ignore the environmental consequences of designs. This is as true in agricultural drainage as in any other engineering design effort. This section briefly discusses some issues to be considered, including quality of irrigation water, soil water, and drainage system effluent water; the quantity of drainage system effluent water; drainage in relation to wetlands; and the reclamation of salinized soils. It is hoped that through discussion of these environmental factors, the designers of agricultural drainage systems will consider a holistic approach to the designs so that the systems work in harmony with the environment rather than in conflict with other uses of the environment. 9.2.1 Irrigation Water Quality Irrigation water quality has little to do with the actual physical design of either surface or subsurface drainage systems. If irrigation water is of poor quality from a salinity standpoint, the amount of water that must pass through the soil to control salinity may be larger than when good quality water is used. The subsurface drainage water is naturally of lower quality than the applied irrigation water; and, if irrigation water quality is poor, the subsurface drainage water may need to be kept separated from the water supply to other users. However, it may be possible to reuse most of the drainage water for irrigation. Surface drainage waters can be reused without restriction unless they happen to contain high concentrations of herbicides, pesticides, or pathogens that may damage crops where the waters are used. 9.2.2 Soil Water Quality For plants to grow unaffected by the quality of the water in the soil pores, some proportion of the applied irrigation water must pass downward through the soil profile. Table 7.4 in Chapter 7 lists the salt tolerance of agricultural crops. If the salinity of the irrigation water is high, a leaching fraction must be added to the water applied in order to keep the soil water salinity below the threshold values shown in the table and avoid crop yield losses. See Chapter 7 for additional details. 9.2.3 Drainage System Effluent Quality The quality of drainage system effluent is a function of (1) the quality of the water entering the soil surface; (2) the timing and amounts of fertilizers and pesticides applied to the crop; (3) the natural materials in the soil; and (4) the proportion of the water extracted by the plants for evapotranspiration. If the plants extract a high proportion of the water from the soil profile, the salt concentration in the remaining water will be increased significantly because most agronomic plants extract nearly pure water and leave most of the salts in the soil. Highly water-soluble fertilizers and pesticides that are not used or degraded by the soil environment will appear in the drainage water. Some pesticides or their daughter products are molecularly bound to the soil particles and do not remain in the percolating water. Nitrate nitrogen from fertilizer applications may appear in drainwater if it is leached through the profile by excess irrigation or rainfall before the fertilizer is consumed by the plants, and is often used as a general indicator of pollution by other agents as well. If there is doubt about the quality of drainage water, samples should be taken over time and analyzed to determine the level of dissolved constituents in the water. Some soils naturally contain toxic elements (e.g., selenium) that may be mobilized if water is leached through the soil.

292

Chapter 9 Drainage Systems

9.2.4 Drainage System Effluent Quantity Water discharged from subsurface agricultural drains must come from water that infiltrated through the soil surface and water seeping into the drains from the water table. In humid areas, the largest volume of drainage water comes from excess rainfall that seeps into the soil and must be removed to prevent a detrimental rise in the water table. At the time of peak discharge, drains in humid areas may discharge as much as 2.3 L/s per ha, or a depth of 20 mm/day, but are normally designed to carry about 0.8 L/s per ha (7 mm/day). In arid areas, drain discharge rates usually do not exceed 0.3 L/s per ha (3 mm/day). Drain discharge rates may be higher if the drains are discharging groundwater that moves into the area from higher lands or canal seepage. 9.2.5 Drainage and Wetlands The basic purpose of agricultural drainage is to remove the excess water from the soil surface and the soil profile to levels such that crops can be grown successfully. Drainage of a land area will therefore change it from having a wet soil to having a soil dry enough to be cultivated. This can create some interesting economic, institutional, and social issues. If an area has been wet for long periods and cannot be cultivated, it will develop an ecology that is typical of a “wetland” and may have a microenvironment that serves as habitat for various species of plants and animals. If biological studies show that the wet area serves a public good, it may be mandated that it cannot be drained. A nearby wetland of similar size can sometimes be substituted, allowing artificial drainage of a troublesome wet area in a field to improve the efficiency of the farming operation. In the U.S., any drainage of an area that could potentially be declared a wetland must be approved according to the procedures and regulations in Section 404 of the U.S. Clean Water Act. To avoid federal penalties, USDA Natural Resource Conservation Service specialists or members of the U.S. Army Corps of Engineers should be contacted for clearance before any construction takes place. 9.2.6 Reclamation of Salinized Soils Soils become salinized in arid areas because the water table is too close to the surface. Water from a shallow water table can move easily to the soil surface where it evaporates. The salt carried by the water is therefore deposited on or near the soil surface. When the salt content of the soil becomes too high, crop seeds will not germinate and the soil itself may be physically damaged by the salt. The first step in reclaiming a salinized soil is to take samples for analysis to determine whether drainage alone will remedy the problem or whether it may be necessary to add some source of soluble calcium, such as gypsum, to aid in the reclamation process. If the soil does not need extra calcium, drains can be installed to lower the water table to an appropriate depth. For arid regions, drains should be installed at a depth of at least 2 m to prevent resalination of the soil following reclamation. When the drains are installed and functioning, the soil can be irrigated a few times, in the absence of a crop, to leach the salts out of the surface soils. Once the salt content of the surface soil has been lowered sufficiently that crop seeds can be germinated, salttolerant crops can be grown economically. Until the salt content of the surface soil is lowered to a satisfactory amount, overirrigation should be practiced. In some cases, it may be desirable to install the drainage system and then pond the soil surface to quickly leach the soil profile. Refer to Section 7.6.4 to estimate the amount of water

Design and Operation of Farm Irrigation Systems

293

that will be needed for reclamation. Once crops are growing, uniform irrigations will eventually remove the excess salt from the profile.

9.3 DRAINAGE SYSTEM REQUIREMENTS In Pakistan, the people appropriately refer to the “double menace” of waterlogging and salinity when describing drainage problems. Thus, drainage systems in irrigated areas should provide control of the water table and allow for leaching water to control soil salinity in the crop root zone. In addition to control of the water table and soil salinity, successful agriculture requires appropriate timing of certain critical field operations. Where natural drainage is inadequate, artificial drains are needed to remove excess soil water at times during the year when the soil is generally too wet for these operations. This may apply to entire farms or to only a small area of a large field. In the latter case, localized drains may be installed in a portion of the field to overcome the lack of natural drainage or to intercept seepage from a canal or other sources. Before proceeding to the details of drainage system design approaches in Section 9.4, it is helpful to consider two important aspects of the need for constructed drainage systems in irrigated agriculture. 9.3.1 Water Table Control Plants grow best when the water table is kept at or below the bottom of the crop root zone. For control of excess water in the root zone, subsurface drains are generally designed to keep the water table midway between adjacent drains from getting closer to the surface than 1 m. (In organic soils, the water table should be maintained at approximately 0.5 m deep to prevent excess oxidation of the organic soil materials.) The depth and spacing of the subsurface drains determines the midpoint water table depth at the design discharge rate. In humid areas the drains are commonly installed at depths of 1.0 to 1.5 m, while in arid areas drains are typically installed at depths of 2 m or more. Drain spacing in arid areas may be 4 to 5 times greater than in humid areas because there is less excess water to be removed. 9.3.2 Leaching and Salinity Control If drains are installed at a 2-m depth or deeper in an arid area and if adequate amounts of irrigation water are applied in the course of normal cultivation of a crop, salinity will be controlled and adequate leaching will occur to prevent soil salination. Drains do not need to be made larger for reclamation of saline soils because all movement of water is downward in the soil profile. If the water temporarily exceeds the carrying capacity of the drains, the flow of water through the soil will be slower and the reclamation may be more effective since the salt in the soil has more time to dissolve in the leaching water.

9.4 APPROACHES TO DESIGN OF SUBSURFACE DRAINAGE SYSTEMS There are two general approaches for calculating drain spacing for a subsurface drainage system, the steady state and the transient approaches. In a steady state design, the water table is assumed to be maintained at a constant level by a continuous slow uniform recharge of water through the soil surface. A steady state design theory was developed by Hooghoudt (1940) in the Netherlands, where rainfall is slow and occurs over a long period of time. When recharge stops, the water level between

294

Chapter 9 Drainage Systems

drains goes down until it reaches the level of the drains. The transient design theory was developed by the U.S. Bureau of Reclamation (USBR, 1993) to more closely represent the periodic recharge to the groundwater that occurs under irrigated conditions. In a transient design, the water table is assumed to rise on the date of irrigation and to fall continuously until the next irrigation or recharge event. In the real world, water tables fluctuate between drains regardless of whether a steady state or a transient method was used to determine spacing. Constructed drainage systems can be classified according to the nature of the water source causing the drainage problem and the protection offered by the drainage system. If the source of water causing the drainage problem originates within the area protected by the drainage system, the system is called a relief drainage system. The drains in this case are referred to as relief drains. The spacing of parallel relief drains may be determined by either the steady state or transient design approach (See Section 9.4.1). If the water originates outside the area being protected by the drains, then the drains intercept this water and are called interceptor drains. Interceptor drains are a special steady state design case and are treated in Section 9.4.2. The plan view geometry (or layout) of drainage systems can be modified to meet special physical conditions. NRCS (2001) shows various layouts for field drains that describe their appearance on a map: random, herringbone, or parallel (see also USBR, 1993). 9.4.1 Relief Drains Drains installed in a field to cause a general lowering of the water table are called relief drains. They effectively “relieve” the high water table. As an alternative to relief drains, networks of small wells can be installed in an area with a high water table for the purpose of lowering the water table. They are called relief wells and have the same effect as relief drains installed horizontally below the soil surface. If the groundwater is of acceptable quality, it can be used directly for irrigation or even municipal use. If the groundwater is salty, horizontal relief drains at a relatively close spacing are preferred over wells or deep drains at wide spacing to minimize the amount of deep saline groundwater removed. 9.4.1.1 Steady flow. Drainage systems in both humid and arid areas can be designed using steady state theory. When the appropriate design parameters are selected, the proper design will result. The amount of water to be removed by the drainage system must be determined. This amount is described as a volume per unit area per unit of time and is called the drainage coefficient. The units of the drainage coefficient are usually expressed in mm/day or m/day, and must be the same as the units of hydraulic conductivity of the soil that are used in the design calculations. In humid areas, the drainage coefficient is based on local experience and is approximately 7 mm/day. The assumption made is that there will be 7 mm/day of excess water from rainfall that will enter the soil profile and that must be removed by the drainage system to prevent the water table from rising closer than 1 m to the soil surface. In arid areas, a steady state drainage coefficient is computed from information about the average peak rate of evapotranspiration (ET) and the fraction of the applied irrigation water that is not consumed by the crop. The fraction of the irrigation water that infiltrates into the soil, but is not consumed by the crop, is related to the irrigation efficiency or water application efficiency. Any surface runoff water from the irrigation should not be included in the

Design and Operation of Farm Irrigation Systems

295

drainage coefficient or the drain spacing will be smaller than necessary. If salinity control is required, a leaching fraction must be added to the drainage coefficient. Equation 9.1 can be used to determine the drainage coefficient for an arid area: q = ET [1 – SF(1 – LF)]/SF (9.1) where q is the drainage coefficient expressed in the same units as ET, ET is expressed as a rate of use per day or as a depth of water consumed over a specific period of time, LF is the leaching fraction (see Chapter 7), and SF is the fraction of the infiltrated irrigation water that is stored in the root zone. The stored fraction (SF) of the irrigation water can be calculated from data for a typical irrigation as SF = ET/[IW(1 – RF) – ET(LF)] (9.2) where IW is the gross amount of water applied, ET is the amount of water consumed by the plants since the last recharge event, and RF is the fraction of the irrigation water that runs off. The stored fraction SF has also been called irrigation efficiency (of the infiltrated water) or water application efficiency (not including runoff). A typical value for the drainage coefficient for an arid area is from 1 to 3 mm/day. If the equation gives a higher value, the wrong efficiency is being used. The leaching fraction used in the drainage coefficient equation will vary from 0.0 to 0.2 depending on the salt concentration of the irrigation water (see Chapter 7) but will seldom exceed 0.05. Improper irrigation practices that result in poor uniformity (low efficiency) should not be assumed to satisfy the annual leaching requirement. Calculation or selection of a steady flow drainage coefficient provides part of the necessary input to steady state drainage system design. The steady state design equation used here was developed for an idealized flow system for two parallel buried pipe drains taken from the middle of a larger set of parallel buried pipe drains. The drains are at some distance D above a horizontal impermeable barrier. This idealized flow system is shown in Figure 9.1. It is assumed in Figure 9.1 that the ground surface and the impermeable barrier are horizontal and the two drains are at the same depth. It is customary to assume that the drains are half full (i.e., the design depth for the drains is to the drain pipe centerline.)

Figure 9.1. Idealized flow system for steady state buried pipe drainage to two drains taken from a set of parallel drains at a distance D above an impermeable barrier.

296

Chapter 9 Drainage Systems

Using the Dupuit-Forchheimer (D-F) assumptions, Hooghoudt (1940) developed an equation of the form: S = (4h / q )( K1h + 2 K 2 De )

(9.3)

where S = drain spacing (in the horizontal direction) q = drainage coefficient h = height of the water table above the drain pipe centerline midway between the two drains K1, = hydraulic conductivity values for the soil above the drains K2 = hydraulic conductivity values for the soil below the drains De is the Hooghoudt equivalent depth coefficient, which is explained below. Note from Figure 9.1 and Equation 9.3 that the Hooghoudt equation applies directly to the special layered soil case where the drains lie at the interface between the two soil layers having different values of hydraulic conductivity. Also note that for K1 = K2, the soil is homogeneous and the equation also applies with K1 = K2 = K. For the case when K2 0.31

(9.6)

Charts to determine Hooghoudt equivalent depth have been prepared and presented in the literature by various authors. We prefer the Moody equations especially for ease of use with computers. By inspection of the Hooghoudt and Moody equations, it is easily seen that the calculation of the proper drain spacing for relief drains is an iterative (or trial and error) procedure because S depends on De and De depends upon S. In fact, the whole design procedure is iterative. Also note that the depth of the drains below ground surface is not expressly included in the Hooghoudt equation (Figure 9.1, Equation 9.3). The iterative design procedure starts with a depth of root zone appropriate for the crop to be grown. If the designer is inexperienced, calculations for all crops that may be grown should be made and then the most critical one should be chosen as the design crop. Once the depth of the root zone, Drz, has been determined, a trial drain depth is chosen. See Section 9.5.2 for discussion of proper drain depths. Since the drain depth is the sum of Drz and h, the choice of the trial drain depth determines the design value for h. Once the depth from ground surface to the impermeable barrier (see Section 9.5.3) is known, D has also been determined. Next, the designer must choose a trial diameter of drain pipe from which the drain radius r is determined. The effective drain radius is half the drain diameter plus the thickness of gravel envelope surrounding the drain pipe. After the drain spacing has been calculated, the drain discharge must be calculated and compared to the pipe capacity to see if this trial drain diameter is satisfactory.

298

Chapter 9 Drainage Systems A

B

C

D

E

1 THIS SPREADSHEET USES HOOGHOUDT'S EQUATION 2 FOR STEADY FLOW WHEN 0 < D/S < 0.31 or D/S = 0.31

F

G

H

FILE NAME: HODESNL

3 FOR LAYERED SOILS 4 INSTRUCTIONS for use of this spreadsheet to calculate drain spacing: 5 1. Enter the proper values for the input data into CELLS A14 through A19 6 2. If you wish to see the effect of a new trial drain spacing, change the first value of S 7 (in CELL A21) 8 3. Accept the last value of S in the first column as the correct drain spacing ONLY if: 9 (a) this last value is equal to the immediately preceding value of S , 10 AND 11 (b) D/S lies in the range 0 < D/S < 0.31 or D/S = 0.31 12 4. If D/S > 0.31, then use the spreadsheet with file name HODSN31L 13 INPUT DATA: 14 = K1 0.75 m/day (hydraulic conductivity above drains) 15 = K2 0.75 m/day (hydraulic conductivity below drains) 16 0.0025 m/day = q (drainage coefficient) 17 0.75 m = h (water table height above drains at mid-point) 18 0.15 m = r (effective drain radius) 19 2.5 m = D (depth of barrier below drains) 20 D e (m) S (m) D/S S 2 (m2) 21 100.00 0.03 2.29 4798 22 69.27 0.04 2.21 4649 23 68.18 0.04 2.20 4641 24 68.13 0.04 2.20 4641 25 68.12 0.04 2.20 4641 26 68.12 0.04 2.20 4641

Figure 9.3. Spreadsheet using the Hooghoudt and Moody equations to calculate steady state relief drain spacing for 0 < D/S ≤ 0.31.

With the foregoing variables determined and knowing the soil hydraulic conductivity K (or K1 and K2) and the drainage requirement q, the designer is ready to use the Hooghoudt and Moody equations to calculate the drain spacing S. While the iterative solution of these equations can be done with a hand calculator, a computer spreadsheet is more useful, especially for the inexperienced designer. Two such spreadsheets are given as Figures 9.3 and 9.4. Figure 9.3 is the one used most often and is for 0 < D/S < 0.31. Figure 9.4 applies when D/S > 0.31. Instructions for use of the spreadsheets are included within the spreadsheets. Use of the spreadsheets will be discussed first and then details of how they are constructed will be given. The known or already determined values of the input variables are placed into cells A14 through A19. Next a trial value for drain spacing S is entered into cell A21. The spreadsheet then calculates the values row by row as shown. The correct drain spacing is given as the last value in column A if the last value of S is identical to the next-to-last value of S immediately above and the values of D/S lie in the range 0 < D/S ≤ 0.31. If D/S > 0.31, then the spreadsheet of Figure 9.4 must be used. The spreadsheets are constructed as follows. In Figure 9.3, the first trial value of drain spacing is entered into cell A21. Then the value of D from cell A19 is divided by

Design and Operation of Farm Irrigation Systems A

B

C

D

299 E

F

G

H

1 THIS SPREADSHEET USES HOOGHOUDT'S EQUATION 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32

FILE NAME: FOR STEADY FLOW WHEN D/S > 0.31 HODSN31L FOR LAYERED SOILS INSTRUCTIONS for use of this spreadsheet to calculate drain spacing: 1. Enter the proper values for the input data into CELLS A14 through A19 2. If you wish to see the effect of a new trial drain spacing, change the first value of S (in CELL A22) 3. Accept the last value of S in the first column as the correct drain spacing ONLY if:: (a) this last value is equal to the immediately preceding value of S , AND (b) D/S lies in the range D/S > 0.31 4. If 0 < D/S < 0.31 or D/S = 0.31, then use the spreadsheet with file name HODESNL INPUT DATA: = K1 (hydraulic conductivity above drains) 0.75 m/day = K2 0.75 m/day (hydraulic conductivity below drains) 0.0025 m/day = q (drainage coefficient) 0.75 m = h (water table height above drains at mid-point) 0.15 m = r (effective drain radius) 50 m = D (depth of barrier below drains) D e (m) S (m) D/S S 2 (m2) 200.00 0.25 12.99 24060 155.11 0.32 10.52 19607 140.03 0.36 9.67 18073 134.44 0.37 9.35 17499 132.29 0.38 9.22 17278 131.44 0.38 9.18 17191 131.11 0.38 9.16 17157 130.98 0.38 9.15 17143 130.93 0.38 9.15 17138 130.91 0.38 9.14 17136 130.90 0.38 9.14 17135 130.90 0.38 9.14 17134 Figure 9.4. Spreadsheet using the Hooghoudt and Moody equations to calculate steady state relief drain spacing for D/S > 0.31.

this trial S to give D/S in cell B21. In cell C21, α is calculated from Equation 9.5, and then De is calculated from Equation 9.4. Next S2 is calculated from Equation 9.3 in cell D21. The square root of this last value gives the new trial value of S in cell A22 and the process is repeated for row 22, etc. The construction and use of the spreadsheet of Figure 9.4 is entirely parallel to that of Figure 9.3. Equation 9.6 is used to calculate De in Figure 9.4, whereas Equations 9.4 and 9.5 are used for this calculation in Figure 9.3. 9.4.1.2 Steady state relief drain design examples. The spreadsheets of Figure 9.3 and 9.4 also show examples of the drain design process. In Figure 9.3, suppose that the root zone depth to be protected by the drainage system is 1.25 m and the drains are to be placed 2 m deep. Then h is 0.75 m. Suppose the drain pipe is to be 10 cm diameter and the gravel envelope surrounding the drain pipe is 10 cm thick, thus the effective drain radius r is 0.15 m. Also, the drainage coefficient is 2.5 mm/day, the soil is homogeneous with hydraulic conductivity of 0.75 m/day, and the impermeable barrier is

300

Chapter 9 Drainage Systems

4.5 m below ground surface. For these conditions, the input variables are as given in cells A14 through A19 of Figure 9.3. If a trial drain spacing of 100 m is chosen, Figure 9.3 shows the calculations of this example resulting in a drain spacing of 68 m (rounded to the nearest meter). After constructing the spreadsheet, a designer can easily show that the procedure converges rapidly to the same drain spacing regardless of the value chosen for the first trial drain spacing. A first trial value of 10 m will produce the same result for this example. Assuming that the drain depth and the root zone depth of this example are satisfactory, all that remains is to determine if the drain pipe diameter of 10 cm is adequate. Since the flow to the drains is steady, the drain discharge per meter length is equal to the product of the drainage coefficient q and the drain spacing S. Thus for this example, the drain discharge is calculated as Qm = qS = 0.0025(68) = 0.170 m3/day per m of drain. Now if the drain laterals are 300 m long, the discharge from the lateral QL is 51.0 m3/day or 5.9×10-4 m3/s. Using a Manning n value of 0.017 for corrugated plastic drain tubing, the capacity of the drain on a grade of 0.001 is calculated to be 1.25×10-3 m3/s from the Manning equation: Q = VA = (1/n) AR2/3 Sf 1/2

(9.7) where V = velocity A = area R = hydraulic radius Sf = friction slope n = the Manning roughness coefficient. Since the capacity of the lateral (1.25×10-3 m3/s) exceeds the calculated discharge (5.9×10-4 m3/s), the drain pipe diameter of 10 cm is adequate. If the discharge computed with the Manning equation is less than the discharge required from the lateral, choose a larger drain diameter and repeat the spreadsheet calculations with the larger trial drain diameter. The example shown in Figure 9.4 is for a very deep impermeable barrier. All other values are as given in the previous example. The drains must handle a discharge of 1.14×10-3 m3/s and the 10 cm drain diameter is still adequate for the 300-m drains on a grade of 0.001 m/m. 9.4.1.3 Transient flow. Transient drain spacing designs require the translation of an amount of recharge water entering the soil into a rise of the water table. When an irrigation occurs, no water will reach the water table and cause it to rise until the complete root zone of the plants has been restored to field capacity. Therefore, the amount of water required to refill the root zone is equal to the amount of water that has been consumed by the crop since the last recharge event. If the water added to the soil does not refill the soil profile to replace the water consumed by the plants, the water table will not rise in response to an irrigation. The amount of the irrigation water entering the soil, as well as the storage space in the soil is expressed as a volume of water per unit area of soil, which is equivalent to a depth of water. When the ET since the last recharge event, expressed as a depth of water, is subtracted from the average depth of irrigation water infiltrating the soil (volume per unit area excluding any runoff), the difference will be a depth of water that will cause the water table to rise. The actual water table rise that will occur is the depth of recharge water divided by the specific

Design and Operation of Farm Irrigation Systems

301

yield of the soil. Specific yield is explained in Section 9.5.6, where it is called drainable porosity. The assumption made for transient flow designs is that any recharge causes an instantaneous rise of the water table on the day of recharge. Periods of one day are used in the computations. The day of recharge is the last day of the discharge from the previous recharge event and the zero day for computation for the rate of fall of the water table until the next recharge event. The transient design procedure presented herein follows closely that developed by Dumm (1968) and others of the U.S. Bureau of Reclamation (USBR, 1993). This design procedure is based upon a concept of dynamic equilibrium wherein the annual cycle of water table fluctuations is relatively consistent from year to year when adequate drainage is achieved for proper water table levels for the crops grown. In some cases this dynamic equilibrium is reached without construction of buried pipe drains, but in many situations natural drainage is inadequate to provide dynamic equilibrium and subsurface drains are required. Glover and Dumm (USBR, 1993), using heat flow analogies which neglect convergence of flow toward the drains, presented an equation with the initial water table given by a fourth degree parabola: y(x,0) = (8/S4) [y(S/2,0)] ( S3x – 3S2x2 +4 Sx3 – 2x4) for 0 < x < S

(9.8)

where the coordinates y(x,z) have their origin at the centerline of one of the drains such that y = 0 when z = 0, y is the water table height above the centerlines of two parallel relief drains taken from the middle of a larger set of parallel relief drains as was done for steady state relief drains (see Section 9.4.1), and S is the drain spacing. By truncating the resulting infinite series solution and using only the first term of the series, an approximate equation results: h/h0 = (36.37/π3) exp(–π K D t/[fS2])

(9.9)

where D = the time averaged flow depth for any drainout period assumed to be given as D = D + y0/2 at any location x (at the midpoint between the two drains, D = D + h0/2) h0 = height of the water table above the drains midway between the two drains at the start of a water table recession or drainout period h = height of the water table above the drains midway between the two drains at any time t after the start of a drainout period K = soil hydraulic conductivity f = specific yield S = spacing between the two parallel drains. Solving Equation 9.9 for the drain spacing gives:

{

1/ 2

}

S = π KDt / ⎡⎣ f ln(36.37h0 /(π 3 h)) ⎤⎦

(9.10)

Van Schilfgaarde (1965) stated that correction for the convergence effect can be made by applying Hooghoudt’s equivalent depth concept. There is nothing inherent in the concept of equivalent depth that restricts its application to steady flow. Thus, it can be expected to apply to spacing calculations based upon various transient theories that neglect convergence of flow toward the drains. The Moody (1966) equations for De

302

Chapter 9 Drainage Systems

(Equations 9.4, 9.5, and 9.6) will be used here. Thus, replacing D with De to correct for convergence of flow toward the drains, Equation 9.10 becomes

{

(

1/ 2

)) }

S = π KDe t / ⎡⎢ f ln 36.37h0 / π 3 / h ⎤⎥ ⎣ ⎦

(

(9.11)

The equation for drain spacing if the drains sit on an impermeable barrier is (Dumm, 1968): 1/ 2

S = {9 Kh0t / [ 2 f (h0 / h) − 1]}

(9.12)

Dumm (1968) presented a summary of the USBR approach for designing pipe drain systems using a transient equation. He reported that Australian and Canadian research workers and a USBR project provided data for checking the validity of the mathematical developments of the transient flow theory under various field conditions. The reader should refer to Tables 1 and 2 and Figures 2, 3, and 4 of the paper (Dumm, 1968) for these comparisons. The concept of dynamic equilibrium as it relates to drainage is illustrated by Figure 9.5 of Dumm (1968) and by Figure 5-3 of the USBR Drainage Manual (USBR, 1993). In addition to the earlier discussion of the dynamic equilibrium concept, the main elements of this concept are explained as follows. When recharge and discharge become equal, the highest level and the range in cyclic annual fluctuation of the water table become reasonably constant from year to year. The water table rises during the irrigation season and reaches its highest elevation after the last irrigation. When year-around cropping is practiced, this usually occurs at the end of the peak portion of the irrigation season. Dumm (1968), gave equations for the transient drain discharge. Replacing D with De in the equation for drains above the barrier gives:

Q = 2 πK D e / S

(9.13)

and for drains on the barrier: Q = 4 Kh 2 / S

(9.14) in which Q is the drain discharge in m /day per lineal meter of drain, K is in m/day, and De , S, and h are in m. Water balance calculations were shown in Tables 5 and 6 of Dumm (1968) by averaging the beginning and end of each drainout period. Note that Q is a discharge rate and is a function of time because h is a function of time. See comparisons in Figures 9, 10, 11, and 12 of Dumm (1968). Information needed for the USBR approach include estimates of deep percolation losses for each irrigation, which is given in Table 9.1, and methods for estimating specific yield, f. In estimating the specific yield, hysteresis is ignored in the water content as a function of capillary pressure. The USBR approach is to estimate specific yield from the hydraulic conductivity determined by the auger hole method. Figure 9.5 shows a curve adapted from USBR (1993) for obtaining the specific yield. 9.4.1.4 Transient relief drain design example. Perhaps the best way to explain the design process for relief drains using transient flow theories is by a detailed example. Some of the steps involved are the same as for steady state design of relief drains. First, choose a trial drain depth. Next, from the root zone depth to be protected and the trial drain depth, determine the maximum allowable height of the water table above the drains midway between the drains (maximum h or h0). Then, from the trial drain 3

Design and Operation of Farm Irrigation Systems

303

Table 9.1. Approximate deep percolation loss below the root zone for use in drain spacing calculations (adapted from Table 3 of Dumm, 1968).[a] On Basis of Texture Deep Percolation Deep Percolation Texture Percentage[b] Texture Percentage[b] Loamy sand 30 Sandy clay loam 14 Sandy loam 26 Clay loam 10 Loam 22 Silty clay loam 6 Silt loam 18 Sandy clay, and clay 6 On Basis of Infiltration Rate Infiltration Rate Deep Percolation Infiltration Rate Deep Percolation (mm/hr) Percentage[b] (mm/hr) Percentage[b] 1.27 3 25.40 20 2.54 5 31.75 22 5.08 8 38.10 24 7.62 10 50.80 28 10.16 12 63.50 31 12.70 14 76.20 33 15.24 16 101.60 37 20.32 18 [a]

[b]

These deep percolation percentages are usually adequate to provide for effective leaching to maintain salt balance in the root zone. However, if the leaching requirement is larger than the percentage given in the table, the larger value should be used in drain spacing calculations. Percentage is based on the application or net input.

Figure 9.5. Curve used by USBR to estimate specific yield from hydraulic conductivity (adapted from Figure 2-4, USBR, 1993).

304

Chapter 9 Drainage Systems

Table 9.2. Seasonal ET of sugar beets used for the design example. Period ET (mm/day) No. of Days 151 May – 31 May 3.0 17 1 June – 10 June 5.1 10 11 June – 27 June 5.6 17 28 June – 5 Sept 6.4 70 6 Sept – 25 Sept 5.6 20 26 Sept – 15 Oct 5.1 20

depth and the depth of the impermeable barrier below ground surface, determine D. Also, choose a trial drain pipe diameter and envelope thickness to select the effective drain radius r. From this point on, additional information must be obtained for transient design beyond that required for steady state design. For transient design, an irrigation schedule must be developed as well as a schedule for all water additions to the soil surface, such as rainfall and/or snowmelt, that will be significant in raising the water table. An ET schedule for water use from the soil root zone must be developed. A criterion must be developed to estimate how much water is used before irrigation is complete. Deep percolation of water below the root zone for each water addition must also be determined. The specific yield of the soil must be determined so that it can be used to estimate the amount of water table buildup (rise) that will be caused by each water addition to the water table. Once all these calculations have been done, the actual calculations of drain spacing can be undertaken. Again, computer spreadsheets will be used for this purpose. Now an example will be given. The details and assumptions given herein comprise just one possible scenario. The scenario used for any specific design must match as close as possible the common irrigation and farming practices for the area. If this example does not match practices for the area where the designer is working, modifications will be required to obtain the best schedule for the specific area. For this example, the silt loam soil has an average hydraulic conductivity of 20 mm/hr as measured by the auger hole method and an available water content of 183 mm/m. The depth from ground surface to the barrier is 12.2 m and the root zone depth for the crop of sugar beets is 1.2 m. Choose a trial drain depth of 2.4 m. To develop an irrigation schedule to be used for design, it is decided that the first irrigation will occur on 15 May and ET prior to 15 May is negligible. The design distribution of ET for the growing season is given in Table 9.2. For the distribution of the sugar beet roots, it is assumed that 40% of the total ET comes from the top quarter of the root zone (top 0.3 m). The irrigation schedule will be based upon use of 75% of the available water in the top quarter of the root zone (top 0.3 m) between irrigations (0.75 × 183 mm/m × 0.3 m = 41 mm). It is assumed that each irrigation application more than fills the root zone to field capacity. Since 40% of the total ET comes from the top quarter of the root zone, then the total root zone ET per irrigation equals 41/0.40 or 103 mm. Assume that the ET on an irrigation day is extracted as part of the drainout period following that irrigation. Now, 17 days at 3.0 mm/day = 51 mm of water and 103 – 51 = 52 mm can still be used before the next irrigation is needed. The next 10 days at 5.1 mm/day uses 51 mm of water and the total use since the irrigation on 15 May has been 51 + 51 = 102 mm for the first 27 days. One more day (11 June) will use 6 mm, bringing the total to 108 mm since 15 May. This exceeds the allowable use of 103 mm, hence the next (2nd) irrigation will

Design and Operation of Farm Irrigation Systems

305

be on 11 June. These calculations are continued (see Table 9.3) to produce the irrigation schedule. Thus, there are nine irrigations scheduled for 15 May, 11 June, 29 June, 15 July, 31 July, 16 August, 1 September, 18 September, and 7 October covering a total span of 145 days (27 + 18 + 16 + 16 + 16 + 16 + 17 + 19). The nonirrigation season then lasts for 365 – 145 = 220 days. Weather records of the area show that the average year will produce about 25 mm of water from snowmelt, which will recharge to the groundwater below the water table on 20 February. The elapsed time from this snowmelt event until the first irrigation on 15 May is 84 days. The rest of the nonirrigation season amounts to 136 days (220 – 84). Because of possible errors using excessively long drainout time periods, the USBR (1993) recommends splitting this part of the nonirrigation season into two nearly equal drainout time periods. Thus, the 136 days is split into two drainout periods of 68 days each. The next steps involve calculation of the deep percolation losses from the irrigations and the water table buildup (rise) from each of the water additions. It is assumed that the silt loam soil has an infiltration rate equal to its hydraulic conductivity. From Table 9.1, for the silt loam soil with an infiltration rate of 20 mm/hr, the deep percolation from each irrigation is estimated at 18%. The net water depth added to the groundwater is then obtained from the 103 mm of water added at each irrigation as follows: Net input = 103/ (1 – 0.18) = 103/0.82 = 126 mm per irrigation. Deep percolation = 126 –103 = 23 mm per irrigation (or 18% of 126). From Figure 9.5 with K = 0.02 m/h, the specific yield is 9% (f = 0.09). Thus for each irrigation, the water table buildup is 23/0.09 = 256 mm or 0.256 m. For the snowmelt the buildup is 25/0.09 = 278 mm or 0.278 m. Table 9.3. Details of irrigation schedule calculations. First irrigation on 15 May Second irrigation on 11 June (17)(3.0) = 51 mm (17)(5.6) = 95 mm 103 – 95 = 8 mm, 8/6.4 = 1.3 days (10)(5.1) = 51 mm 27 days 102 mm (1)(6.4) = 6 mm 18 days 101 mm Third irrigation on 29 June 103/6.4 = 16.1 days (16 days)(6.4) = 102 mm

Fourth irrigation on 15 July 103/6.4 = 16.1 days (16 days)(6.4) = 102 mm

Fifth irrigation on 31 July 103/6.4 = 16.1 days (16 days)(6.4) = 102 mm

Sixth irrigation on 16 August 103/6.4 = 16.1 days (16 days)(6.4) = 102 mm

Seventh irrigation on 1 September (5)(6.4) = 32 mm 103 – 32 = 71 mm, 71/5.6 = 12.7 days (12)(5.6) = 67 mm 17 days 99 mm

Eighth irrigation on 18 September (8)(5.6) = 45 mm 103 – 45 = 58 mm, 58/5.1 = 11.4 days (11)(5.1) = 56 mm 19 days 101 mm

Ninth irrigation on 7 October (9)(5.1) = 46 mm, so a tenth irrigation is unnecessary and ET beyond 15 October is neglected due to beginning of harvest on 16 October.

306

Chapter 9 Drainage Systems

As discussed in Section 9.4.1, the USBR approach is based upon a concept of dynamic equilibrium in which the water table fluctuates during the irrigation season, but generally rises and reaches its highest elevation immediately after the last irrigation. Spreadsheets are now used to calculate the proper drain spacing. This example uses the spreadsheet with filename USBR. This spreadsheet is designed for the USBR transient flow approach using the USBR equation for drains above the barrier. The spreadsheet uses the Moody equations for Hooghoudt equivalent depth. Figure 9.6 shows the first spreadsheet calculation using the input values for this example and a trial drain spacing of 150 m. The instructions for spreadsheet use are included as part of the spreadsheet. An effective drain radius r of 0.15 m (a 10-cm drain diameter with 10-cm thick gravel envelope) is used. Also the trial drain depth is 2.4 m (below ground surface) so that D = 12.2 – 2.4 = 9.8 m. The choice of this first trial drain spacing is not critical since we can use linear interpolation or extrapolation after two trials to get values much closer to the correct drain spacing. Note that since the value of h0 (1.473) at the bottom of Figure 9.6 is greater than the desired value (1.200), the assumed drain spacing of 150 m is too large. Choose a smaller value, say, 120 m, and recalculate using the spreadsheet. Figure 9.7 shows the resulting calculated spreadsheet. Since the last value of h0 (1.016) is too small, the assumed drain spacing (120 m) is too small. The designer may interpolate between the two calculated drain spacings (150 m and 120 m) to obtain a new trial value of drain spacing (132 m). Instead, once the spreadsheet has been constructed, the designer may just wish to enter new trials of S into cell A18 until the last value of h0 matches the desired value (1.200). This latter approach was used to produce Figure 9.8 showing the correct drain spacing to be 132.6 m, rounded to 133 m. 9.4.1.5 Discussion of relief drain design examples. Because of all the data requirements and many assumptions upon which the transient method rests, it would be interesting to know how close results of using a simple approach such as Hooghoudt’s steady state equation would be to those of the transient method. The spreadsheet HODESNL (Figure 9.3) can be used to calculate drain spacing for comparison. The question arises as to how to choose the value of the drainage coefficient, q. Suppose q is calculated as the average recharge rate for the peak irrigation period, i.e., when the interval between irrigations is the shortest. Then for the transient example given in Section 9.4.1, q = (0.023 m)/(16 days) = 0.00144 m/day The calculation of drain spacing with this value of q in Hooghoudt’s equation as shown in Figure 9.9 gives a value of 152.9 m for drain spacing S. This spacing is 15% larger than given by the transient method of calculation (Figure 9.8). Now assume that the transient method gives a good description of water table position throughout the irrigation season. The spreadsheet USBR (Figure 9.6) can be used to find a starting value of h0, to satisfy dynamic equilibrium conditions so that the effect on water table position throughout the year from drains placed at the spacing calculated from the Hooghoudt equation can be predicted. This is done by trial and error until the starting and ending values of h0 in column D match for a drain spacing of 153 m as given by the Hooghoudt equation (Figure 9.9). This result is shown in Figure 9.10 where the dynamic equilibrium value of h0, is given as 1.527 m for a drain spacing of 153 m.

Design and Operation of Farm Irrigation Systems

A 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49

B

C

D

307

E

F

G

H

DRAIN SPACING CALCULATIONS - TRANSIENT FLOW FILE NAME: USBR APPROACH, EQUATION 9.11 USBR INSTRUCTIONS for use of this spreadsheet to calculate drain spacing: 1. This spreadsheet is designed for a specific irrigation and recharge schedule. You MUST modify the first three columns of the table for your particular schedule! 2. Enter data into CELLS A14 through A18. 3. Enter desired height of water table above drain centers as the first h 0 into CELL D28. 4. Print the spreadsheet. 5. Enter a new estimate of drain spacing into CELL A18 and print the new spreadsheet. 6. Use interpolation between spreadsheets to get an improved estimate of drain spacing. 7. Repeat process until h 0 at bottom of column 4 agrees with the value at top of column 4. 8. For D = 0, do not use this spreadsheet! Use file name USBRDOB. INPUT DATA: 0.48 m/day = K (hydraulic conductivity) 0.09 = f (specific yield) 0.15 m = r (effective drain radius) 9.8 m = D (depth of barrier below drains) 150 m = S (estimate of drain spacing) CALCULATIONS: 0.065333 = D/S 3.454004 = α Equation 9.5 Hooghoudt's equivalent depth for 0 < D/S < 0.31, or D/S = 0.31 6.668109 = D e 10.23053 = D e Hooghoudt's equivalent depth for D/S > 0.31 6.668109 = D e Hooghoudt's equivalent depth for conditions of this design h0 K(D e +h 0/2)t h/h 0 t Buildup h IRRIG. (days) (m) (m) (m) NO. f S2 9 68 1.200 0.1172 0.369 0.443 68 0.000 0.443 0.1110 0.392 0.174 SM 0.278 84 0.452 0.1373 0.303 0.137 1 0.256 27 0.393 0.0439 0.760 0.299 2 0.256 18 0.555 0.0296 0.876 0.486 3 0.256 16 0.742 0.0267 0.901 0.668 4 0.256 16 0.924 0.0270 0.898 0.830 5 0.256 16 1.086 0.0273 0.896 0.973 6 0.256 16 1.229 0.0276 0.893 1.097 7 0.256 17 1.353 0.0296 0.876 1.185 8 0.256 19 1.441 0.0333 0.845 1.217 9 0.256 1.473

Figure 9.6. Spreadsheet using the USBR transient approach to calculate parallel relief drain spacing showing first trial drain spacing that is too large.

308

Chapter 9 Drainage Systems

A 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49

B

C

D

E

F

G

H

DRAIN SPACING CALCULATIONS - TRANSIENT FLOW FILE NAME: USBR APPROACH, EQUATION 9.11 USBR INSTRUCTIONS for use of this spreadsheet to calculate drain spacing: 1. This spreadsheet is designed for a specific irrigation and recharge schedule. You MUST modify the first three columns of the table for your particular schedule! 2. Enter data into CELLS A14 through A18. 3. Enter desired height of water table above drain centers as the first h 0 into CELL D28. 4. Print the spreadsheet. 5. Enter a new estimate of drain spacing into CELL A18 and print the new spreadsheet. 6. Use interpolation between spreadsheets to get an improved estimate of drain spacing. 7. Repeat process until h 0 at bottom of column 4 agrees with the value at top of column 4. 8. For D = 0, do not use this spreadsheet! Use file name USBRDOB. INPUT DATA: 0.48 m/day = K (hydraulic conductivity) 0.09 = f (specific yield) 0.15 m = r (effective drain radius) 9.8 m = D (depth of barrier below drains) 120 m = S (estimate of drain spacing) CALCULATIONS: 0.081667 = D/S 3.432672 = α Equation 9.5 Hooghoudt's equivalent depth for 0 < D/S < 0.31, or D/S = 0.31 6.168004 = D e 8.514399 = D e Hooghoudt's equivalent depth for D/S > 0.31 6.168004 = D e Hooghoudt's equivalent depth for conditions of this design h0 K(D e +h 0/2)t h/h 0 t Buildup h IRRIG. (days) (m) (m) (m) NO. f S2 9 68 1.200 0.1705 0.218 0.262 68 0.000 0.262 0.1586 0.245 0.064 SM 0.278 84 0.342 0.1972 0.167 0.057 1 0.256 27 0.313 0.0632 0.628 0.197 2 0.256 18 0.453 0.0426 0.770 0.349 3 0.256 16 0.605 0.0383 0.803 0.486 4 0.256 16 0.742 0.0387 0.800 0.594 5 0.256 16 0.850 0.0391 0.798 0.678 6 0.256 16 0.934 0.0393 0.796 0.743 7 0.256 17 0.999 0.0420 0.775 0.774 8 0.256 19 1.030 0.0470 0.737 0.760 9 0.256 1.016

Figure 9.7. Spreadsheet using the USBR transient approach to calculate parallel relief drain spacing showing second trial drain spacing that is too small.

Design and Operation of Farm Irrigation Systems

A 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49

B

C

D

309

E

F

G

H

DRAIN SPACING CALCULATIONS - TRANSIENT FLOW FILE NAME: USBR APPROACH, EQUATION 9.11 USBR INSTRUCTIONS for use of this spreadsheet to calculate drain spacing: 1. This spreadsheet is designed for a specific irrigation and recharge schedule. You MUST modify the first three columns of the table for your particular schedule! 2. Enter data into CELLS A14 through A18. 3. Enter desired height of water table above drain centers as the first h 0 into CELL D28. 4. Print the spreadsheet. 5. Enter a new estimate of drain spacing into CELL A18 and print the new spreadsheet. 6. Use interpolation between spreadsheets to get an improved estimate of drain spacing. 7. Repeat process until h 0 at bottom of column 4 agrees with the value at top of column 4. 8. For D = 0, do not use this spreadsheet! Use file name USBRDOB. INPUT DATA: 0.48 m/day = K (hydraulic conductivity) 0.09 = f (specific yield) 0.15 m = r (effective drain radius) 9.8 m = D (depth of barrier below drains) 132.6 m = S (estimate of drain spacing) CALCULATIONS: 0.073906 = D/S 3.442674 = α Equation 9.5 Hooghoudt's equivalent depth for 0 < D/S < 0.31, or D/S = 0.31 6.396233 = D e 9.241689 = D e Hooghoudt's equivalent depth for D/S > 0.31 6.396233 = D e Hooghoudt's equivalent depth for conditions of this design h0 K(D e +h 0/2)t h/h 0 t Buildup h IRRIG. (days) (m) (m) (m) NO. f S2 9 68 1.200 0.1443 0.282 0.339 68 0.000 0.339 0.1354 0.308 0.104 SM 0.278 84 0.382 0.1678 0.224 0.086 1 0.256 27 0.342 0.0538 0.690 0.236 2 0.256 18 0.492 0.0363 0.820 0.403 3 0.256 16 0.659 0.0326 0.850 0.560 4 0.256 16 0.816 0.0330 0.847 0.691 5 0.256 16 0.947 0.0333 0.844 0.799 6 0.256 16 1.055 0.0336 0.842 0.889 7 0.256 17 1.145 0.0359 0.823 0.942 8 0.256 19 1.198 0.0403 0.788 0.944 9 0.256 1.200

Figure 9.8. Spreadsheet using the USBR transient approach to calculate parallel relief drain spacing showing correct drain spacing obtained by successive trials of S.

310

Chapter 9 Drainage Systems A

B

C

D

E

1 THIS SPREADSHEET USES HOOGHOUDT'S EQUATION 2 FOR STEADY FLOW WHEN 0 < D/S < 0.31 or D/S = 0.31

F

G

H

FILE NAME: HODESNL

3 FOR LAYERED SOILS 4 INSTRUCTIONS for use of this spreadsheet to calculate drain spacing: 5 1. Enter the proper values for the input data into CELLS A14 through A19 6 2. If you wish to see the effect of a new trial drain spacing, change the first value of S 7 (in CELL A21) 8 3. Accept the last value of S in the first column as the correct drain spacing ONLY if: 9 (a) this last value is equal to the immediately preceding value of S , 10 AND 11 (b) D/S lies in the range 0 < D/S < 0.31 or D/S = 0.31 12 4. If D/S > 0.31, then use the spreadsheet with file name HODSN31L 13 INPUT DATA: 14 = K1 0.48 m/day (hydraulic conductivity above drains) 15 = K2 0.48 m/day (hydraulic conductivity below drains) 16 0.00144 m/day = q (drainage coefficient) 17 1.2 m = h (water table height above drains at mid-point) 18 0.15 m = r (effective drain radius) 19 9.8 m = D (depth of barrier below drains) 20 D e (m) S (m) D/S S 2 (m2) 21 100.00 0.10 5.74 20274 22 142.39 0.07 6.55 22895 23 151.31 0.06 6.69 23318 24 152.70 0.06 6.71 23381 25 152.91 0.06 6.71 23390 26 152.94 0.06 6.71 23391

Figure 9.9. Drain spacing estimated from Hooghoutd’s equation with q as the average recharge rate for the peak irrigation period of this example.

Careful study of Figure 9.10 shows that after irrigations 6, 7, 8, and 9, h0 exceeds the desired value of 1.200 m and h exceeds this value at the ends of the drainout periods following irrigations 7 and 8. At all other times of the year, the water table is below the 1.200 m depth as desired for the crop of this example. A designer must evaluate the economic damages of such water table positions before deciding whether one is justified in using the simpler steady state approach instead of the much more datademanding transient approach to design the drainage system. Of course, many more sets of conditions and situations than considered here would have to be evaluated before one could reach any general conclusions on this question. Further work is left to the individual designer. Of course, as was mentioned in Section 9.4.1 on steady state design, the designer must complete the work by ensuring that the trial drain pipe diameter is sufficient to carry the expected drain discharge (refer to Equation 9.7). One could also develop a spreadsheet for the special transient design case where the drains rest upon the impermeable barrier using Equation 9.1. The designer is cautioned to use the design tools presented here with judgment tempered by experience. Especially with the transient approach, an appropriate irrigation schedule must be built into the spreadsheet to handle the specific situation for the design conditions at hand.

Design and Operation of Farm Irrigation Systems

A 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49

B

C

D

311

E

F

G

H

DRAIN SPACING CALCULATIONS - TRANSIENT FLOW FILE NAME: USBR APPROACH, EQUATION 9.11 USBR INSTRUCTIONS for use of this spreadsheet to calculate drain spacing: 1. This spreadsheet is designed for a specific irrigation and recharge schedule. You MUST modify the first three columns of the table for your particular schedule! 2. Enter data into CELLS A14 through A18. 3. Enter desired height of water table above drain centers as the first h 0 into CELL D28. 4. Print the spreadsheet. 5. Enter a new estimate of drain spacing into CELL A18 and print the new spreadsheet. 6. Use interpolation between spreadsheets to get an improved estimate of drain spacing. 7. Repeat process until h 0 at bottom of column 4 agrees with the value at top of column 4. 8. For D = 0, do not use this spreadsheet! Use file name USBRDOB. INPUT DATA: 0.48 m/day = K (hydraulic conductivity) 0.09 = f (specific yield) 0.15 m = r (effective drain radius) 9.8 m = D (depth of barrier below drains) 153 m = S (estimate of drain spacing) CALCULATIONS: 0.064052 = D/S 3.455722 = α Equation 9.5 Hooghoudt's equivalent depth for 0 < D/S < 0.31, or D/S = 0.31 6.710662 = D e 10.39937 = D e Hooghoudt's equivalent depth for D/S > 0.31 6.710662 = D e Hooghoudt's equivalent depth for conditions of this design h0 K(D e +h 0/2)t h/h 0 t Buildup h IRRIG. (days) (m) (m) (m) NO. f S2 9 68 1.527 0.1158 0.374 0.571 68 0.000 0.571 0.1084 0.402 0.230 SM 0.278 84 0.508 0.1333 0.315 0.160 1 0.256 27 0.416 0.0426 0.771 0.320 2 0.256 18 0.576 0.0287 0.884 0.509 3 0.256 16 0.765 0.0259 0.909 0.696 4 0.256 16 0.952 0.0262 0.906 0.862 5 0.256 16 1.118 0.0265 0.903 1.009 6 0.256 16 1.265 0.0268 0.901 1.140 7 0.256 17 1.396 0.0287 0.884 1.233 8 0.256 19 1.489 0.0323 0.853 1.271 9 0.256 1.527

Figure 9.10. Dynamic equilibrium at a drain spacing of 153 m for this example.

312

Chapter 9 Drainage Systems

9.4.2 Interceptor Drains Interceptor drains are single drains installed nearly along a ground surface contour at a depth to intercept subsurface water moving downslope from higher lands. In contrast, relief drainage assumes that all the water removed by the drains originates from the soil surface above the system of drains and that there is no lateral movement of water into the area drained. Interceptor drains can be open drains or buried pipes. Both kinds are equally effective if functioning as designed. We prefer the buried pipe interceptors. The main advantages of buried pipe interceptors are no loss of farmed area, no restriction to equipment movements, no weeds, and low maintenance. Open ditch interceptors introduce problems with weeds, instability of banks, restrictions of farming operations, loss of significant farm area, and difficutly of maintaining hydraulic grade. The depth of the drains and the proportion of the subsurface flow cross-section intercepted by the drain determine the amount of water that will be collected by the drain. The principal design factor for interceptor drains is the quantity of water entering the drains, which determines the best pipe size. The depth and location of the drain is controlled by the local site conditions, particularly the hydraulic conductivity of the soil, the slope of the groundwater table in the vicinity of the drain, and the depth from the water table to any low-permeability (restrictive) layer in the soil profile. Darcy’s law is used for the design of interceptor drains. Consider a vertical section aligned with the direction of flow of water to be intercepted. The original flow rate under normal conditions is calculated for a unit width of soil (length of drain) in this vertical section from Equation 9.15. Q = Kd(dh/dl) (9.15)

where Q = flow rate, m3/day per m width of soil (length of drain) K = hydraulic conductivity, m/day d = depth from the water table to the restrictive layer, in m dh/dl = hydraulic gradient. The depth of the drain is selected based on soil profile characteristics and the kind of excavating equipment available for construction. If the low-permeability layer can be reached, the drain should be placed on top of the layer. The drain should not be installed in or under a restrictive layer. The distance of the interceptor drain below the original groundwater surface divided by the original total depth of the flowing water above the restrictive layer multiplied by the calculated original flow will be the flow per unit length of interceptor. That is, the interceptor can be considered to skim all water arriving between its depth and the original (pre-construction) groundwater surface. The flow per unit length of drain can then be used to choose a pipe size. The post-construction groundwater surface downslope of the interceptor will be parallel to the original groundwater surface and at the depth of the interceptor. The groundwater surface in the soil upslope from the drain will be at the original level beyond a distance of 10 times the depth of the drain below the original groundwater surface. The upslope drawdown curve to the drain has an approximate parabolic shape. The designer is referred to USBR (1993) for additional details on interceptor drains, including how to place additional drains downslope in an irrigated area requiring a combination of interceptor and relief drains. Keller and Robinson (1959) presented results of an extensive sand-tank flume study of interceptor drains that can be used as a guide for design. Their results include ways

Design and Operation of Farm Irrigation Systems

313

of estimating the increased seepage loss from a canal when an interceptor drain is installed near the canal. The interceptor drain actually induces more seepage from the canal than existed before the drain was constructed. As long as the proper size of drain pipe is used, the interceptor drain will protect the land downslope as intended.

9.5 DETERMINING DESIGN VARIABLES Drainage design is not an exact science because of the great variability that exists in natural soils. Frequently, the design variables needed can only be approximated, but every possible attempt should be made to use accurate and precise information. Rainfall is not constant and irrigations are not perfectly uniform, so drainage coefficients are not precise. Likewise, neither the soil hydraulic conductivity nor the depth to the impermeable barrier can be determined with a high degree of accuracy under any circumstances, so drainage design also requires the use of reasonable engineering judgment. 9.5.1 Water Table Depth As given in Section 9.3.1, the design depth to water table at the midpoint between a system of parallel relief drains is normally taken to be 1.0 m. The USBR (1993) specifies a design depth to water table of 1.2 m. The deeper depth gives a factor of safety in areas where salinity control is important. If a steady state drain spacing is calculated using a water table depth of 1 m, a drain spacing calculated by the transient method will be less, even though the physical variables used in the calculation are identical. The preferred method is to calculate the steady state drain spacing with a water table depth of 1 m and then use the calculated steady state drain spacing to calculate the maximum height of the water table at the time of recharge using a transient design equation. The fluctuations of the water table will be controlled by the drain depth and spacing and the recharge events, and not by the design method, steady state or transient, used to determine the drain spacing. 9.5.2 Drain Depth Drains should be installed on or above a low-permeability layer and in highly permeable layers that are at or near the proper depth. Beyond those guidelines, the depth of drains is determined primarily by the type of excavating equipment available. Plows and trenchers used to install subsurface drains have a limited depth capability, depending on their design. Open drains can be as deep as necessary, but to maintain stable side slopes, deep open drains take large areas of land out of service and create serious maintenance problems. Special trenchers have been manufactured to install buried subsurface drains to depths of 4 to 5 m, but they are uncommon. The USBR (1993) made an economic analysis of drain depth. Deeper drains can be installed at wider spacing, but excavation costs are higher. The balance between excavation costs, speed of construction, and cost of materials resulted in an optimum drain depth of approximately 2 m. Shallower drains will cost more because more materials will be needed for the closer spacing. Deeper drains will cost more because labor and excavation costs are higher. Local conditions may cause the true economic drain depth to be different. 9.5.3 Depth to Barrier In the design of subsurface drainage systems, one of the necessary variables is the distance from the drain to some layer in the soil that would be a barrier to vertical water movement. Since permeability of the soil is a relative value, any horizontal soil layer with a hydraulic conductivity less than a tenth to a fifth of the weighted average hydraulic conductivity of all the overlying layers that are below the water table is con-

314

Chapter 9 Drainage Systems

sidered to be a barrier. Work of Toksoz and Kirkham (1971a,b) can be used to evaluate this assumption of barrier depth. The designer may also use the appropriate spreadsheet in this chapter to evaluate the effects of different barrier depths on calculated drain spacing. In most drain spacing equations, the depth to barrier is replaced by an “equivalent depth” (see Section 9.4.1). Since the equivalent depth equation contains the drain spacing and the drain spacing equation contains the equivalent depth, the solution to the equations must be by trial and error. If the depth to the barrier is unknown from soil borings or well logs or some other source, the equivalent depth can be assumed to be 25% of the drain spacing. 9.5.4 Hydraulic Conductivity In subsurface drain spacing calculations one of the most important factors is the permeability or hydraulic conductivity of the soil. Hydraulic conductivity can be determined in the field by means of an auger hole test (USBR, 1993; Matsuno, 1991) or a shallow well pump-in test (USBR, 1993). In the absence of any other information, the basic intake rate of the soil can be used as an estimate of hydraulic conductivity. It should be remembered, however, that the basic intake rate applies to the highly structured surface soil, but the water flowing to the drains is moving through the deeper soil layers. 9.5.5 Drainage Requirements The drainage requirements and goals in humid and arid areas are different (see Section 9.1.1). In humid regions, excess water removal is the objective of subsurface drainage. The hydraulic conductivity of soils in humid areas tends to decrease with depth so that drains must be closely spaced at a shallow depth. The normal installation depth for humid area drains is 1.2 m and the selected design steady state water table height above the drains is 0.2 to 0.3 m. The design depth to the midpoint water table is therefore about 1 m. Local experience is used to decide on the drainage coefficient, which is in the range of 7 to 8 mm per day. Humid area soils should not be overdrained, i.e., drained to such depths that the plants become short of water between precipitation events or irrigations. The depths below the water table can be used as a soil water storage reservoir. Organic soils should not be drained below a depth of 0.5 m to protect the soils against accelerated oxidation. In arid areas, the primary purpose of subsurface drainage is to keep the water table well below the soil surface so that excessive movement of salts from the water table to the root zone or the soil surface is avoided. In fine-textured soils, water can move upward by capillarity a distance of 2 m or more, so to avoid salinity problems drains must be installed at a depth of at least 2 m. In undrained arid areas, natural water tables should be below 3 m. In arid areas that are irrigated throughout the year or throughout the full growing season, drains can be installed at depths shallower than 2 m, but the land should not be allowed to stand fallow for any significant period of time or the surface will become salinized from capillary rise of water. With relatively frequent irrigations, salts will be leached downward and will not accumulate on the soil surface. The drainage requirement can be calculated using Equation 9.1 to determine the proper drainage coefficient for use in drain spacing equations. 9.5.6 Drainable Porosity A saturated soil, in the absence of restrictive drainage conditions, will drain to field capacity. The difference in water content of soil between saturation and field capacity is the drainable porosity. If the drainable porosity of the soil is 10%, adding 10 mm of

Design and Operation of Farm Irrigation Systems

315

water to a soil at an equilibrium field capacity water content will cause the water table to rise 100 mm. A better term to use in the design of drainage systems is specific yield. The definition of specific yield is the volume of water released by a unit area of soil for a unit drop in the water table. Specific yield changes with depth to the water table until the water table is deeper than about 3 m. Specific yield is correlated with hydraulic conductivity for normal drainage depths and the USBR (1993) has published a curve that can be used to estimate specific yield when the hydraulic conductivity is known. Field measurements of drain discharge as the depth to water table changes can also be used to estimate specific yield. The most common laboratory method for determining the specific yield requires interpretation of a soil water retention curve for the soil in question.

9.6 DRAINAGE SYSTEM MATERIALS Drainage design is not finished nor can discussion of the design procedure be complete without some consideration of the materials from which the system is to be constructed. This section considers the tubing and pipe, envelope, outlet, and manholes for buried pipe drainage systems. 9.6.1 Tubing and Pipe The first pipes used for agricultural subsurface drains were short lengths (0.3 m) of unglazed fired clay pipe with an internal diameter of 0.1 m. Since the pipes were made in the same kilns as other types of clay tiles, they were called drain tile and drains constructed of these rigid pipes were called tile drains. The reader is referred to Weaver (1964) for a more complete history of tile drains in the U.S. The pipes were merely butted together in the bottom of a trench and depended on the irregularity of the ends of the pipes to admit the drainage water. Standards were developed for strength and the pipes were used successfully for many years. Concrete pipes also have been used for subsurface drains. They were first made in 0.3-m lengths with square ends and were called concrete tiles. Later, special concrete drain pipes were made in approximately 1-m lengths with tongue and groove or bell and spigot joints that fit loosely to admit water. In the 1960s, corrugated and perforated plastic tubing, made primarily of high density polyethylene in the U.S. and polyvinyl chloride in Canada and Europe, was developed and was rapidly accepted as a substitute for clay tile and concrete pipe as subsurface drains. Today, concrete pipes are normally used for buried subsurface collector drains larger than 0.3 m and corrugated plastic tubing is the dominant material for smaller-diameter collector drains and laterals. Clay and concrete pipe are rigid pipes and have strength and loading characteristics that are very different than the flexible corrugated plastic tubing. Corrugated plastic tubing (called simply plastic drain in the following) is classified as flexible even though it feels rigid. Plastic drains must be carefully bedded to prevent being crushed by the weight of the backfill in the drain trench. If a gravel envelope is not used, plastic drains should be laid in a specially formed 90° or semicircular groove (with the same diameter as the outside of the drain tubing) made in the bottom of the trench. Otherwise, the tubing may be deformed or crushed. The full resistive strength of plastic drains can only be attained with the proper bedding described above. Placing a plastic drain in a flat-bottomed trench with loose material along the sides as backfill can result in flattening or failure of the drain. Surrounding plastic drains with a fine gravel or coarse sand envelope will provide the drain with adequate structural support in a flat-bottomed trench.

316

Chapter 9 Drainage Systems

There are ASTM standards for both rigid pipes and flexible drain tubing. The standards provide specifications for strength and durability and protection of plastic drains from ultraviolet light damage. The USBR (1993) and the NRCS (2001) have drainage handbooks containing additional information about pipe and tubing strength and installation requirements. Studies in Europe (Dierickx, 1992) have shown that the open area for entry of water into drain pipes or tubing should be 20 cm2 per meter length as a minimum and that open area up to 50 cm2 per meter improves drain performance. Plastic drains should have at least one perforation in every corrugation. The advantages of plastic drains over clay or concrete pipe are light weight, ease of handling, lack of loose joints to become misaligned, low cost, and durability. Disadvantages are its low structural strength without proper bedding, susceptibility to crushing from point loads during backfilling, and greater hydraulic roughness than rigid pipes. Any pipe that will not deteriorate while buried in the soil can be used as a drain pipe if it has adequate open area for the entry of water. Pipe perforations must be small enough to prevent the entry of sediment if the pipe is not protected by synthetic or natural envelope materials. 9.6.2 Envelope Porous materials such as coarse sand or fine gravel are placed around drains in unstable soils to prevent movement of soil particles into the drains. These materials are called drain envelopes. Early in the history of buried pipe drainage these envelopes were called filters. A filter is exactly what we do not want to place around our drains. Since a filter is designed to trap particles, it will clog over time and prevent water movement into the drain. Conversly, a well-designed envelope of properly graded granular material will naturally “bridge” to prevent most soil particles from passing through but will allow very fine soil particles into the drain. These very fine particles will be transported on out through the system with the normal water flow. The NRCS (2001) and USBR (1993) have published criteria for the design of granular material envelopes for drains. Soils (particularly fine sands and coarse silts) in arid regions are frequently structurally unstable and thus most subsurface drains require the use of envelopes. Drain envelopes will always improve the hydraulic functioning of a drain. Soils in humid areas tend to be structurally stable so that drain envelopes are not normally necessary. In recent years, various synthetic fabrics called geotextiles have been used for drain envelopes. While they may be cheaper and easier to install, they are not as effective as granular materials. Where success with geotextiles has been reported, probably no envelope was needed. We recommend envelopes of properly graded granular materials. 9.6.3 Outlet As stated in Section 9.1.1, the outlet(s) is the discharge point(s) of the drainage system where the water enters an existing channel or waterway. Small drainage systems may have only one outlet but larger, more complex systems usually have multiple outlets. The importance of the outlet should not be underestimated. The adequacy of the entire drainage system depends on a properly designed, constructed, and maintained outlet. Gravity flow of all water from the drains to the receiving bodies of water is usually most desirable. When this is not possible or is difficult to do, sumps equipped with water-level controlled pumps may be used to lift the drainage effluent to the proper elevation.

Design and Operation of Farm Irrigation Systems

317

Regardless of whether the drainage system uses sump pumps or discharges by gravity flow, the last section of buried pipe should be solid rigid pipe that extends over the receiving body of water. This pipe is necessary to prevent damage to the system from livestock and/or wildlife. Usually the last section of pipe is specified as corrugated metal culvert pipe of suitable diameter. To protect the outlet pipe from encroachment by muskrats or other animals, flap valves, grates, or other devices may be installed at the exit end of the pipe (NRCS, 2001). We recommend that the invert of the outlet pipe be at least 0.3 m above the high-water stage of the waterway. 9.6.4 Receiving Bodies of Water Although formally not a part of the drainage system, the existing waterway or other receiving body of water must be carefully considered. The stage-frequency relationship of the waterway must be known, as well as its capacity to carry the drain discharge together with its normal flow. The quality of water in the receiving water body must be such that the drainage water can be safely discharged into it. Subsurface drainage outlets have been classified as point source pollution if they damage the water quality of the receiving stream. In determining the effective elevation of a proposed drainage outlet, the elevation of the water surface throughout the year should be determined to eliminate the possibility of submerging the drain system outlet at a time when drainage is most needed. 9.6.5 Manholes Manholes in subsurface drainage systems serve the same purpose as manholes in a sewer system; they provide access for inspection and cleaning. As a general rule, manholes should not be used in a subsurface drainage system unless absolutely necessary. Manholes interfere with cultivation if they are placed in a field. If they are left open they can be dangerous to children and animals, and are entry points for trash that may clog the drainlines. Manholes are useful at points where collector systems come together so that performance of individual systems can be observed. Rather than installing manholes at regular intervals, it is more economical to dig down to a drain with a suspected problem.

9.7 CONSTRUCTION METHODS AND EQUIPMENT Subsurface drains are installed in trenches that may be hand dug or opened by a variety of digging machines. Most drains are installed by mechanical trenchers designed and equipped especially for drain installation. Fouss (1974) shows photographs of drainage equipment which has not changed much in recent decades except for more widespread use of laser grade control. Fouss shows a modern chain-digger trencher equipped with a shield to prevent collapse of the trench before the pipe can be installed in an unstable soil, and a drainplow that can install drains to depths of 1.5 m without making an excavation. (Drainplows are uneconomical for installations deeper than 1.5 m.) Trenching machines are classified according to the digging method used: a wheel, flexible ladder, or digging chain. The chain-type trencher is used for narrow trenches common for small diameter lateral drains. Wheel and ladder trenchers are used for the wide trenches needed for large diameter drains. Drainplows are used for small-diameter drains installed with or without geotextile envelope materials, but not with gravel/sand envelopes. Trenches can be opened with a backhoe or other type of excavating machine, but extra effort is then needed to bring the bottom of the trench to the necessary firm uni-

318

Chapter 9 Drainage Systems

form grade. Drain pipes should be laid straight or on gently curving uniform grades without vertical deviations greater than 15% of the pipe diameter. Vertical deviations (humps or valleys) in a drainline reduce the flow gradient and make a deadwater area upslope of the hump where sediment can collect and reduce the flow capacity of the pipe. Depressions in a drainline also collect sediment and reduce flow capacity. Overexcavation of a trench and bringing the trench to grade by throwing in loose material may result in vertical grade deviations, pipe misalignment, and structural failure of the pipe. Thus, bringing an over-excavated trench back onto grade should be done only with gravel material—an expensive process. The foregoing discussion points to the need for good grade control during the construction of drains. Controlling grade with a laser is now commonly used with drain plows and drain trenchers to assure adequate control of grade and alignment.

9.8 DRAINAGE SYSTEM OPERATION AND MAINTENANCE Subsurface drainage systems with gravity outlets operate continuously and automatically with minimum maintenance. If the outlet is a sump with a pump, then continuous maintenance of that part of the system is necessary to assure adequate drainage. The operation of the system is fixed at the time of the design and consists of the design rate of water removal or the rate of lowering of the water table at the midpoint between drains. If the drain flows respond to irrigations or rainfalls in excess of the water storage capacity of the soil, the drains are assumed to be operating properly. The appearance of new wet spots near a drainline in a field is an indication that the drain is clogged at that point. The drain can then be excavated, examined, and repaired. The crop can be watched for indications of inadequate drainage, such as salt accumulations on the soil surface or differences in crop growth (poor growth in wet periods or good crop growth in dry periods). In periods of water shortage, plant roots can invade and clog the drains. Sugar beet, cotton, willows, poplars, and deep-rooted desert plants can clog drains in a relatively short period during times of water shortage. Use of rigid drain pipes is recommended if such conditions are expected. Maintenance of buried subsurface rigid-pipe drains is primarily accomplished by high-pressure water jetting machines similar to sewer cleaners used in the city (Grass et al., 1975). Sections of drains up to 200 m long can be easily cleared of roots, sediment, and iron ochre deposits by jet cleaning. Pump pressures are approximately 4 MPa. The drains are cleaned in sections from excavated access holes. After an appropriate plug is placed into the downstream section of drain at the access hole, cleaning is done in an upstream direction from this plug and the water and debris are pumped out of the access hole. When cleaning of this section is finished, the drain pipe and envelope are repaired and the access hole is filled. Cleaning costs are usually 20% or less than the cost of installing a new drain.

9.9 DRAINAGE SYSTEM PERFORMANCE EVALUATION Drainage system performance criteria are fixed at the time of design. Drain depths are set by the capacity of the available digging machines, the requirements of the area, and the stratification of the soil. Drains should not be installed in or below soil layers that are slowly permeable. The minimum allowable depth to the water table at the midpoint between drains is fixed by local experience and is usually approximately 1 m. The drainage requirement or design drainage coefficient determines the rate of water removal necessary and then the soil hydraulic conductivity is used in a suitable

Design and Operation of Farm Irrigation Systems

319

equation to determine the appropriate drain spacing. If the soil drains acceptably, the performance criteria are assumed to be met. If the soil does not drain properly, further investigation will be needed to determine whether the drain is faulty or whether additional drains are necessary.

REFERENCES Dierickx, W. 1992. Use of drainage materials. In Proc. of 5th Drainage Workshop, Vol. I. New Delhi, India: ICID-CIID. Dumm, L. D. 1968. Subsurface drainage by transient-flow theory. J. Irrig. Drain. Div., ASCE 94(IR4): 505-519. Fouss, J. L. 1974. Drain tube materials and installation. In Drainage for Agriculture, 147-177. Monograph No. 17. J. van Schilfgaarde, ed. Madison, Wis.: American Soc. Agronomy Grass, L. B., A. J. Mac Kenzie, and L. S. Willardson. 1975. Inspecting and cleaning subsurface drain systems. Farmers Bulletin No. 2258. Washington, D.C.: USDAARS. Hooghoudt, S. M. 1940. Bijdragen tot de kennis van eenige natuurkundige grootheden van den grond, 7, Algemeene beschouwing van het probleem van de detail ontwatering en de infiltratie door middel van parallel loopende drains, greppels, slooten, en kanalen. Versl. Landbouwk Ond. 46: 5 15-707. Keller, J., and A. R. Robinson. 1959. Laboratory research on interceptor drains. J. Irrig. Drain. Div., ASCE 85(IR3): 25-40. NRCS. 2001. Chapter 14: Water management drainage. In Engineer Field Handbook, Part 650, 14-56. Washington, D.C.: USDA, Natural Resources Conservation Service. Matsuno, Y. 1991. Improved measurement of hydraulic conductivity in an auger hole. MS thesis. Pullman, Wash.: Washington State Univ. Moody, W. T. 1966. Nonlinear differential equation of drain spacing. J. Irrig. Drain. Div., ASCE 92(IR2): 1-9. Toksoz, S., and D. Kirkham. 1971a. Steady drainage of layered soils. I: Theory. J. Irrig. Drain. Div., ASCE 97(IR1): 1-18. Toksoz, S., and D. Kirkham. 1971b. Steady drainage of layered soils. II: Nomographs. J. Irrig. Drain. Div., ASCE 97(IR1): 19-37. USDI-BR. 1993. Drainage Manual. 3rd ed. Denver, Colo.: U.S. Department of Interior, Bureau of Reclamation. van Schilfgaarde, J. 1965. Transient design of drainage systems. J. Irrig. Drain. Div., ASCE 91(IR3): 9-22. Weaver, M. M. 1964. History of Tile Drainage (in America Prior to 1900). Waterloo, N.Y.: M. M. Weaver.

CHAPTER 10

LAND FORMING FOR IRRIGATION Allen R. Dedrick (USDA-ARS, Beltsville, Maryland) Ronald J. Gaddis (A B Consulting, Lincoln, Nebraska) Allan W. Clark (Clark Brothers, Dos Palos, California) Alan W. Moore (Cameron County Drainage District 5,Harlingen, Texas) Abstract. Land forming for irrigation is a key part of the land improvement process in all regions where crop production is dependent upon irrigation. The state of the art of land forming as it relates to irrigation development, particularly surface irrigation, is presented. Topics addressed include system layout, soil survey and allowable cuts, topographic surveys, earthwork analyses, types of equipment, field operating procedures, cost and contracting, and safety. Computer-aided design, including various computer software packages, and extensive changes in equipment for land forming and its control emerging over the past 25 to 30 years are included. Keywords. Irrigation, GPS, Land forming, Land improvement, Land levelers, Land leveling, Land shaping, Land smoothing, Lasers, Scrapers, Surface drainage, Surface irrigation, Water conservation, Water management.

10.1 INTRODUCTION Land forming for irrigation is a key part of the land improvement process with the goal of sustaining high crop productivity. It provides a land surface ideal for surface irrigation and drainage. Both irrigation and drainage systems require land surfaces that transport water sufficiently slowly to allow adequate infiltration and minimize erosion, while at the same time transporting water rapidly and uniformly to prevent injury to plants. The benefits of land forming are extensive (Springer, 1966; Lyles, 1967; Quackenbush, 1967; Buras and Manor, 1970; Sewell, 1970; Harris, 1971; Kriz et al., 1973; Thomas and Cassel, 1979; Shih et al., 1981; and Shih, 1992) and include: ƒ improving surface drainage by eliminating low areas and adding grade to areas of no slope; ƒ permitting more efficient use of irrigation water and making possible the irrigation of a larger area with a limited water supply;

Design and Operation of Farm Irrigation Systems

321

ƒ improving efficient control of water by providing for uniform distribution; ƒ reducing the number of surface field ditches required for good drainage since each furrow carries its own water; ƒ reducing erosion by preventing water from concentrating in natural drainage ways; ƒ reducing herbicide and fertility losses because of better control of water; ƒ reducing ditch maintenance by minimizing the number of deep ditches; ƒ saving machinery-operating time in turning around at the end of large fields as compared to small fields; ƒ increasing the cropping area by removing ditches and fencerows; ƒ increasing machinery efficiency because point rows are eliminated and machinery can be operated at higher speeds; ƒ increasing the precision of inter-row tillage; ƒ increasing the accuracy of height or depth of the cutting edge of harvesting machines; ƒ reducing wear and damage to equipment working in the field; ƒ providing better crop and soil management because the uniform ground surface allows precision planting and fertilizer application, which results in more uniform germination and crop maturation; ƒ increasing crop yield because of better soil, water, and crop management; ƒ reducing organic soil subsidence by maintaining a more even water table depth within a field; ƒ reducing excessive organic soil mineralization and nitrification by eliminating the high spots; and ƒ saving energy in paddy rice production by reducing amount of water pumped for flooding the fields. The benefits touch on all aspects of productivity. Further, the list addresses issues pertaining to all climatic regions, i.e., irrigated areas where production is dependent upon irrigation (arid, semi-arid regions) as well as areas where irrigation is supplemental to precipitation. In irrigated areas with higher precipitation, land forming is extremely important to control surface drainage and improves the uniformity of water application by surface irrigation. In some instances, the ideal land forming for surface drainage and surface irrigation systems have conflicting requirements: flatter slopes for surface irrigation systems generally result in higher irrigation efficiencies, but drainage may be inadequate. Irrigation systems (whether surface, sprinkle, or micro) are selected and designed considering various factors applied to specific site conditions. These factors include: ƒ water supply—source location, delivery flow rate, quality, delivery schedule, delivery point; ƒ crop characteristics—effective root zone, growth characteristics, cultural practices; ƒ soil features—water-holding capacity, intake rate, depth; ƒ topographic features—surface irregularities, steepness, direction of slope; ƒ geographic features—field shape, natural drains, buildings, utilities or obstructions; ƒ labor—amount required, quality, quantity available, cost; ƒ energy—amount required and available, associated cost;

322

Chapter 10 Land Forming for Irrigation

ƒ system cost—hardware and other ancillary equipment, installation, operation, maintenance; ƒ farm equipment—type commonly used and/or available; and ƒ farmer’s preference. All irrigation methods may require some land forming to attain maximum production, depending mostly on the climatic region and irrigation method used. However, the performance of surface irrigation methods is impacted most by the field surface topography. To guide the process of land forming, topographic features influenced by the soil depth and type are paramount. Depending on the irrigation system selected, the field may be graded with a single or a series of slopes in the direction of irrigation, and either with or without side fall, i.e., field slope other than that in the direction of irrigation (discussed further in Section 10.2). As used herein, land forming (or shaping) refers to both land grading and land smoothing. Land grading refers to the movement of soil associated with meeting the general slope requirements of the irrigation system design, while land smoothing (or finishing) generally refers to the removal of nonuniformities in that soil surface. The average finished slope is a statistical measure of how accurately the field was graded to meet the design slope. Land smoothing can be statistically described by the precision of the field elevations around the design slope (i.e., the standard deviation of the field elevations from the design slope). Land grading and smoothing improve the distribution uniformity of irrigation water. This improved uniformity provides an opportunity to improve the irrigation application efficiency of the system. Generally, uniform irrigations improve the resultant crop yield, because of the potential negative impacts of low and high areas. Low areas are generally areas of over-application, resulting in poor soil aeration, associated oxygen deficiencies, and potential excess leaching of nutrients. High areas are generally areas of under-application, thus resulting in insufficient water to meet crop needs and leading to potential build-up of salts. Many times irrigators will apply additional irrigation water to sufficiently irrigate the high areas, which potentially exacerbates the negative impacts associated with the low areas and lowering the overall irrigation efficiency. 10.1.1 Scope The objective of this chapter is to review the state of the art of land forming as it relates to irrigation development, particularly surface irrigation. Land forming associated with nonirrigated land is not addressed directly, although much of the information presented would apply. Topics addressed include system layout, soil survey and allowable cuts, topographic surveys, earthwork analyses, types of equipment, field operating procedures, cost and contracting, and safety. Computer-aided design, including various computer software packages, and extensive changes in equipment for land forming and its control have emerged in the past 25 to 30 years.

10.2 SYSTEM LAYOUT Before the land forming for a field can be designed, the farm irrigation system, drainage system, main field boundaries, roads, and other physical features must be planned. From soil information, topography, method of water application, crops to be grown, and the farmer’s own personal likes and dislikes, alternative field arrangements are developed. Field arrangement and design are important parts of the process of land forming. It is essential that planning be done with imagination. All of the practical potential layouts should be considered and the one best suited to the site selected.

Design and Operation of Farm Irrigation Systems

323

The planner should consider and include, when appropriate, the following items when developing the irrigation system layout: ƒ location and elevation of the water supply; ƒ amount and arrangement of supply ditches and/or pipelines, as well as water control requirements; ƒ location, number, size, and elevation of drainage ditches; ƒ location, size and method of operation of the return-flow system; ƒ type of farm equipment to be used and ease of access to and operation within all fields for that equipment; ƒ field location and arrangement, to minimize soil variations within each field; ƒ combining small fields into larger fields, especially if larger farming equipment is to be used; ƒ minimizing non-cropped area; ƒ ease of management of irrigation water and cultural operations; and ƒ current and future crops to be grown. Design criteria must be established before the field can be subdivided and the design for land forming prepared. Both irrigation and drainage affect the allowable length of fields. Excessively long fields may create erosion problems during irrigation and/or rainfall. Short fields require more delivery and drainage facilities for a given area. Efficiency of machinery operation depends on the type of machinery used, but generally increases with length of field. Distribution uniformity (DU), when used as the design criterion for surface irrigation systems, can be described as DU = f (I, n, so, L, Qin, Zn) (10.1) where I = infiltration characteristic, generally defined by a relationship between cumulative infiltrated depth and infiltration opportunity time n = flow resistance (Manning n) so = slope (or series of slopes) in the direction of irrigation of the area to be irrigated L = length of the furrow or border qin = unit flow rate, i.e., flow rate per unit width (Qin/W) or flow rate per furrow (Qin is the inflow rate to the field, and W is the width of the area being irrigated in a particular set) Zn = net depth to apply. Infiltration and flow resistance are characteristics inherent to the irrigation site. At the time of design, the other four factors (slope, length, unit flow rate, and net depth to apply) are adjusted within various constraints to provide an irrigation system that will meet the targeted design performance criteria. The design slope in Equation 10.1 reflects land grading requirements. In the U.S., irrigation and drainage guidance for the particular type of surface irrigation system(s) being considered, i.e., how the last four factors of Equation 10.1 are combined, can be obtained from various sources such as the USDA-Natural Resources Conservation Service (formerly USDA-Soil Conservation Service), Cooperative Extension Service, universities, USDA-Agricultural Research Service, or private consultants. These design guidelines, when combined with soils, depth of water to apply, and flow information, provide optimum field slopes and lengths to achieve the targeted design criteria. Additional information on surface hydraulics is given in Chapter 13 and design procedures are given in Chapter 14.

324

Chapter 10 Land Forming for Irrigation

10.3 SOIL SURVEY AND ALLOWABLE CUTS Land forming should only be done after first knowing the soil profile conditions and the maximum cut that can be made without permanently affecting agricultural production by exposing infertile and/or rapidly draining subsoils. General soil survey maps are available for many areas and, if not available, the services of a soil scientist should be obtained. In the U.S., soils information may be obtained from the local Natural Resources Conservation Service (NRCS) office, local Cooperative Extension Service, or a state university. The standard soil survey generally delineates areas larger than 2 ha. The soil profiles are described to a depth of 1.5 m. In alluvial soils, sand and gravel pockets may occur within a soil-mapping unit of some other soil type. A detailed site investigation should be made to determine the extent of such material before the land grading begins so that adequate fill material can be made available. Sand and gravel pockets that occur near the water inlet end of a surface irrigation system may greatly limit the irrigation efficiency because of high intake rates. It is often advisable to remove these sand and gravel materials and replace them with soil similar to the rest of the field. If there are doubts as to the soil depths, investigation with an auger in several representative locations within the land area is mandatory. Soils with deep, well-drained subsoil generally have few limitations on depth of cut. Some shallow soils may be suitable for irrigation, but may not have sufficient depth (e.g., shallow soils over bedrock, gravels, caliche, heavy claypan, or fragipans) to permit the necessary land forming needed for surface irrigation, while they could be successfully irrigated with sprinkle or microirrigation systems. Cross-slope benching is sometimes used on sloping land to reduce the depth of cut needed to obtain suitable field grade and can reduce the cost of land forming (Dedrick, 1989).

10.4 TOPOGRAPHIC SURVEY It is extremely important that the entire farm be studied before any grading work is attempted, even though land grading is usually done on a field-by-field basis, because the most economical design for an individual field may be undesirable when the entire farm system is considered. The factors noted in Section 10.2 should be considered by the designer in planning the overall farm development. A general topographic map of the entire farm showing the location of all physical features can provide much of this information. The designer must select the irrigation layout; the most desirable location for the various elements of the farm irrigation, water reuse, and drainage systems; and location of field boundaries and field roads. By presenting this type of information, several alternative designs can be developed for consideration and selection by the farmer or farm manager. The selected plan should permit land grading to be carried out on individual fields, even over a period of years, if necessary. Planning is aided by development of a detailed topographic map. When collecting data for the map, consider: ƒ What is the size of the area to be planned? ƒ What are the availability and capability of personnel? ƒ What are the time constraints for getting the job done? ƒ How will the topographic data be analyzed, i.e., how will field slopes, earthwork quantities, and haul distances be calculated? ƒ Does the type of grading and smoothing equipment and the procedure used during the grading and smoothing operations dictate a certain topographic layout? ƒ Will grid stakes be required for the grading operation?

Design and Operation of Farm Irrigation Systems

325

ƒ What type of surveying equipment will be used to collect the data? ƒ How will the topographic map be generated, by hand or by computer? In simplest terms, the topographic data collected includes the elevation, z, of variously located x and y coordinates, including boundaries to the area. The survey data collected in the field will be structured either according to a uniformly spaced, gridbased pattern or irregular spacing based on documenting the extremes and changes of the surface. The choice of data collection pattern will depend on the general smoothness of the area, the method to be used to analyze the data, type of grading equipment to be used and associated design required, and how the field surfaces will be mapped. Highly accurate global positioning systems (GPS) have been introduced to the land forming industry during the past 10 to 15 years. In one system, the elevation information is obtained from a laser transmitter/receiver system and the x-y coordinates from the GPS. A second system gets both the position and elevation from the GPS. Both systems can be used with either grid-based data collection or irregularly spaced patterns. 10.4.1 Grid-Based Pattern 10.4.1.1 General description. The grid-based pattern is best suited to areas that are relatively smooth. If, during construction, the method of grading is to be based on lanes and controlled by periodic cut/fill stakes, a grid layout has the advantage of defining the lanes. If the method of grading depends on manual control, grid locations must be preserved until construction when these same locations will be used for the cut/fill stakes. If laser-controlled scrapers are used for grading, only a few points of known elevation are required and systematically located cut/fill stakes are not needed. Information needed regarding the finished slope will depend on the type of laser hardware used. The main disadvantage of the grid pattern is that critical high and low points and changes in slope do not often occur at grid points, and thus they are missed in data collection and analyses leading to inaccurate surface portrayal. For ease in computing volume and area, especially when not computer aided, the grids should be made square; however, stakes may be set at any spacing desired by the designer. The most commonly used spacing is 30 m. Larger spacing makes it more difficult for the operator of manually controlled grading equipment to maintain the grade from one stake to the next. Closer spacing greatly increases the work of staking, surveying, and computing the grading design. Occasionally, it may be necessary to locate additional points in the grid pattern to record significant topographic changes in the land surface or to show irregularly shaped boundaries. If closer spacing or numerous additional points are needed to describe the topography, irregularly spaced data collection may be more appropriate. Grid-based elevation readings are either taken in the center of each square or at the corners. Each data set type requires a different method of analysis. An elevation reading taken in the center of each square is assumed to represent the mean elevation of that entire square. The field surface analysis resembles a checkerboard with each square at a different elevation. Exact elevations around the border of the surface are unknown and either must be estimated or can be defined by adding readings at a few locations. The field layout is simple, usually adequate for flat, uniform surfaces, and lends itself readily to simplified approximations with regard to slope design and earthwork calculations. When the elevation readings are taken at the edges of the squares, a more accurate surface analysis model (based on triangulation) is possible since each square actually

326

Chapter 10 Land Forming for Irrigation

defines two triangles. This type of grid data can be reduced to the first type of grid representation if one of the simple approximation methods is to be used for design of land forming. 10.4.1.2 Staking the field. Wood laths (10 mm × 45 mm × 1.2 m) are excellent stakes for setting the grid points. They are easily set by the technician from a standing position, and cut or fill markings near the top of the lath are easily seen by the equipment operator. The laths should be sharpened and driven into the ground far enough so that they will remain standing until the grading work is completed. A procedure is outlined by Anderson et al. (1980) to guide the complete staking of a field. The procedure entails the establishment of key rows of stakes across the ends of the field from which the remaining grid stakes can be set by eye. For larger fields it may be desirable to establish the guide rows near the center of the field to eliminate some of the errors that occur in setting long lines by visual sighting. 10.4.1.3 Survey equipment and techniques. Since the x-y coordinates are defined by the grid layout, the elevation is the only remaining unknown. The elevations are usually determined with an engineering level and rod or a rod with a laser receiver. To facilitate field readings, the rodman may use an all-terrain vehicle or pickup truck to move quickly from reading site to reading site. In some cases, the laser-control equipment attached to the tractor-scraper unit is used. 10.4.2 Irregularly Spaced Pattern 10.4.2.1 General description. Irregularly spaced patterns can be used to document the field surface. Irregularly spaced data points are useful in mapping the area to be irrigated as well as to map important planning features outside the irrigated area. No prior layout is required. The location of data points is selected in the field and should reflect various controlling features or break lines. Break lines are topographic features that have more or less uniform slope, such as natural and man-made surface drains, streams, lakeshores, roads, railroads, ditches, and ridgelines. Numerous data points can be used to adequately represent any feature. On uniform portions of the ground surface, the distance between readings can be increased without compromising the integrity of the data analysis model. 10.4.2.2 Surveying equipment and techniques. Analyses of irregularly spaced data can result from any surveying technique that provides both the location (x-y coordinates) and the associated elevation readings. Typical equipment used includes the transit, plane table and alidade, automated total station, aerial photographs with stereo mapping, and laser-GPS and GPS surveying systems. If accurate and detailed contour maps for the area already exist, electronic digitization can be used to develop a database for reproducing the topographical map and for calculation of land grading . Transit. The transit provides both location and elevation. The surveyor uses the stadia readings to determine distance from the instrument and, with the horizontal and vertical angles, to calculate the coordinates of a point. Electronic distance measuring equipment (EDM) can be used instead of the stadia. EDM is rapid and eliminates some hand calculations and associated errors, especially when combined with handheld, battery-powered electronic notebooks or data loggers to register and store data, thereby reducing the need for hand recording all survey records. Plane table and alidade. The plane table and associated alidade are best suited for situations where the contour map itself is the end product. The method is of limited use where data are needed for land grading design. Often, an engineering level is used

Design and Operation of Farm Irrigation Systems

327

with the plane table and alidade to supplement the data collection by determining the data point elevations. The alidade is used to sight the rod. Stadia crosshair readings are used to determine the distance to the sighted point. The rectangular base (also a ruler or scale) of the alidade rests directly on the map as it is drawn. The line of sight and the scaled horizontal distance to the rod are plotted by hand to locate the x-y coordinates of each data point as the field is being surveyed. Location of the point replaces the recording of the horizontal angle. The plane table and alidade can be used to develop contour lines directly by fixing a target on the rod at a height that will result in the elevation of a desired contour line. A series of irregularly spaced points, found to be at the desired elevation using the level, are located and drawn on the map with the alidade. Connecting the points draws the contour lines. Total station. A popular, current surveying technique involves the use of the automatic total station. This apparatus is a battery-powered, computerized transit using EDM technology that also performs various data calculation, recording, and transfer functions. The total station electronically registers the x-y distances and elevations of points selected by a rodman carrying a mirror-like prism. The instrument operator transmits a beam of light to the prism enabling the instrument to measure horizontal and vertical angles (like a transit) and the distance from the instrument to the target. The 3D coordinates of each selected ground location are electronically calculated and stored, later to be transferred for data analysis. It is often possible to survey a very large area from a single location since such equipment functions accurately over long distances, with actual distance depending on atmospheric conditions. Aerial mapping. Aerial photography with stereo plotting is best used when a large area (several thousand hectares) is to be planned as a unit or where work is to take place on several large adjacent farms at the same time. The photography should be done during the season of low vegetative growth and the maps plotted, as they are needed. This procedure requires that vertical and horizontal control points be established before the photos are taken. These points, which must be visible on the photos, usually are made with white plastic strips in the form of a cross. Low-level photographs are taken from an airplane. The height of the plane above ground level will depend on the camera used and the detail required for the topographic map. For example, a 150-mm focal length lens in the camera at about 900 m to 1200 m above the ground will provide the resolution needed to plot 0.6-m contour lines. The topographic map is made with a stereoscopic plotter using stereo pairs of photos. While excellent for general planning purposes, the individual areas to be graded will require additional surveying to provide accurate data for grading design calculations. Laser-GPS system. This system combines the use of laser and GPS technology to provide survey information similar to that provided by the total station. In addition, such surveying systems are coupled with computer-aided land leveling software that produces various design and display information that is used directly with land forming equipment and/or as support material for the equipment operator. GPS surveying system. This recently introduced system performs all of the functions of a total station utilizing the GPS to obtain both the field coordinates and the elevation. 10.4.2.3 Digitization. Digitizing provides a means of transforming a contour map from any source (including plane table and alidade data) into a database for analysis.

328

Chapter 10 Land Forming for Irrigation

This can be done either manually or using a digitizer. If done manually, either a grid is established prior to the survey so that the x-y coordinates are known, or the contour map of the area to be irrigated is drawn using irregular survey points as needed. A clear plastic overlay, marked with grid locations, is superimposed on the map enabling the elevations to be interpolated. Digitizing technology registers the location of each x-y coordinate of interest from maps or data points. The elevation is entered either by defining a series of x-y coordinates on a contour line or by entering individual elevations for the selected points. When using digitized contour lines in surface computations, the user needs to insure that points on one contour line triangulate with those of an adjacent contour. Computer software exists that allows such user intervention. Further, the user must prevent three points on the same contour from forming a triangle, which would then erroneously form a level triangle. 10.4.3 Digital Elevation Models (DEMs) The availability of computers with the capability to rapidly process large quantities of data has made hand or calculator analyses almost obsolete. The same skills once used to carefully draw maps and make various earthwork calculations are now applied toward managing inexpensive computer software to rapidly produce, revise, and refine contour maps and system design options. The digital representation of the topographic survey data, i.e., an array of points whose horizontal positions are given by their x and y coordinates and whose elevations are given as z coordinates, are known as digital elevation models (DEMs) or digital terrain models (DTMs). Generally, when an irregularly spaced DEM is utilized, a triangulated irregular network (TIN) surface model of the area is formed. The TIN model is constructed by connecting points in the array to create a network of adjoining triangles. (The same procedure can be applied to grid-based data.) Each of the triangles is formed by lines joining together three of the actual survey points, one at each corner of the triangle. Creation of a network of “nearequilateral” triangles is a criteria commonly used in the development of TIN models. Further, the TIN depends on proper identification of break lines. Break lines must be identified to the computer in the input array. Once the break lines are appropriately identified, the break lines become sides of triangles in the network of triangles created when developing the TIN model. Information from this network of triangles can then be used in all subsequent data analysis and display. 10.4.4 Automated Contouring Systems Available contouring software allows a topographic map to be drawn by technicians with no experience with hand-drawn contours. Automated contouring systems generally make two assumptions concerning TIN models: (1) each triangle side has a constant slope, and (2) the surface area of any triangle is a plane. Based on these assumptions, elevations of contour crossings are interpolated along triangle edges. After review of the initial contour map, the user may modify any portion of the resulting triangulation as needed, and thereby insure that the surface model accurately simulates the actual ground surface. Although the contouring program forms the surface model automatically, most programs incorporate some means by which the user designates the boundary of the surface to be mapped. In some cases, the boundary is assumed by the program and can be altered if it is incorrect. Various mathematical techniques are used to approximate field contours. Data interpolation is a common technique that involves interpolation between adjacent grid points to determine the coordinates of any contour line that may pass between the

Design and Operation of Farm Irrigation Systems

329

points. Without smoothing, however, this method may produce erratic values and contours that are difficult to interpret. Polynomial regression has been used (Portier and Shih, 1982) to estimate parameters for a mathematical model that produces a smooth, continuous surface with well-behaved contours. A disadvantage is that none of the original observations are likely to be found on the fitted surface, however, this may not be a problem if the goal of the analysis is to obtain a general impression of the field for making decisions. The software utilized for developing topographic maps generally consists of a data editing program, a contouring program, and a computer-aided design (CAD) program. Within the data-editing program, the data are screened to exclude points such as benchmarks and other references that are not on the field surface. A 3D model of the ground surface is formed within the contouring program. From this model the contour map is then drawn according to the interval selected by the user. Usually, topographic maps with contour lines using 0.3-m intervals are adequate for description and design. On very flat or irregular topography, 0.15-m increments may be desirable. When the slope exceeds 1%, the planner may desire a contour interval of 0.6 m or greater to have a more readable map. Contouring software is available that enables designers to produce 3D views of the present (existing) terrain and as it will appear after the proposed modification. These views can be extremely helpful in assisting landowners and/or project officials to visualize, and assess, the various development alternatives. The contour map can be output to a computer printer, a plotter, and/or a CAD program. It is important that the topographic map illustrates other physical features of the farm. Usually within a separate CAD program, the designer combines the contour map with other data that show the location and elevation of benchmarks, location and source of irrigation water, existing field boundaries, drainage patterns and outlets, farmstead, farm roads, location of both buried and aboveground utilities, and any other physical features that may affect the planning of the system and the design of the land grading for each field.

10.5 EARTHWORK ANALYSES Design and analysis for land forming include processes for defining the finished field surface, adjusting for cut and fills to compensate for shrinkage of the soil, and making the necessary earthwork calculations. At the present time, computers are used nearly exclusively to optimize cuts and fills. Powerful, easy to use, and relatively inexpensive software for land forming design is commercially available from vendors of surveying software. Many consulting engineering firms, federal agencies, and state universities involved in irrigation development have such programs. Most commercial software is able to handle irregular field shapes, irregularly spaced data, and other complex design situations. Cut-fill maps and an analysis of the work and haul patterns required to achieve the design are important options now available. The following discussion is intended to provide an overview of design procedures commonly used in design and analysis of land forming, supplemented with an ample number of references. A limited number of computational equations are given to provide examples of how certain issues are addressed. In most instances, numerous procedures are available for solving various aspects of the earthwork analysis process; it is left to the interested readers to decide which procedures best accommodate their needs.

330

Chapter 10 Land Forming for Irrigation

10.5.1 Defining the Finished Field Surface As noted earlier, the survey data collected can be according to either a grid-based or irregular pattern. Generally, a mathematical description of the design surface is developed that meets the functional requirements of the land forming operation. An equation for a linear plane is used to describe the linear land-grading plane where the elevation, H(x,y), at any point (x,y) is given by H ( x, y ) = a + bx + cy

(10.2)

where a is the elevation of the plane at the origin of the x and y coordinate system, and b and c are the two edge slopes of the plane in the x and y directions, respectively. The best-fit linear plane can be determined by linear regression analysis techniques. Best-fit approaches are useful, especially, for drainage purposes and when the slopes, b and c in Equation 10.2, are within tolerances for the irrigation system. In many surface irrigation design situations, the design plane will differ from the best-fit plane (e.g., the design plane generally has two fixed design specifications which leaves only one dimension in which the plane can be changed to balance cut and fill volumes). For example, in border irrigation there normally is no cross slope in the border strips and the field elevation at the inlet may be fixed. This means that the plane has only one factor, the longitudinal slope, which can be varied in the design. Level basins, where the basins are designed to be level in both directions (Erie and Dedrick, 1979; also see Chapter 14), are another example. In this case, elevation is the only factor that can be varied for land grading. To minimize earthwork volumes, iterative adjustment of the slope or elevation of the design plane within prescribed limits is done by many earthwork software packages. Numerous procedures for defining the finished field surface have been based on uniformly spaced, grid-based patterns. In this work, much of the land grading design has involved determining a linear plane that balances cut and fill volumes to minimize earthwork. Further, design procedures for nonlinear surfaces have been presented (Hamad and Ali, 1990; Ebne-Jalal, 2004). The following paragraphs provide an overview of some of the linear plane procedures developed for analyzing grid-based data. Note that the focus is on minimizing the earthwork requirement. Givan (1940) introduced the first systematic procedure to perform land-grading designs for rectangular fields. Givan used least squares procedures to define a plane surface that fits the natural ground surface with a minimum of earthwork. Chugg (1947) developed the first land grading design method for irregularly shaped fields. He applied the least squares method to calculate the best-fit slope, using transparent and coordinate papers to simplify and organize the computational process. The procedure requires the graphical determination of rectangular distances from the origin to the centroid and a separate determination of the centroid elevation. Raju (1960) developed a method to calculate the slopes of the graded plane, which he called the fixed volume center method. His method is based on the criterion that the total volume of earth and the center of volume will be the same before and after grading. This assumption is required to ensure a balance between cut and fill and to obtain the least amount of grading and movement of earth. The fixed volume center method has proven to be an accurate method of land grading, and although time consuming when done by hand, can readily be used with computers. Harris et al. (1966) introduced what they called the best-fit warped surface, in which they recognized the need for other types of design by suggesting that a field-

Design and Operation of Farm Irrigation Systems

331

surface shape that allows row and cross-row grades to vary within prescribed limits would be useful, especially when surface drainage was needed. Such a surface has variable slope in the irrigation direction, while the cross slope follows the natural ground profile with minor modifications. Design limits are maximum and minimum row grades and cross slopes, maximum change in row grade and cross slope at a station, and the cut-fill ratio. Shih and Kriz (1971) presented the symmetrical residuals method of land grading design for both rectangular and irregularly shaped fields. The method is based on residual properties, Newton’s divided-difference interpolation procedure, and statistical properties of the best design with an unbiased estimate and minimum variance. Sowell et al. (1973) applied linear programming techniques to the land-grading designs presented by Shih and Kriz (1971) and found that the linear programming technique gave a smaller total sum of depth of cut. Paul (1973) developed two methods to calculate the slopes of the best-fit plane, the double-centroid and the computer-minimized cost methods. Hamad (1981) proposed a method for quick estimation of the volumes of earthwork by using probability theory. Manela (1983) applied the least squares method to find the best-fit plane. Scalloppi and Willardson (1986) further developed a practical procedure to calculate the slopes of a graded plane using the least squares method. All of the above design methods involved trial and error procedures to determine the plane that balances cut and fill volumes, considering the shrinkage of the soil, to minimize earthwork. Easa (1989) presented formulas for direct land-grading design that eliminates the need for trial and error procedures to minimize earthwork. The method explicitly considers the required design specifications, which may include the two edge slopes of the plane, one edge slope and a control point, or two control points. Such an approach is important in many practical situations in the field where many times the finished plane must conform to certain boundary conditions. Procedures also have been developed for analyzing irregularly spaced data. The bestfit design plane generally can be generated by a modified least-squares technique in which each data point is weighted in the analysis in proportion to the area it represents. 10.5.2 Computation of Cut and Fill Adjustments In land shaping it is essential that the volume of material excavated be adequate to make the fills. If the cuts equal the fills without borrowing or wasting material, the earthwork is in balance. Experience has shown that the volume of cuts computed from the design must exceed the fills within a certain percent. The cut-fill volume ratio, R, is defined as C′ R= (10.3) F′ where C′ and F′ are the cut and fill volumes, respectively, after the parameters of the plane have been adjusted. Cut-fill ratios generally range from 1.2 to 1.6, but can be as low as 1.1 and as high as 2.0 (Michael, 1978). The cut-fill ratio depends on factors such as soil type and its moisture content, both of which generally vary with depth of cut; the earthmoving equipment weight, its tire size, and number of passes made over the soil; and the depth of the layer (commonly referenced as “lift”) of soil deposited with each pass of the equipment. All of these factors influence the extent of compaction. The correct estimation of the cut-fill ratio is an important factor in the earthwork design. Shih (1992) provided a generalized guide for selecting cut-fill ratios dependent upon soil texture,

332

Chapter 10 Land Forming for Irrigation

depth of cut and fill, organic matter content of soil, and roughness of the soil surface (Table 10.1). Generally, it is best to base this estimate upon the judgment of those having local experience with the soils in question. For large projects, it may be appropriate to use a portion of the area as a test of the cut-fill ratio to be able to adjust the earthwork design for the remainder. The best-fit linear plane (Equation 10.2) assumes that the cut and fill volumes are equal. This formula for the best-fit plane can be used directly if the shrinkage of the soil during leveling is negligible. However, under most conditions, to obtain the desired cut-fill ratio (Equation 10.3) the parameters, a, b, and c, of the design land grading plane (Equation 10.2) must be adjusted. Clearly, the adjustments of the parameters of the land-grading plane should have the effect of lowering the plane. Most adjustment procedures have focused on lowering the plane, i.e., reducing a in Equation 10.2 by some amount, Δa, where Δa = a – a′ and a′ is the elevation of the adjusted plane: Δa =

1+ R 2 RN t

(R ∑ F - ∑ C)

(10.4)

where Δa = increment the grade line is lowered, m R = cut-fill ratio Nt = number of both cut and fills coordinates ΣC = summation of cuts before lowering the plane, m ΣF = summation of fills before lowering the plane, m. A trial and error procedure generally is used to adjust a, b, and c to satisfy R in Equation 10.3. As noted earlier, Easa (1989) developed approximate adjustment formulas for all design parameters a, b, and c that can be applied directly to determine the position of the plane that satisfies the cut-fill ratio. Approximation formulas like those presented by Easa (1989) are essential when calculations are being done manually and can be a useful tool within computer programs. Since R is simply an estimate based on experience, the adequacy of the cut-fill ratio should be checked as the land shaping progresses. Sometimes it is necessary to use material from the field to construct farm roads or elevated ditches, or material excavated from a drain may need to be spread on the field. The adjustment in the plane to allow for this material can be estimated by the following equation: V Δa ′ = K (10.5) A where Δa′ = increment the grade line is lowered, m V = volume of compacted fill in a road or elevated ditch to be taken (“borrowed”) from the graded surface, or the volume of soil to be distributed (“wasted”) over the graded surface when in its compacted state, m3 A = area of the field, m2 K = dimensional constant equal to 1.0 for Δa′ in meters. Table 10.1. General guidelines for selecting cut-fill ratios for land grading design (Shih, 1992). Soil Depth of Cut Organic Roughness of Cut-Fill Ratio Texture and Fill Matter Field Surface Low, 1.10–1.25 coarse deep low smooth Medium, 1.25–1.40 medium moderate medium moderately smooth High, 1.40–1.70 fine shallow high rough

Design and Operation of Farm Irrigation Systems

333

10.5.3 Earthwork Calculations The designer of the land grading program must be concerned with minimizing the land grading costs while carefully considering the layout of the system, the irrigation efficiency attainable, labor requirements, cost of operation and maintenance, and farmability. Thus, the optimum or minimum cost of land forming will be the minimum cost associated with meeting the above criteria. The volume of excavation, or cut, is generally used as a basis for contracting and for estimating equipment requirements and job schedule. Various procedures have been developed to minimize cuts and fills. In most instances, minimizing cuts and fills does not minimize costs (flat, uniform fields being an exception). The goal is to minimize the amount of work that must be done which includes minimizing the volume of cut and haul distance. Type of equipment also can affect the costs of land forming. In some instances, a distance criterion such as average haul distance is used to supplement the purely volumetric method generally used. The following paragraphs discuss some of the commonly used methods for computing volume of cut or fill. They all represent ways to calculate the volumetric difference between two surface models, one being the existing or original surface and the second representing the proposed design surface. The summation method, the least accurate of the methods presented, is used to provide quick estimates of volumes to be excavated (USDA-SCS, 1961). The method assumes that a given cut or fill at a grid point represents an area midway to the next grid point. The grid point is considered as being in the center of the grid rather than at the corner. The product of the cut in meters and the area of the grid in square meters provide the required excavation volume in cubic meters. Where excavation involves the entire grid, the volume of earthwork could be obtained by multiplying the area of the grid by the average cut at the four corners of the grid. Since cut and fill both often occur within a grid, a procedure known as the fourpoint method is used. This method is based on the equations V c=

2 L2 ( ∑ C ) K ∑C + ∑ F

(10.6)

Vf=

2 L2 ( ∑ F ) K ∑C + ∑ F

(10.7)

where Vc = volume of cut, m3 Vf = volume of fill, m3 ΣC =sum of cuts on the four corners on a grid square, m ΣF = sum of fills on the four corners of a grid square, m L = length of the grid square, m K = dimensional constant equal to 4 when C, F, and L in m and V in m3. An accurate method of computing the volume of cut or fill in land forming makes use of the prismoidal formula, L V = ( A 1 + 4 Am + A2 ) (10.8) 6 where V = volume, m3 L = distance between end planes, m A1 and A2 = end plane areas, m2 Am = middle section plane area, m2.

334

Chapter 10 Land Forming for Irrigation

This formula is easily adapted to computer applications, but it is laborious by hand or calculator. Many currently used computer-aided land forming programs determine the soil volume through the use of a triangulated irregular network (TIN) of elevation differences between the original and design surfaces (digital terrain model, DTM) to generate tetrahedrons from which volume can be calculated. The tetrahedron is used since it is the simplest enclosed figure that can be made from the least number of planes. By representing the original surface by its survey locations, it preserves the exact elevations as surveyed. The procedure includes the following steps: 1. Calculate the difference in elevation between the existing ground DTM and the proposed design surface DTM for all key points. Key points might include all data points on the existing ground model, defining data points for the design surface (e.g., corners of the area to be graded), and various intermediate points that may be formed by the intersection of triangle sides between the two DTMs. The calculation is done by projecting each point onto the other surface, interpolating the corresponding elevation, and calculating the difference. 2. A TIN is generated to represent the DTM created by the elevation differences. The triangles for the elevation difference TIN are generated honoring all break lines (e.g., elevations along a ridge and/or along the bottom of a drainage way) and points common to both DTM surfaces. 3. The TIN is used to determine volumes by breaking the data for each triangle into the appropriate number of tetrahedrons. The volume of each tetrahedron is calculated and the volumes are added together to compute the cut and fill volumes between the original and the design DTMs. 4. A form of the prismoidal formula (Equation 10.8) is used to calculate the volume of the tetrahedrons. Various solution procedures are used within the earthwork software packages to apply the formula. Most procedures are proprietary.

10.6 TYPES OF EQUIPMENT There is a great variety of equipment that can be used for land forming. In some instances, machinery has been made more efficient, more reliable, more compact, and can move more soil than formerly. Each type of machine has its own capabilities and limitations. Over the past 25 to 30 years, laser-control technology has been adapted to

Figure 10.1. GPS-based leveling systems use signals from satellites to control a scraper or scrapers as they move through a field, shown graphically on the left. The system is anchored in the field by a base station reference (right). A display, control unit, and GPS antenna are mounted on the tractor.

Design and Operation of Farm Irrigation Systems

335

most tools used in land forming; and a class of scrapers that can both carry, and in some instances, precisely smooth the design surface has been developed. As noted in Section 10.4, highly accurate global positioning systems have been introduced in the industry of land forming, and more recently have been adapted to its tools (Figure 10.1). These technologies have significantly impacted the choice of equipment and how it is used. The examples of equipment described in this section were selected to illustrate some of the types that are being used (Gattis et al., 1959; Haynes, 1966; Fisher et al., 1973; Shih, 1992; Wiedemann et al., 1977; Dedrick et al., 1982). All equipment for land forming must be kept in good repair. The best type and size depends on the particular job. 10.6.1 Power Units Prior to the 1960s, crawler tractors were used to pull most scrapers. Crawler tractors are still used in special circumstances since they have low soil compaction characteristics, provide excellent traction on a variety of soils, and can move large volumes of earth when used with 6- to 15-m3 carrier-type scrapers. They are well suited for short hauls but their slow speed (about 8 km/h) limits their suitability for broad use. Most of the land forming is now being done with either self-propelled rubber tired scrapers or scrapers pulled with rubber-tired wheel tractors. Their high field speeds and maneuverability makes them well suited for this type of work. Two-wheel units are more maneuverable than four-wheel units. Either can be used with scrapers to great advantage on land-grading jobs requiring long-distance hauling. They need to be appropriately matched to the earth-moving equipment as certain combinations are better than others. Bulldozers are suitable for making heavy cuts and moving earth a short distance. Earth moving distances by bulldozer should not exceed 90 m, but less than 60 m is preferred. Bulldozers are well-suited for certain rough grading work and are essential as pusher tractors when extra power is required for loading large scrapers. 10.6.2 Equipment for Land Grading Most land grading where large quantities of earth are to be moved over appreciable distances is done with earth-moving scrapers ranging in capacity from about 1.5 m3 to 20 m3. The advent of the fixed-blade scraper has made it much easier to control depths of cut and fill allowing much larger-volume scrapers to be pulled by smaller power units (e.g., farm tractors). Motor graders and blade-type graders pulled by tractors may be used for land forming on small fields, especially for narrow benches or where only minor grading work is needed. They are used more extensively for shaping ridges on the downhill side of benches and the slope from one bench to another; excavating relatively small channels, such as shallow drainage ditches, where the excavated material can be spread or shaped along one or both sides of the channels; constructing divisions between irrigated borders or basins and crowning land; and maintaining shallow drainage ditches or field laterals. Many types of scrapers can be, and generally are, equipped with laser and/or GPS control systems. The laser technology was originally applied in agriculture for automatic grade control for subsurface drainage equipment (Fouss et al., 1964) and was later adapted to equipment used in land forming. Development of highly precise GPS technology over the past 10 years has facilitated its use with such equipment. Laser and GPS precision-guidance systems have revolutionized the process of land forming.

336

Chapter 10 Land Forming for Irrigation

The carrier-type scraper is a tractor-drawn implement that excavates, loads, hauls, and spreads earth. It consists of a bucket mounted on rubber-tired wheels with a blade and apron across its front end for cutting, scooping and retaining a load of earth. Carrier-type scrapers can be used efficiently on many land-grading jobs and are most efficient where the depth of cut is sufficient for rapid loading. The scraper must be pulled by the right size and type of tractor. Often, on large earth-moving jobs, these scrapers are operated in groups with a pusher tractor. Carrier-type scraper sizes range from 2 to 20 m3 capacity with the smaller capacity being designed for use with farm tractors. Elevating scrapers are used for land grading, especially where intermediate depths of cut are required. Farm tractors commonly used with four-row equipment can handle the 4-m3 size. Large farm tractors or contractor-type tractors are used to pull the larger scrapers. A very desirable feature is the ability of the elevating scraper to load under most depths of cut and soil textures. In addition, cutting and spreading can be done evenly. The ejector scraper, a relatively recent addition to the land grading and smoothing industry, is a fixed-blade scraper that can be used in place of the elevating scraper in some instances and is exceptionally useful where relatively small cuts are required (Figure 10.2). Ejector scrapers range in capacity from a few cubic meters to more than 10 m3 and can be used singly or in tandem. The back wall of the bucket is moveable. Soil is forced or ejected out of the scraper when hydraulic cylinders push the back wall forward. The soil flows out of the scraper at an even rate, simplifying the grading operation. Further, the cutting edge can be precisely positioned with laser-control equipment resulting in precisely finished grades. The moveable back wall of the ejector scraper can be fixed in a forward position to create a land-smoothing device similar to the bottomless drag scraper discussed below. Thus, the scraper is a tool that can be used to cut, haul, and fill over relatively large distances while also being used to provide the final smoothing operation. 10.6.3 Equipment for Land Smoothing Normally, it is impractical and too expensive to finish land surfaces to exact grade with heavy earth-moving machines. The heavy scraper work should provide a field surface such that two or three passes over the field with land smoothing equipment, designed to remove small grade irregularities, will produce the desired uniform surface. 10.6.3.1 Bottomless scrapers. The bucket or bowl of a bottomless scraper has little or no bottom and earth moving is accomplished by scraping soil from areas that are above grade, dragging it a short distance, and depositing it in areas that are below grade. Bottomless scrapers work best in a loose soil that has adequate moisture so that a smooth, relatively firm surface remains after the smoothing is completed. Dry, powdery soil conditions should be avoided. The field should be watered when such conditions exist, then when the field has sufficiently dried, the surface is worked with a disk or shallow chisel before the final smoothing is done. There are two basic types of bottomless scrapers, the land plane and the drag scraper, which differ in the way in which grade is controlled. The land plane is a bottomless scraper with a long frame extending both in front of and in back of the scraper, with wheels mounted at the ends of the frame. The blade on the land plane is mounted midway on the frame and is adjustable vertically so that the depth of cut and the amount of earth carried by the bucket can be regulated. Some

Design and Operation of Farm Irrigation Systems

337

Figure 10.2. Ejector scrapers are a type of fixed-blade scraper used for moderate land grading and smoothing, operated singly (above) and in tandem (below).

planes are equipped with hydraulic controls so that the tractor operator can control the blade level from the tractor. Other planes are manually adjusted. When in use, the blade is set at a level that will maintain about one-third to one-half load in the bucket. If the blade is set too low, the soil in the bucket will become excessive, spilling around

338

Chapter 10 Land Forming for Irrigation

the ends and over the top of the bucket. If it is set too high, there will be insufficient soil to adequately fill the low areas. When the adjustment is properly made, the plane will automatically remove the high spots and fill in the depressions of a diameter equal to nearly half the machine length. Cuts of 6 to 9 cm or less can be made within the length of the machine. Land planes are best suited for relatively large fields. They are manufactured in lengths up to about 27 m and with blade widths up to 4.5 m. From a smoothing standpoint, the longer the plane the better the smoothing job. From a maneuverability standpoint, the longer machines require a wider turning area, and thus will leave a wider strip around the boundaries of the field. Smaller fields will have greater proportions that cannot be properly smoothed. On fields of 8 ha or less and on narrow benches, smaller and more maneuverable planes are more desirable. The two-wheel bottomless scraper, commonly known as a drag scraper, can be operated either manually by the tractor operator or by using laser-guided controls. Lasercontrolled drag scrapers (Figure 10.3) and ejector scrapers (Section 10.6.2, Figure 10.2) have become the precision, land smoothing tools of choice. These machines are manufactured in a number of widths, up to 5.5 m, to serve various purposes. The drag scraper is well-suited for handling large volumes of earth over a short haul, while the ejector works for both short and long hauls. Manual smoothing can be done with the scrapers through visual control by the operator observing the grade to control the quality of finishing. Through the use of the hydraulic controls, the tractor operator can cut, drag, and unload as desired. When the drag scraper is used with operator control, it may be necessary to land plane the area to attain the required results. The lasercontrolled drag scraper will do a more accurate job than a skilled operator using a manually operated drag scraper and/or land plane. Further, it relieves the operators of some of their machine control responsibilities, and if combined with available field

Figure 10.3. Laser-controlled drag scraper used for precision surface finishing. The battery-powered laser transmitter, with solar charger, is shown in the foreground and the receiving unit is shown mounted on the scraper.

Design and Operation of Farm Irrigation Systems

339

surveying equipment can reduce or eliminate the need for traditional survey crews. Often manually operated drag scrapers are used before land planing to remove surface irregularities too minor to be taken care of economically with a carrier-type scraper. 10.6.3.2 Other levelers and floats. Farmers in their farm shops as well as various commercial companies have made excellent levelers and floats in many variations. Generally, the earth-moving principles are the same as for land planes, and the effectiveness is dependent on the length of the frame. Levelers and floats that can be pulled by the average farm tractor are more important in restoring the smoothness of a field after tillage than in removing the irregularities left by heavy earth-moving equipment.

10.7 FIELD OPERATING PROCEDURES The land grading work must be planned and scheduled to assure satisfactory field conditions. Some of the factors that must be considered are timing, weather, and field condition. ƒ Timing—The field operations must fit into the cropping sequence. When appropriate, the grading should be scheduled so that the work can be carried out between the harvest of one crop and the planting of the next crop. The time available between crops will help determine the type of equipment to be used. ƒ Weather—While weather conditions cannot always be predicted, there is generally a time in each region when conditions are most favorable for grading work. If possible, the work should be planned for the most favorable season to avoid long shut-down periods. In particular, fieldwork should not be scheduled for periods of high rainfall, freezing temperatures, or other weather conditions that may cause excessive delays in earth moving or cause wind or water erosion. ƒ Field conditions—Some field conditions during land grading can have serious adverse effects on the quality of the work and crop production after it is done. Grading should not be done when the fields are excessively wet. Operating scrapers in wet soils is difficult and can cause compaction and other serious damage to the soil structure, which results in poor soil conditions for crop production. Grading of very dry soils should also be avoided. Cuts are difficult to make, fills become loose and powdery, and the result is poor quality work, higher cost of operating the equipment, and a dust hazard. Grading with excessive crop residues also should be avoided since it can result in poor compaction and excessive settling in deeper fill areas. The quality of the land forming operation depends on the equipment used, condition of the field, and the skill of the operator. There are probably as many different ways of approaching the actual earth moving process as there are equipment operators. The earth-moving work should be carried out so that the grid stakes are not disturbed until the job is ready for the final smoothing operation. The farmer or contractor is normally given a map showing the work to be done. The information shown on the map depends on local custom and contractor preferences. In some cases it may be a map showing just the amount of cut or fill. A typical cut-and-fill map is shown in Figure 10.4. Numbers indicate the cut or fill in tenths of a meter. For example, C02 indicates that the cut is to be 0.2 m. If the grid point were marked F01, the fill would be 0.1 m. Some engineers make a detailed study of the contractor’s map and prepare earthwork balance areas suggesting the direction and location that earth is to be moved. Most contractors or operators study the map and make their own decision on

340

Chapter 10 Land Forming for Irrigation

Figure 10.4. A typical cut-fill map used by the equipment operator. Cuts and fills are in tenths of a meter.

Figure 10.5. Field sketch of suggested haul lanes to be used by the land forming equipment operator to optimize the earthmoving process.

Design and Operation of Farm Irrigation Systems

341

matching cut and fill areas. Figure 10.5 is an example of how the earth movement patterns might be visualized by the operator. On large, more complicated, jobs the haul patterns may be sketched out on the contractor’s map. In some instances, the contractor may want a complete map showing original ground elevation, planned finished grade elevations, and the cut and fill figures. Other procedures include such things as indicating the cut with red figures and the fill with blue, or by cross-hatching the cut areas in red and the fill areas in blue. 10.7.1 Land Grading Cutting and filling is normally done parallel to the grid stake rows. The procedure varies with operator preference and equipment. In some instances, when the scrapers are equipped with laser control equipment, field stakes are not used. The operator merely uses a cut-fill map of the area, some idea as to the most efficient method of going about the grading job, and uses the laser controls to provide the grade guidance needed. If the grade stakes are used, some operators cut and fill a strip the width of the scraper adjacent to the stake rows. The same procedure is followed for the opposite side of the lane along the next row of stakes. When these are brought approximately to grade, work is carried on in the intermediate area. When done without laser control, the grade between stake rows is carried visually across the lane. This process is continued until the entire field has been shaped. Another procedure is to make a cut along the stake row as deep as can be made on the first pass. The second strip is made just far enough over so that the tractor can straddle the strip that is left. The strip that was left is then removed as a third strip. This procedure is repeated across the cut area. Regardless of the procedure followed, the cuts and fills should be made in uniform layers. This reduces uneven settling of the field, facilitates carrying the proper grade from one stake row to the other, permits faster travel, and is easier on equipment. Normally the small area around each stake is left until the shaping is completed. If deep cuts are made leaving a large, high mound at the stake, it may be desirable to remove the stake, cut away the mound and replace the stake. Cut areas should be disked as soon as possible after the cuts are made to keep the cut surface from drying and becoming hard. This will make the preparations for the final smoothing easier. Normally grades are checked with a permissible tolerance of ±60 mm. On level or flat grades it may be necessary to check within ±30 mm. In no case should reverse grades in the direction of irrigation be permitted. When the field construction has been approved, the stakes should be removed, the mounds and depressions at the grid stakes graded to the level of the adjacent surface, and the field disked or chiseled to prepare it for the final smoothing operation. 10.7.2 Land Smoothing 10.7.2.1 Land plane. Since the cut-and-fill operation by the heavy earth-moving equipment is normally carried out parallel to the rows of the grid stakes, it is generally desirable to carry out the first planing operations diagonally to the grid pattern. It is customary to land plane the field in three directions. The first two operations are in diagonal directions, perpendicular to each other, and the third is in the direction of irrigation. During the smoothing operations, considerable time is lost in making the turns at the field boundaries. Figure 10.6 shows how the first two diagonal planing operations can be carried out to cover the field with two passes of the plane in the minimum amount of time. This can reduce the time required by about 25% over planing the entire field in one diagonal direction and then repeating the operation at right angles to the first operation.

342

Chapter 10 Land Forming for Irrigation

Figure 10.6. A method of performing two diagonal land planings in which distance traveled is minimized.

10.7.2.2 Drag and ejector scrapers. In recent years, the drag scraper or the ejector scraper with the moveable back of the scraper in the forward position, has become an excellent final smoothing machine, with laser control becoming state of the art after first being introduced in the mid-1970s as a tool to assist in land forming. Figure 10.3 shows a scraper and laser equipment in the field. The laser transmitter is mounted on a tripod, special stand, or trailer. It generates a thin laser beam that is rotated rapidly to form a virtual plane of light over the working area. The plane either can be level (for level basins) or sloping. The maximum range of the laser is about 300 m, but dust, wind, and heat can be limiting. However, these do not affect GPS-controlled systems, which have been adapted to land forming tools like the drag and ejector scrapers. There are many personal preferences as to how the smoothing operation will be completed. If there are significant high and low areas, a typical approach is to use a field map as a guide to move soil from high to low areas. Once the field is close to on grade, or if there are no significant high and low areas to start with, the operator normally runs the scraper over the field in two or three directions with the last pass in the direction of irrigation. Fields smoothed in this way normally do not require further smoothing (e.g., land planing not required). As noted earlier, land smoothing can be described statistically by the variability of the field elevations around the design slope. Field elevations are normally distributed; thus, the standard deviation of the field elevations from the design slope is an appropriate criterion for assessing smoothing adequacy. With well-trained operators and carefully adjusted equipment, standard deviations of the finished field elevations from the targeted field plane finished with laser-controlled scrapers range from about 12 to 15 mm on 4 to 16 ha fields (Dedrick, 1979; Dedrick et al., 1982). On level basins in the southwest, NRCS standards require 80% of a basin to be within +15 mm and 100% within ±30 mm. To meet the 80% criteria the standard deviation must be slightly less than 12 mm. Typically, standard deviations on “level” basins prior to laser-controlled smoothing ranged from 25 to 30 mm. 10.7.2.3 Water leveling. Water leveling for rice land is a method of field benching that was used extensively in the past, but in the United States it essentially has been

Design and Operation of Farm Irrigation Systems

343

replaced with laser-controlled scrapers. Rice is normally grown on soil having a restrictive layer with a permeability rate of about 0.5 mm or less per hour. If this restrictive layer is thick enough to support a tractor and scraper under a water-saturated condition, water leveling can be used. The equipment commonly used for water leveling is a bottomless scraper or a blade mounted on the back of a farm tractor. The general procedure is as follows: ƒ Smooth the field with a land plane to remove as many of the surface irregularities as possible. ƒ Locate and construct the permanent levees. Levees should be straightened by filling across the low areas. ƒ Flood as much area as can be leveled in one day to a depth that will cover the high point. ƒ Carry out the leveling work using the tractor-mounted blade or the bottomless scraper. Tractor movement produces water waves that assist the leveling operation. More than one tractor is often used to speed up the work and to create more wave action. ƒ Soil movement can be determined by water depth measurements. ƒ After the suspended soil material has settled and the water has become clear, the remaining water should be drained off. ƒ When the soil has dried sufficiently, the area between levees should be worked and smoothed with a land plane. In some rice production areas, land forming is done to reduce the number of levees and to construct parallel levees. Final smoothing of the areas could be done with either water-leveling techniques or laser-controlled scrapers. Faulkner (1964) provides more detailed information on water leveling. 10.7.3 Maintenance Land forming often requires a large investment of capital. Annual maintenance is an important part of the farmer’s operation to obtain the proper return on this investment. Special treatment and annual maintenance is needed to bring land back to full production and keep the surface uniform. To improve soil condition after smoothing, tillage (e.g., plowing, disking, or chiseling) mixes the disturbed soil and loosens areas that have been compacted. In cold climates, leaving the soil surface in a roughened condition over winter will allow the frost action to break down clods and improve infiltration. If land forming has exposed infertile subsoils, soil testing is recommended. Reclamation practices might include a green manure crop the first year to add organic matter and/or applications of barnyard manure. These reclamation procedures may need to be supplemented with the application of fertilizer to meet crop requirements. Trace minerals, such as iron or zinc, are sometimes deficient in cut areas. Following land shaping, irrigation, and farming operations, fill areas tend to settle and cut areas “fluff up” leaving the surface uneven. In some instances, final smoothing after major land forming may not be justified until the land has been irrigated. To provide an opportunity to refinish the area, farmers generally plant an annual crop the first year after land forming. The extent of the continuing maintenance procedures will depend on several factors such as the type of tillage equipment that is being used, time since major land forming processes were completed, type of crop to be grown, value of the crop to be grown, and the susceptibility of the crop to damage caused by soil surface irregularities and associated nonuniform water application.

344

Chapter 10 Land Forming for Irrigation

Cultivation should be done carefully to help maintain precisely smoothed surfaces. Plowing should be done with a two-way plow. In some cases, sweeps or chisels are used to cut through the soil, raising and dropping the soil to provide a loosening effect.

10.8 COST AND CONTRACTING The cost of earth work in land forming will vary with a number of factors: topography, field conditions, length of haul, total volume and volume of earth movement per unit area, type of equipment used, and the operator’s skill. Competition among contractors and the need for off-season work often affects the cost. In most areas the major earthwork is contracted based upon a volumetric basis. Final smoothing generally is contracted on a per unit area basis. In some instances, where the contractor does both the major land moving and finishing, the quoted price per unit volume includes final smoothing. Most contracting is done on a verbal agreement between the contractor and the farmer or landowner. This sometimes leads to misunderstandings. The following items are suggested as some of the things that should be covered by a written agreement: ƒ field location and description of site conditions; ƒ basis for the construction such as the contractor’s work map, special construction not shown on the map, allowance for settlement in heavy fill areas, cut-tofill ratio, etc.; ƒ facilities, materials, and assistance which the farmer will provide to the contractor; ƒ who will do the smoothing and when it is to be done, if the final smoothing is to be done as a separate job; ƒ completion date with provisions for extension if factors beyond the contractor’s control delay the progress of the work; ƒ basis for payment (if based on volume, the method to be used for determining the volume and who is going to do the computation should be stated; agreement should be made on how the field will be checked for completion); ƒ liability insurance; ƒ responsibility for locating and protecting buried utilities; ƒ final clean-up of rubbish, excess materials and equipment, broken equipment, etc; and ƒ pay schedule.

10.9 SAFETY All equipment must have the required safety devices and should be operated in a reasonable and safe manner. Safety shields must be in place and securely fastened. Special precautions should be taken when operating equipment around electric power lines, buried utilities, and communication lines. Buried utilities and communication lines have been damaged and lives lost during land grading operations. The general location of the buried utility should be indicated on the contractor’s map. Most utility companies will, on request, stake the exact location of their utility line and indicate the depth of the cover over the line. While some governmental agencies serve written notice to the farmer that he is responsible for notifying the utility company of the work to be done on his farm, all parties involved in the land forming must be safety conscious. Dust associated with land forming can be a serious problem, especially during and after the final smoothing operation. Dust affects nearby residents, creates hazardous conditions on nearby roads, and can damage crops on adjacent farmland. As noted

Design and Operation of Farm Irrigation Systems

345

earlier, land forming on dry soils should be avoided, both from a quality of job standpoint (i.e., loose soil is difficult to smooth) and from a dust (health and safety) standpoint. In some instances, temporary irrigation facilities, such as portable sprinklers, may need to be used to provide better working conditions. If soil moisture is adequate, working the field with small chisels or a spring-tooth harrow may provide sufficient surface roughening to control blowing dust. The finished field should be irrigated and planted as soon as possible. If the planting of the next crop is to be delayed for any length of time, it may be advisable to plant a cover crop to provide adequate protection from wind erosion.

REFERENCES Anderson, C. L., A. D. Halderman, H. A. Paul, and E. Rapp. 1980. Land shaping requirements. In Design and Operation of Farm Irrigation Systems, 281-314. M. E. Jensen, ed. St. Joseph, Mich.: ASAE. Buras, N., and G. Manor. 1970. Field comparison of land smoothers. Trans. ASAE 13: 639-640, 643. Chugg, G. E. 1947. Calculations for land gradation. Agric. Eng. 28(11): 461-463. Dedrick, A. R. 1979. Land leveling-precision attained with laser-controlled drag scrapers. ASAE Paper No. 79-2565. St. Joseph, Mich.: ASAE. Dedrick, A. R., L. J. Erie, and A. J. Clemmens. 1982. Level basin irrigation. In Advances in Irrigation, 105-145. D. I. Hillel, ed. Academic Press. Dedrick, A. R. 1989. Surface-drained level basin systems. In Proc. Int. Conf. on Irrigation: Theory & Practice, 566-575. Southampton, U.K. Easa, S. M. 1989. Direct land grading design of irrigation plane surfaces. J. Irrig. Drain. Eng. 115(2): 285-301. Ebne-Jalal, R. 2004. Weighted average method for land grading design. J. Irrig. Drain. Eng. 130(3): 239-247. Erie, L. J., and A. R. Dedrick. 1979. Level-basin irrigation: A method of conserving water and labor. USDA Farmers’ Bulletin No. 2261.Washington, D.C.: USDA. Faulkner, M. D. 1964. Leveling Rice Land in Water. Crowley, La.: Louisiana Sta. Univ., Rice Exp. Sta. Fisher, C. E., H. T. Wiedemann, C. H. Meadors, and J. H. Brock. 1973. Chapt. 7: Mechanical control of mesquite. In Mesquite-Growth and Development, Management Economics, Control, Uses. Res. Mono. 1, Texas Agricultural Energetics. Westport, Conn.: AVI. Fouss, J. L., R. G. Holmes, and G. O. Schwab. 1964. Automatic grade control for subsurface drainage equipment. Trans. ASAE 7(2): 111-113. Gattis, J. L., K. A. Koch, and J. L. McVey. 1959. Land grading for surface irrigation. Agric. Ext. Serv. Circ. 491. Little Rock, Ark.: Univ. Arkansas. Givan, C. V. 1940. Land grading calculations. Agric. Eng. 21(1): 11-12. Hamad, S. N. 1981. Using probability theory in land grading earthwork. Al-Muhandis 80: 23-28. Hamad, S. N., and A. M. Ali. 1990. Land-grading design by using nonlinear programming. J. Irrig. Drain. Eng. 116(2): 219-226. Harris, W. S. 1971. Crop yields as affected by row grades. Trans. ASAE 14: 860-863. Harris, W. S., J. C. Wait, and R. H. Benedict. 1966. Warped-surface method of land grading. Trans. ASAE 9: 65-65.

346

Chapter 10 Land Forming for Irrigation

Haynes, H. D. 1966. Machinery and methods for constructing and maintaining surface drainage on farm lands in humid areas. Trans. ASAE 9: 185-188, 193. Kriz, G. J., L. D. Hunning, R. E. Sneed, and S. F. Shih. 1973. Land-forming designs for water management. J. Soil Water Cons. 28: 134-136. Lyles, L. 1967. Effect of dryland leveling on soil moisture storage and grain sorghum yield. Trans. ASAE 10: 523-525. Manela, M. 1983. Construcao de planos geometricos em sistematizacao de solos pelo metodo dos minimos quadralos generalizados. Irrigacao e Tecnologia Moderna 10: 32-34, 36-37. Michael, A. 1978. Irrigation: Theory and Practice. Vikas, New Delhi, India: Vikas Pvt Ltd. Paul, H. A. 1973. Theoretical derivation and economic evaluation of the double centroid and computer-minimized cost methods of calculating slopes for land grading. Thesis. Logan, Utah: Utah State Univ. Portier, K. M., and S. F. Shih. 1982. Spline fitting in land-forming design. Soil CropSci. Soc. Fla. Proc. 41: 39-43. Quackenbush, T. H. 1967. Land modification for efficient use of water. J. Irrig. Drain. Eng. 93: 7014. Raju, V. S. 1960. Land grading for irrigation. Trans. ASAE 3(1): 38-41, 54. Scalloppi, E. J., and L. S. Willardson. 1986. Practical land grading based upon least squares. J. Irrig. Drain. Eng.112(2): 98-109. Sewell, J. I. 1970. Land grading for improved surface drainage. Trans. ASAE 13: 817819. Shih, S. F. 1992 Energy requirements for land clearing, forming, and ditching. Energy in World Agric. 6: 101-116. Shih, S. F., and G. J. Kriz. 1971. Symmetrical residuals method for land forming design. Trans. ASAE 14(6): 1195-1200. Shih, S. F., D. E. Vandergrift, D. L. Myhre, G. S. Rahi, and D.S. Harrison. 1981. The effect of land forming on subsidence in the Florida Everglades’ organic soil. Soil Sci. Soc. America J. 45: 1206-1209. Sowell, R. S., S. F. Shih, and G. J. Kriz. 1973. Land-forming design by linear programming. Trans. ASAE 16: 296-301. Springer, G. 1966. Land reshaping for cherry orchards. Trans. ASAE 9: 292-294. Thomas, D. J., and D. K. Cassel. 1979. Land-forming Atlantic Coastal Plain soils: Crop yield relationships to soil physical and chemical properties. J. Soil Water Cons. 34: 20-24. USDA-SCS. 1961. Chapt. 12: Land leveling. In National Engineering Handbook, Sect. 15. Washington, D.C.: USDA Soil Conservation Service. Wiedemann, H. T., B. T. Cross, and C. E. Fisher. 1977. Low-energy grubber for controlling brush. Trans. ASAE 20: 210-213.

CHAPTER 11

DELIVERY AND DISTRIBUTION SYSTEMS John A. Replogle (ARS-USDA, Maricopa, Arizona) E. Gordon Kruse (ARS-USDA, Fort Collins, Colorado) Abstract. The supply of irrigation water may be on-farm or delivered to the farm through a water-user association or irrigation district from more distant, often more reliable, and larger water supplies. We summarize farm water delivery policies and their impacts on farm labor, crops, and on-farm investment needed to achieve effective water management. Water delivery usually implies that several farms share access to the water. The irrigation schedule, including flow rate, duration of access, and the time to access the flow again, can be rigid or flexible. Hardware items, including automatic gates, methods of flow metering, and systems used to convey irrigation water from the farm supply to the fields are described. Canal and pipeline water measurement is presented. Pipelines or lined ditches offer greater seepage control, ease of water diversion, and reduced maintenance, but have a higher initial cost than earthen ditches. There is a tendency for mechanizing and automating supply, distribution, and application systems in order to deliver water with a flexible schedule. Requirements of system management and equipment, including reservoir use and management, structures and water conveyance design, system operation and maintenance, and mosquito control, are presented. Keywords. Canals, Ditches, Flow control, Gates, Irrigation district, Irrigation management, Irrigation scheduling, Low-pressure pipe systems, Mosquito control, Pipelines, Reservior, Water management.

11.1 INTRODUCTION When the supply of irrigation water exists on the farm, either from a well (as are an estimated 42% of the water supplies for irrigated farms) or directly diverted from a stream (as are an estimated 16%), the farm operator can usually control the timing and quantity of the delivery. Most other farm water supplies are surface waters delivered to the farms through a system operated by a water-user association or irrigation district, whose purpose is to facilitate water delivery, application, and removal. Such organizations give the farmer access to more distant, often more reliable, and larger water supplies by sharing the construction and operating costs with many other users, but there are usually more restrictions and external control of amounts delivered and flow rates.

348

Chapter 11 Delivery and Distribution Systems

The first portion of this chapter summarizes the various forms of farm water delivery policies and their impacts on farm labor, crops, and on-farm investment needed to achieve effective water management. Delivery policies and methods are classified in terms of distinguishing physical parameters. Physical features of the systems and hardware items, like automatic gates and methods of flow metering, that can significantly contribute to their improved operation are described. The delivery system is purposely viewed as a subsystem of the farm production scheme, as part of the total water handling cost of the farm unit, because the farmer ultimately pays essentially all costs. Improvements and new developments are stressed. The remainder of the chapter describes systems used to convey irrigation water from the farm supply—a canal, reservoir, or well—to the fields. Such systems generally have a capacity less than about 0.3 m3/s. Ditches or pipelines used for such conveyance should be well designed and maintained to minimize water loss by seepage and to allow efficient irrigation. Generally, pipelines or lined ditches offer greater seepage control, ease of water diversion, and reduced maintenance, but have a higher initial cost than earthen ditches.

11.2 IRRIGATION WATER DELIVERY Water delivery through canals or pipelines usually implies that several farms must somehow share access to the water in terms of flow rate, duration of access, and the return time to access the flow again. The canal or pipeline system must accommodate these three factors. The implied goal is to provide water at the right time in the right amount and at the right place. Canals and pipelines within the farm must then be operated to accommodate the delivery restrictions. 11.2.1 Schedules The delivery policy used by the water supply organization to meet this three-part implied goal is called a scheduling system or schedule. This scheduling system may have been established by law or based on economic and engineering factors that are important, or were important when the organization was established. Schedules consist of the three factors: delivery flow rate to the farm, delivery frequency to the farm, and delivery duration. The relations between these three factors control the capital cost and operating expenses of the delivery system. The same factors bear heavily on the effective and economical use of water, labor, energy, and capital investment on the farm. They may even limit the crops that can be considered. An appreciation of the operational range available for different types of schedule can be obtained by examining the three scheduling components in more detail. ƒ Flow rate—For the delivery flow rate, policy may specify a constant flow at each delivery cycle throughout the season. One limit is a continuous, unchanging, uninterrupted flow. Variations would include continuous but seasonally modified flow rates. The more common cycled flow is usually in some kind of rotation system among users. The canal delivery equipment associated with constant deliveries is frequently a fixed orifice or fixed pipe size. Seasonal changes in flow rate to users can be implemented within narrow limits by adjusting the canal water surface elevation in some instances. ƒ Frequency—Delivery frequency, like flow rate, can vary widely. Water can be delivered periodically throughout the season, and this period can be changed seasonally. For example, the cycle may repeat at two-week intervals in the hot season and be changed to three- or four-week intervals in the cooler seasons.

Design and Operation of Farm Irrigation Systems

349

ƒ Duration—Delivery duration, the length of time that flow is allowed through the farm gate or valve, can vary from continuous throughout the season to a few minutes or hours per irrigation cycle. In some projects, all farms on a canal take small flows for the entire period the canal is operating. The canal may be filled on alternate weeks. Other projects deliver in 24-hour units or in some other standard but fixed time block. An additional consideration is who controls the rate, frequency, and duration. This can be local farm-operator control, central-authority control or some combination. 11.2.1.1 Schedule types. Schedule types may be broadly classified as either rigid (predetermined) schedules or flexible (modifiable) schedules. Variations are summarized in Table 11.1, with decreasing flexibility from top to bottom. Descriptions of the three components, how they are handled, and who controls each component are useful for understanding the advantages and limitations. In general, the farm water delivery policy must be compatible with the distribution system on the farm. In many cases this has required on-farm designs to be adapted to an existing delivery system with the net result that the total production system is severely limited. Often, economical improvements in the design and operation of on-farm systems are made possible by appropriate modifications in the water delivery system. Thus, improvements in the distribution system, although they may increase costs, may be justified by more effective water use and reduced on-farm labor, capital, or energy costs. A system of schedule naming was proposed by Replogle and Merriam (1980). The names for various schedules differed slightly from names commonly used, but not always with the same meanings among users. The proposed names were subsequently revised by Replogle (1984, 1987). Clemmens (1987) grouped the schedules according to the degree of flexibility provided to the farmers in terms of rate, frequency, and duration, and the level within the delivery system at which irrigation delivery decisions are made. He otherwise retained most of the naming system described above. Table 11.1 summarizes some of the more prominent schedule combinations and the source of control. The constraints on each of the schedules are indicated to help provide an understanding of the possible implications of comparative schedules. 11.2.1.2 Rigid-schedule irrigation systems. Rotation (rigid) schedules can sometimes improve irrigation efficiencies in areas where farm operators are new to irrigation technology, compared to the free access of a flexible system, which could be misused. Rotation schedules require relatively low capital investment in canals or distribution pipelines and involve the least water agency management. The rotation schedules have one of the lowest staffing requirements in relation to the capital investments in the system. Field tasks may be simply to open or close a valve or gate as instructed, with no precision regulation required. Without on-farm water storage, this schedule forces the farm operators to operate at low water use efficiency. The irrigator may have to forego crops that must be watered often, require precise timing of applications (some vegetable and hay crops), or require careful water management to discourage pests and diseases. The only communication required between farm operator and water agency is a posting (or other general unilateral notice) of the schedule, a desirable simplicity in areas with numerous small farms and limited telephone service. However, demand systems (see Section 11.2.1.3 below) also have limited communication requirements because they allow the farm operator to access the water supply without notice to anyone.

350

Chapter 11 Delivery and Distribution Systems

Table 11.1. Definitions of delivery scheduling methods, after Clemmons (1987). System Constraints Schedule Categories Rate Frequency Duration Local control Demand schedules Demand U U U Limited rate demand L U U Arranged frequency demand L A U Intermediate Control Arranged schedules Arranged A A A Limited rate arranged L A A Restricted arranged C A C Fixed duration arranged C A F Fixed rate/restricted arranged F A C Central control Central system schedules Central system V V V Fixed amount V F F Rotation schedules Rotation F F F Varied amount rotation F F(V) F(V) Varied frequency rotation F F(V) F(V) Continuous flow F(V) Terminology: U: Unlimited, no restriction, under user control. L: Limited to maximum flow rate, but still arranged. A: Arranged between user and water authority. C: Constant during irrigation as arranged. F: Fixed by central policy. V: Varied by central authority, at authority’s discretion. (V): Varied by central authority, seasonally, by policy.

The use of rigid rotation-scheduling policies should at best be considered as an interim delivery policy and efforts should be made to direct rehabilitation, redesign, and education towards providing better-informed users and more flexible delivery schedules that can respond to soil- and crop-water needs and conserve water and energy. Rigid-schedule systems can sometimes be made more effective if some flexibility in water delivery can be introduced. Farm operators, for instance, might make their needs known through a preseason meeting with the water delivery authorities. Thus, these listed schedules could be further categorized as central-controlled schedules or schedules for which the farm operator has some input, but not enough day-to-day operational input to be classified among the arranged-schedule groups. Typical flow rates of rigid-schedule systems are from 15 to 30 L/s; typical frequencies are 7 to 14 days; and typical durations are 8, 12, or 24 hours. The larger flows are usually in canals and the smaller flows are frequently in low-pressure pipelines. Limiting cases, as mentioned above, are the continuous flow schedule, and durations that last as long as flow is in the canal. 11.2.1.3 Flexible-schedule irrigation systems. Demand (flexible) schedules, in which the irrigator may have water as desired—flexible in frequency, rate, and duration—are often too expensive, especially if large rates or complete time flexibility are

Design and Operation of Farm Irrigation Systems

351

offered. The capacity of such a system would be too large for reasonable capitalization and operating costs, because it must be large enough to meet the combined probable demand of all the users at any one time. Thus, the actual schedule is a limited-rate demand schedule, which has often been called a modified-demand schedule. It allows flexible rates, up to some maximum, as well as flexible durations and frequencies. In some irrigation districts, requests must be made from 1 to 3 or 4 days in advance of the desired delivery date. The maximum rate is set fairly high to economically use labor on the farm, but it is limited by the economics of the total farm and distribution systems. With this limited-rate demand schedule the rate may be varied by the irrigator during irrigation, and the irrigator controls the duration within the requested 24-h period. Obviously, this schedule requires operational spillage, storage, or automation of the supply because the irrigator may vary the flows. Regulating reservoirs established at several locations along the mains or at heads of laterals are suitable. They can be filled at night when on-farm demand drops off. This permits mains and sub-mains to operate near peak capacity most of the time. When sub-mains and laterals are automated by use of level-top, float-controlled canals, or closed or semi-closed pipelines, operations become very simple (see Section 11.2.3). Main canals need to be automated so that gates at all checks can be simultaneously set to pass any flow demanded any place in the system. Peak flow demands are met from the strategically placed reservoirs so main canal flows are relatively stable. Non-automated, sloping main canals can also function satisfactorily if they have reservoirs to accumulate water that would otherwise be spilled. They must have adequate flow to satisfy the day-to-day peak demand in the upper reach, and an adequate automated service area below the reservoirs to utilize the operational variations. On-line system reservoirs may be located slightly above or below the canal and the necessary flows pumped into or out of them. The lifts and quantities are often small. The grade may be reduced on a steep canal to provide the elevation differences needed for the reservoir either by reconstruction or by adding check structures. The limited-rate demand schedule is highly practical for the water user. It enables the farm operator to irrigate each crop when needed, and to use a stream size that is economical and efficient for the particular situation. Thus, differences in soils and crop requirements can normally be accommodated. Limitations on crop types that can be successfully grown are reduced. A ditch rider, who is an agency employee, may need to be on call 24 hours a day to reduce operational spillage as the farm operator changes his flow rate or finishes irrigating, unless the system has automated canals or closed or semi-closed pipelines, and suitable reservoirs. Nearly all large projects in the northwestern U.S. operate on a type of arranged schedule. In some cases it is possible to deliver water on a straight demand schedule during part of the season, and change to one of the arranged systems or even to a rotation schedule during peak crop-growth periods. The demand schedules are practical when the discharge from the supply system has adequate capacity, can be varied, and has automated canals and/or closed or semi-closed pipelines. A totalizing meter measures the volume of water, but the irrigator controls the flow rate and duration (Merriam, 1973, 1977). With a fully adequate capacity, it is not necessary to schedule the delivery day, thus greatly reducing operation and maintenance costs.

352

Chapter 11 Delivery and Distribution Systems

11.2.2 Flow Control and Measurement to Farms Good irrigation water management requires knowledge of the quantity of water available. Water at rest (i.e., stored water) is measured in units of volume, such as cubic meters (m3). Water in motion is measured in units of flow rate, as volume per unit time, such as cubic meters per second (m3/s) or liters per second (L/s). Many methods are available for measuring water deliveries to farms and for measuring rates and volumes of water on-farm, but accuracy varies widely. Generally, flow meters can be classified into two major groups: those for measuring the total quantity of flow (totalizing meters) and those for measuring flow rate (rate meters). With appropriate auxiliary equipment, meters from either group can be converted to the other group. 11.2.2.1 Totalizing meters. The most common totalizing meters are the propeller meters that measure flow in pipelines. Direct totalizing meters for open channels are rare. One example is the Dethridge meter, which is used mainly in Australia to measure accumulated delivered volume flows in canals. More recently, open channel meters based on use of ultrasonics can provide flow rate and accumulated volume and are thus totalizing meters. Propeller meters. Propeller meters are often used to measure discharges from wells and for pipeline deliveries to farms. In such service, they are usually fitted to show both totalized flow quantity and flow rate. These meters use a multi-bladed propeller with two to six blades, made of metal, plastic, or rubber, that drive the indicating device to give any desired volumetric and rate units. Modern meters may use magnetic coupling between the propeller and the indicator, and the indicator may be electronic rather than mechanical. The meters are designed and calibrated for operation in closed conduits, usually round pipes, that are flowing full. The propeller diameter usually ranges from 50% to 80% of the pipe diameter. Meters are commercially available for pipe sizes ranging from 0.05 to 1.8 m in diameter. The effective low-to-high flow range for an individual meter is about 1:10. Ordinarily, these meters will operate within flow velocities ranging from 0.15 to 5.0 m/s. The principle involved is simply counting the revolutions of the propeller as the water passes, the rate of revolution being proportional to rate of flow. Changes in the pattern of flow approaching the meter or changes in the bearing friction or the resistance of the propeller surfaces affect the accuracy. Routine accuracy of a wellmaintained propeller meter is usually about ± 5%. Higher accuracy may sometimes be achieved, particularly if calibrated in place. Replogle (1999) and Replogle and Wahlin (2000) describe one method of checking propeller meters in systems with a free outflow. Propeller meters are sometimes used to measure flow in irrigation canals by directing the flow to be measured through a length of pipe that is flowing full. Often a submerged culvert or pipe turnout serves this purpose. Because of debris problems, the meter may be placed in the flow just long enough to get a valid measurement of flow rate and then removed. Flow rate is then assumed to be constant throughout the delivery period to produce totalized flow. However, it may be left in place for the entire time of a delivery and actually totalize the flow, as may be needed for the limited-rate demand schedule. Dethridge meters. Of primarily historical interest, Dethridge meters were designed by J. S. Dethridge of Australia in 1910 (Dethridge, 1913) and have since been used extensively on that continent (Bos, 1989; Replogle, 1970). A Dethridge meter is basi-

Design and Operation of Farm Irrigation Systems

353

cally an undershot waterwheel turned by the canal water passing through its emplacement, a short concrete section specially shaped to provide only minimum practical clearance of the lower portion of the wheel. The Dethridge meter registers volume of delivered water. It is simple and sturdy in construction and operates with a small head loss. These are being phased out in favor of rate meters with totalizing and rate indication in the form of overshot weirs, on the argument that the Dethridge meter is a safety hazard. It has not been used, to the authors’ knowledge, in the U.S., probably because of its bulkiness and safety concerns. This is a direct totalizing meter for open channels, but flow rate can be obtained by using a stopwatch or other time records. 11.2.2.2 Rate meters for canal-system deliveries. Rate meters that are most commonly used in canal-system deliveries to farms are orifices, weirs, and flumes. Orifice, Venturi, ultrasonic, electromagnetic, and vortex meters are available for pipeline deliveries. Vortex and electromagnetic meters are not commonly used in irrigation practices. For a discussion of these, refer to USBR (1997). Orifices. Orifices are used for both canal flow measurements and in pipelines. Orifices are openings, of well-defined area, in a plate or bulkhead, the tops of which are well below the upstream water level in the case of canal flows. If the orifice discharges into the air, it is a free-discharging orifice. If it discharges under water, it is a submerged orifice. Free-discharging orifices require considerable head and are usually limited to special locations where head loss is not critical. Submerged orifices require less head and are more commonly used. Orifice shapes vary widely. Circular, sharpedged orifices are easily machined. Segmental orifices, which can be made by inserting a plate part way into the side of a pipe to form semicircular openings, are used in some pump discharge pipes because they will pass both air and sand if properly oriented. In practice, circular, sharp-edged orifices placed into canal service are fully contracted so that the bed and sides of the approach channel and the upstream water surface are remote enough to avoid influencing the discharging jet. Usually orifices formed by partly opening a gate are suppressed from one or more sides and may require special rating procedures to be sufficiently accurate. The term “suppressed” means that the contraction of the flow does not occur, that is, the contraction is suppressed. A general disadvantage of orifices is debris clogging that prevents accurate measurements. It is frequently difficult to tell if partial clogging has occurred, especially in sediment-laden water. To prevent canal overtopping, the elevation of the top of the orifice wall or other emergency overflow should be at or slightly below the maximum expected upstream water level so that flows can be safely bypassed if the orifice plugs. The head-discharge relation for circular, sharp-edged, submerged orifices is:

Q = C d A 2 g ( h1 − h2 ) 3

(11.1)

where Q = discharge rate, m /s Cd = discharge coefficient, ranging between 0.57 and 0.61 A = area of the orifice opening, m2 g = gravitational constant, 9.8 m/s2 (h1– h2) = head differential across the orifice, m. For freely discharging orifices, h2 = zero and hl is measured from the center of the orifice. Orifices should be installed and maintained so that the velocity of approach is negligible. Circular orifices can produce accuracies within about ± 1% if properly machined, installed, and maintained.

354

Chapter 11 Delivery and Distribution Systems

Other orifice devices. Orifices are much less sensitive to head variations than overfall weirs and flumes because they respond only to the square root of the head change. A variation that is even less sensitive is the NEYRTEC module gate, which delivers water at a nearly constant flow rate, even though the head may vary considerably (Goussard, 1987; Bos, 1989). Fixed guide vanes cause flow that modifies the vena contracta in the module, reducing or increasing the area of the exit flow enough to approximately compensate for changing head within a limited range. These modules are equipped with a separate slide panel that acts as a simple on-off gate. This gate is not intended for flow regulation. Flow regulation is accomplished by changing the number of parallel mounted modules. The constant-head orifice (CHO) is a combination regulating and measuring structure used on turnouts to farms. It usually consists of two sluice-type gates in tandem, installed about 1 m apart. The CHO was developed by the U.S. Bureau of Reclamation (USBR, 1997) and is so named because it is operated by adjusting the downstream gate to maintain a constant head differential (h1 – h2), across the orifice formed by the upstream gate. The differential head is usually 0.06 m. The discharge will then be at the required value to an accuracy of about ± 7%. Because of the small head differences, extreme care is needed for setting the required head readings. Sharp-crested weirs. Sharp-crested weirs (Brater and King, 1976; USBR, 1997) are overflow structures made from thin plates whose thickness at the overfall edge is less than 2 mm. The weir opening is usually rectangular, triangular, or trapezoidal, although other shapes have been used. These weirs are accurate, i.e., will repeat their calibration to within 1% or 2%, especially the rectangular and triangular types. The trapezoidal versions are considered to be reliable to ± 5%. The most popular trapezoidal style is the Cipoletti weir with 0.25:1 side slopes (horizontal:vertical). To obtain high accuracy, the nappe (which is the falling sheet of wather springing from the weir plate) must be ventilated on the downstream side of the weir plate so that the pressure there will be atmospheric. If the weir is not ventilated, the resulting vacuum will cause variable and undetermined over-discharging. The downstream water level should be kept at least 5 cm below the crest of the weir. This means that the required head loss always exceeds by 5 cm the upstream head over the weir crest. This required high head loss is one of the major limitations to using sharp-crested weirs for irrigation flows, especially in relatively flat irrigated areas. The discharge over rectangular sharp-crested weirs can be approximated with a power function of the form Q = CLH 3 / 2 (11.2) where C incorporates the gravitational constant, the discharge coefficient, and other effects; L is the weir crest length; and H is the total energy head on the weir (the flow depth plus the approach velocity energy). Methods for determining C are given by Brater and King (1976). For design purposes, C = 1.66 for SI units or 3.0 for footpound (customary) units. Methods to more precisely define K for rectangular sharpcrested weirs are also given by Kindsvater and Carter (1959) and Bos (1989), and include other shapes of opening and the effects of weir suppression. Critical-flow flumes. Critical-flow flumes are rate meters particularly suited for measuring irrigation water in open channels. They consist of a specified constriction built into the canal floor and/or sidewalls that raises the upstream flow level above that which would have existed without the constriction. This rise must create enough drop

Design and Operation of Farm Irrigation Systems

355

to cause critical flow in the contracted section, or throat, of the flume, and is frequently less than 10% or 15% of the original flow depth. This makes them attractive for measuring canal flows in many relatively flat irrigated areas of the world. Of historical interest, because of their wide applications for more than a halfcentury, are Parshall flumes (Parshall, 1950; USBR, 1997), followed by H-flumes (Bos, 1989; USDA, 1979), cutthroat flumes (Skogerboe et al., 1972), and special trapezoidal and triangular flumes (Robinson, 1968). These styles, however, permit curvilinear flow to occur in the control section, making them hydraulically complex with their calibrations depending on laboratory ratings. Additionally, most of the trapezoidal and triangular flumes require depth measurement in a converging flow section, which is not favored by modern hydraulics practice. Because of the complex hydraulic behavior and method of head sensing, these flumes must be carefully constructed throughout their length, and can become relatively expensive devices. Another general type of flume is the long-throated flume, which may be contracted from the sides and bottom, and which includes some broad-crested weirs that are contracted from the bottom with rounded upstream edges. These have been used for many years in Europe (Ackers and Harrison, 1963; Ackers et al., 1978; Bos, 1989). These are called long-throated flumes because the contraction length is about twice the maximum flow depth for styles resembling broad-crested weirs, causing nearly parallel flow as it passes through the flume throat. For side contractions only, the throat length may need to be based on a length-to-width ratio. This is because the shock waves from the contraction at the edges must cross to the center of the flow for the central region to “know” that the side contraction exists. A study of this effect is presented in Wang et al. (2005). This shock wave propagates at about 45° across the channel, which would suggest a minimum throat length exceeding one throat width. The cross-sectional shape can be almost any regular geometric pattern— rectangular, triangular, trapezoidal, circular, parabolic, or complex. The flow depth is sensed in an upstream channel section, again, in an area of parallel flow. The stagedischarge ratings are based partly on calculated relations, combined with laboratorydetermined discharge coefficients, and usually meet the desired ± 5% accuracy requirement demanded of the flumes. More recent calibration procedures, involving mathematical modeling, velocity distribution, and considering boundary layer development caused by frictional effects, improve the reliability of these calibrations to within ± 2% (Bos et al., 1991; Clemmens et al., 2001). Simplified field installation techniques have greatly decreased the skill requirements of field construction. Individual discharge ratings can be restored to the original intended accuracy in spite of most field construction anomalies. An important aspect of flumes is the high submergence limit that can be tolerated as compared with sharp-crested weirs. (Submergence limit is defined here as that point of submergence caused by downstream water level at which the real discharge deviates by 1% from that indicated by the observed flow stage.) A high submergence limit minimizes required head loss for the measuring device in the canal system. Because sharp-crested weirs can withstand no submergence, the minimum required head loss is equal to the flow depth over the weir plus about 5 cm, to assure free flow. Flumes, on the other hand, have submergence limits that range from about 50% to as high as 95%. This corresponds to absolute head losses through flumes of about 5% to 50% of the canal flow depth.

356

Chapter 11 Delivery and Distribution Systems

The submergence limit of critical-flow devices, like flumes and weirs, is a function of the streamline curvature at the control section and the losses in the downstream expansion. Broad-crested weirs and long-throated flumes, which have straight and parallel streamlines at the control section, can approach a submergence limit of 95%, if part of the kinetic energy is recovered by a proper exit expansion (Bos, 1989). The submergence limits for truncated long-throated flumes, i.e., flumes that have no expansion section, are about 70% to 85%, depending on whether the throat area approximates a rectangular or a triangular shape. These submergence limits, even for truncated versions, are higher than those routinely listed for H-flumes, cutthroat flumes, and Parshall flumes, which are 25%, 65%, and 70%, respectively (Chow, 1959; Clemmens et al., 2001) for the usual sizes of flumes on farms. The older flumes, however, are quite usable, and many are in continual service and most certainly need not be replaced if they were properly installed and subsequently maintained. Abt et al. (1989) and Abt and Staker (1990) present ratings corrections for flumes that have settled out of level, either longitudinally or laterally. Accurate and much less expensive designs are now available for new installations. These have the added capability of being tailored to most existing channels. Many open channels can now be retrofitted with a low head-loss measuring device that can operate at high submergence with a single depth reading. When installing any of the critical-flow flumes, care must be taken to assure adequate free flow discharge at both high and low ranges. Checking only at high flow is risky. For example, a rectangular flume placed in a trapezoidal channel, where the channel, rather than a check gate, controls the flow level, may be free flowing at high stages but submerged at low flows (Replogle, 1968). Usually, combinations of these shapes need to be checked for submergence at low flows. On the other hand, a trapezoidal or triangular flume placed into a rectangular channel will require checking the calculations at high flows and will tend to become less submerged at lower flows. Rectangular flumes, including the Parshall and the cutthroat, are sometimes installed with a vertical offset to avoid submergence at low flows (Replogle, 1968; Humphreys and Bondurant, 1977), at the expense of higher head loss at maximum flow. The trapezoidal, broad-crested weir with upstream approach ramps is a variation on the long-throated flume. It is particularly suited for the common, small concrete canals, constructed in a continuous manner with sliding forms, which are called slipformed canals. Such weirs can inexpensively and accurately measure irrigation flows. They are usually truncated at the downstream end of the weir sill, although an exit ramp can improve the submergence limit from about 85% to 95% in some configurations. Clemmens et al. (2001) present design information for broad-crested weirs in trapezoidal and other channel shapes. Analysis shows that the error in discharge is almost proportional to the error in constructed sill width (Replogle, 1977; Replogle et al., 1997; Clemmens and Replogle, 1980), and corrections to the rating tables due to these types of construction anomalies can usually be accurately adjusted. A rectangular broad-crested weir discharges nearly equal quantities of water over equal widths. The flow is nearly two-dimensional over the weir, so that rating tables can provide the flow rate, q, in m3/s per meter width of sill for each value of h1. Design and calibration information for these weirs is given in Clemmens et al. (2001) and Wahl et al. (2005).

Design and Operation of Farm Irrigation Systems

357

Portable flumes for flow rates up to about 50 L/s are described in Bos et al. (1991) for trapezoidal and rectangular cross-sections. Adjustable sill-height flumes are now available that allow easy installation at an estimated elevation and applying postinstallation adjustment to flow conditions. They can be used in both lined and unlined small canals. Commercial offerings are available in sizes up to 1000 L/s, although only the smaller sizes of less than about 300 L/s maximum capacity are considered portable. Simple constructions for small trapezoidal canals. The structural requirements of broad-crested weirs of the long-throated type described above allows simple construction methods in the common-sized trapezoidal canals with 30- and 60-cm bottom widths, 1:1 side slopes and capacities on the order of 500 L/s. Construction suggestions are shown in Figure 11.1. The concrete requirements are quantities that can be hand mixed. Structures for circular channels. The broad-crested style of long-throated flume has been adapted for flow measurements in circular channels and pipes with a free water surface because of its relatively simple construction. Other shapes, such as side contractions and center line piers, have been used. The reader is referred to the computer program and design procedure for most calibrations (Clemmens et al., 1993a, and the update to Windows XP provided in Wahl et al., 2005). There are two situations of particular interest for designing a weir in a pipe. The weir can be placed somewhere in the middle of a straight pipe, or at the end of a pipe, such as the entrance to a deep manhole, which is probably more common because the site is more accessible. When the weir is placed in the interior of a straight pipe, the presence of the weir must cause a rise in the upstream water surface to develop the required head loss. This increase in flow area upstream causes a proportional decrease in velocity and may cause subsequent sediment deposition. Where this is a problem, a downstream ramp should be considered to reduce the required head loss and thus accommodate a smaller sill height. Sediment problems are further aggravated at low flows because the freeflowing (normal depth) water surface drops proportionally faster than the water surface upstream from the weir. Thus, the velocity difference becomes greater and greater with decreasing discharge. For situations where wide fluctuations in flow rates exist and sedimentation is a problem, an alternative weir shape or location should be considered. An alternative for the measuring location is the end of a pipe, particularly where there is a large drop in water surface. A weir can often be designed so that the water level upstream from the weir either matches or is below the normal depth of water in the pipe. This may avoid aggravating sedimentation problems. Portable flumes using plastic pipe. Circular pipes can also be used as the basic format for portable and permanent structures (Figure 11.2). For portable use, the translocated stilling well is convenient because it renders the device less sensitive to longitudinal and transverse leveling. The installation and use of the translocated stilling well is described in detail in USBR (1997), and is similar to that shown in Figure 11.2. A broad-crested weir, usually 0.2 to 0.5 times the pipe diameter in height, is placed in the pipe. The free-flow limiting depth in the pipe is about 0.9 times the pipe diameter. The most convenient approach to the general calibration computations is application

358

Chapter 11 Delivery and Distribution Systems Drain tube allows canal to empty without pumping. Usually needs cleaning with a rod when drainage is needed.

1. Select a concrete lined

2. Fill with field soil to within 5 to 10 cm

3. Put steel angle-iron forms in place.

canal that is free of major

(2 to 4") of finished broad-crested weir.

These must be carefully shimmed to

cracks. Place a drain tube

Temporarily form a downstream

level in all directions. See below for

(about 3 cm (1") plastic) in canal.

ramp with the soil even if a concrete

example construction of these forms, which are left in the concrete.

Concrete thickness Reinforcing bars may be added to ramp.

Soil

3

should

1

wash away from under ramp when canal flow starts.

4. Assure that the forms stay

5. Pour the ramp on about a 3:1

6. Mount sidewall gauge at proper location

level while filling with concrete.

slope. This slope is not critical.

and elevation. Usually set elevation of most used flow rate. Check zero and evaluate if gage and canal wall slope are well enough matched (±3 mm ( ±1/8")) is usually acceptable)

e ug Ga

Mo

tin un

gB

ra

et ck

Wall-gage mounting bracket made from galvanized sheet metal.

Wall gauge can be made using

Attach to canal wall with screws.

a chisel and punch to make the marks and numbers. A paper pattern useful as a chisel guide.

Use 4 cm (1-1/2 ") angle or larger. Cut ends to match canal side slope.

After sliding gauge into Need be only approximate

Form for sill

bracket to proper vertical

concrete to lock it tightly.

Make as accurate as practical

Figure 11.1. Simple construction method for long-throated flumes in small farm canals.

of the WinFlume calibration software for long-throated flumes of many crosssectional shapes presented in Clemmens et al. (2001) and updated in Wahl et al. (2005). The following design and calibration information was produced using the WinFlume computer model and its predecessors. Qcfs = ( D ft )2.5 K1, ft ( h1, ft / D ft ) + K 2 , ft )U

(11.3)

Design and Operation of Farm Irrigation Systems

359

Stilling Well Location

To D y1 Stilling Well

bc

h1

y2

p1 La

Lb

L

Figure 11.2. Broad-crested weir for flow measurement in a circular culvert.

where Qcf s = discharge, ft3/s (for SI units, multiply Qcfs by 28.32 to obtain L/s) Dft = diameter of pipe, ft (multiply SI units in m by 3.28 to obtain ft) K1,ft = constant from Table 11.2 K2,ft = constant from Table 11.2 h1,ft = head measured from sill top (bottom of contracted section), ft (multiply SI units in m by 3.28 to obtain ft) U = exponent. As mentioned, the flumes represented in Table 11.2 are Froude models of each other. These and all the long-throated flume shapes can be similarly scaled without using the computer model, as long as all dimensions remain proportional. Small differences from direct computer results are to be expected because the roughness of construction materials is not usually scaled. Smooth concrete roughness was used to develop the values in Table 11.2. The calibration equations from Table 11.2 will usually perform satisfactorily to within ± 3%, not counting the error of the readout detection method, for scale expansions or contractions of about 5. Scale expansions of 10 tend to over-emphasize roughness and will under-predict discharge by 5% to 10%. Accuracy at the ± 2% range requires individual computation of the constructed device using the computer model of Wahl et al. (2005). As with the other broad-crested weirs, and flumes, the width of the sill surface, w, is one of the two most important dimensions in the flume. The other is the accurate referencing of the zero elevation of the head-measuring device. For portable measuring weirs, the translocated stilling well, which references the upstream head to the sill floor without the necessity of accurately leveling the flume, is recommended. This minimizes errors in head determination even if the weir is not exactly level (Bos et al., 1991; Wahl et al., 2005). The upstream gage should be used only if the flume is accurately leveled or is permanently installed.

360

Chapter 11 Delivery and Distribution Systems

Table 11.2. Equation parameters for sills in circular flumes. These values represent minimum lengths in direction of flow and may be increased 30% with only slight change in calibration, and should be suitable for most applications.[a] Pre-Gage Dist., Lpg ≥ hmax Sill Height = p1/D hmin Approach, Lap ≥ hmax = 0.07 D = [0.85 D – p1] Converging, Lcv = 3 p1 hmax Control, Lcn ≥ 1.5 D – p1 K2,ft K1,ft Qrange/D5/2 (ft) (ft) (ft3/s) p1/D La/D Lb/D L/D U hrange/D bc/D 0.20 0.65 0.60 1.200 4.130 0.004 1.736 0.080-0.65 0.056–1.975 0.800 0.25 0.60 0.75 1.125 3.970 0.004 1.689 0.070-0.60 0.048–1.689 0.866 0.30 0.55 0.90 1.050 3.780 0.0 1.625 0.070-0.55 0.050–1.434 0.917 0.35 0.50 1.05 0.975 3.641 0.0 1.597 0.065-0.50 0.046–1.202 0.954 0.40 0.45 1.20 0.900 3.507 0.0 1.573 0.060-0.45 0.042–0.991 0.980 0.45 0.40 1.35 0.825 3.378 0.0 1.554 0.055-0.40 0.037–0.807 0.995 0.50 0.35 1.50 0.750 3.251 0.0 1.540 0.050-0.35 0.032–0.640 1.000 [a]

The values in the table are only valid for foot-pound (customary) units. To use SI units, first convert D and h1 from m to ft by multiplying by 3.28 ft/m for entry into table, then convert the resulting Qcfs to m3/s or L/s. Alternately, use the computer program of Wahl et al. (2005).

In Figure 11.2, the stilling well is placed at the usual location. However, it can be placed at another location, such as near the outlet, provided the sensing of depth at the gage location is accurately transmitted to the alternate location. It can even be placed in the channel if it does not significantly obstruct flow or divert flow to one side and cause erosion. Placing it in the upstream channel often causes detrimental flow patterns that can affect the function of the flow-measuring device, unless it is dug deep into the bank or placed at substantial distance upstream. Alternately, a static head tube, consisting of several holes about 3 mm in diameter drilled into a length of 2- or 3-cm diameter PVC pipe, can be used to monitor the pressure at the gage location as shown in Figure 11.3. These holes are about 30 cm from the end of the capped pipe so that flow separation around the end of the pipe is negligible by the time the flow passes the pressure sensing holes. The water level sensed here is transmitted to the stilling well where the depth can be observed by any of the several flow depth methods discussed elsewhere. The system shown can be used as the primary readout method or it can be used to properly set the zero for another depthsensing device. If a static pressure tube is a permanent part of, for example, a stilling well placed downstream, then the tube should be clamped tightly to the wall of the flume so that debris trapping is minimized, and well above the floor level to reduce sediment plugging. The area obstruction of the pipe crossing the sill control area is small and can usually be ignored, except in small flumes where the pipe represents more than about 1% of the flow area. Head reading, h1detected in suspended cup

Inlet holes for static head tube h1

h1

SIDE VIEW

END VIEW

Figure 11.3. Using a static head tube for a portable flow-measuring system in a half pipe.

Design and Operation of Farm Irrigation Systems

361

11.2.2.3 Rate meters for pipe system deliveries. Rate meters commonly used in irrigation pipes are the propeller meters mentioned in 11.2.2.1, which can also totalize flow. Differential-head meters, such as the orifice meter for pipe flows, operate on the principles described in section 11.2.2.2.1. Other selected meters described herein include Venturi meters and special ultrasonic meter applications. Venturi meters. Venturi meters are applicable to irrigation pipeline flows but are not widely used for routine farm turnout use because their length and the straight-pipe requirements usually make them impractical. Despite this, they represent one of the older and more reliable of the differential head meters. The head loss is relatively low, and they can pass slurries readily. In irrigation works, small Venturi meters can be used to create a vacuum for chemical injection applications. In sizes compatible with irrigation wells, they are usually considered too costly. The flow range is narrow but similar to that of the orifice meters (about 3 to 1, high- to low-flow rate). These devices are well covered in the literature. The reader is referred to ASME (1971), and to Brater and King (1976), for a more complete treatment. Field-constructed Venturi meters. Special Venturi meters constructed from plastic pipe fittings address the economic limitations of commercially available meters, fouling problems, and water management requirements. Individual calibration of each constructed meter is not required for effective applications in irrigation practice. Variability between constructions is usually within ± 2% and the head loss is between 18% and 25% of head reading (compared to 10% to 14% for the “standard” Herschel-type Venturi tube) (ASME, 1971). The discharge coefficient is also slightly less than the Herschel-type tube. A constant value of Cd = 0.95 is suitable for most irrigation applications. More complete details are given in Replogle and Wahlin (1994). General construction details are shown in Figure 11.4. App ly suc tion or pressure to ad just the differential pressure to a c onvenient

Clamp

Pre-m arked flow m ea surem ent stic k

rea ding level

No bubbles in pressure tube s Air vent (o ptiona l)

Standard Venturi

Plastic -Pipe Venturi

Figure 11.4. Construction of plastic-pipe Venturi meter showing differential head arrangement for direct discharge indication.

362

Chapter 11 Delivery and Distribution Systems

Venturi meters and orifice meters can be made more convenient for the user if a scale is designed to indicate flow rate directly. For any given meter, a scale can be produced that is used by placing the bottom at the water level in one leg of a manometer and reading flow rate directly at the level of the other leg. No subtracting of readings and no table look-up is needed (Replogle and Wahlin, 1994). The liquid levels in the manometer can be raised or lowered to suit the observer without changing the differential (Figure 11.4). Ultrasonic meters in irrigation applications. An exception to high costs that usually preclude on-farm applications is a single-path sonic meter in 76-cm concrete pipe used on farm turnouts in some irrigation districts in Arizona. Such meters operated satisfactorily in most installations (Replogle et al., 1990). Large random error fluctuations were associated with partly opened vertical inlet gates combined with typically 45° horizontal pipe bends that tended to send jets in a spiral motion to the meter section downstream. Three-minute averages were required to produce acceptable flow rate readings. Totalized flow volume over 24 h or longer using a 16-s sample time every 15 min was sufficiently accurate because the instantaneous errors were random and tended to cancel. However, observing a single reading that was often 15% (and once observed to exceed 30%) of average flow was not acceptable for on-farm water management use. Hydraulic fixes, such as installing large-diameter orifices with openings of 85% to 90% of the pipe diameter downstream from the last pipe bend before the meter, appeared to cause enough cross mixing to mitigate the jets. The resulting flow readings were more stable, fluctuating less than 5%. Ultrasonic Doppler meters. In the early 2000s several commercial sources began promoting ultrasonic Doppler meters that can be used in both canals and pipelines. These devices and commercial offerings are changing rapidly, and the reader is advised to check manufacturers’ literature for operational applications and costs. 11.2.3 Delivery Systems Suited to Mechanized Irrigated Systems With the increased appreciation of the value of higher on-farm irrigation efficiency to save water, labor, and energy, there is an increased tendency for mechanizing and automating supply, distribution, and application systems. To obtain these savings, the whole system must be able to deliver water in a way that is flexible in frequency, rate, and duration. For best operation, the flow must be controlled at the point of application by automatic controls, or mechanized so that flow is easily controlled when the irrigator is making on-the-spot decisions. Automatic systems are usually thought of as being capable of initiating application of water to fields when irrigation is needed, sequencing sets, and shutting off flow when irrigation is complete. Semiautomatic (sometimes called mechanized) systems may need an operator’s periodic attention to initiate a sequence or to reset components, such as drop gates. The ability to control the rate and duration of flow at the point of application is essential to obtain high irrigation efficiency. For surface methods, large stream sizes reduce the application time and labor requirements. For all irrigation methods, it is essential that the water be turned off when the proper depth has been applied, because additional water is wasted. Automated application systems have in-field controls responsive to flow volumes, head, duration, soil moisture, etc., to apply the desired irrigation. Therefore, the delivery system must also automatically make the desired flow conditions available. Such automated application systems require a fair amount of supervision and maintenance. They require a large capital investment but have low nonsupervisory labor requirements. Their flow capacity requirements are moderately large. The systems

Design and Operation of Farm Irrigation Systems

363

may receive nearly continuous use because around-the-clock and weekend operations are less likely to be avoided by users. The semiautomated systems may require the irrigator to be in the field to operate the controls, adjust the flow rates, head, durations, etc., in a fashion that can be preplanned as is done for the automated systems. Moreover, while applying water, the irrigator can make on-the-spot decisions, repairs, etc. The in-field application equipment for semiautomation consists of simple, standard types of controls requiring little maintenance or supervision. Such systems require larger flow capacity because nighttime operation should be avoided, and the stream size should be as large as the irrigator can handle with good manual equipment and design. Semiautomatic systems can allow unattended nighttime irrigation, but may require the operator to check the operation, reset gates, or move components in the evening and again the next morning. Some manual systems may approach or exceed the convenience of semiautomated systems, for example, 2- to 5-ha level basins supplied through manually operated liftgates. These are especially practical on soils with low intake rates, require moderate capital investment, and have low supervisory and moderately low non-supervisory labor requirements (Dedrick et al., 1978; Dedrick, 1984, 1986, 1989, 1990; Dedrick and Reinink, 1987; Clemmens, 1992; Clemmens et al., 1993b). Because of their large capacity and low per-hectare labor requirement, they justify a well-trained and wellpaid irrigator. On average, one irrigator is employed to irrigate about 150 to 200 ha. In a special arrangement of level basins, made possible by gently sloping topography in southern Colorado, one irrigator serviced about 700 ha with manually operated gates (Dedrick, 1984). Highly automated microprocessor-controlled systems, like the liftgate method with air cylinder actuators (Dedrick et al., 1978), require more capital investment and seem most easily justified economically when replacing systems where very frequent irrigations are necessary, using large quantities of labor and small flow rates, or where flow rates are large relative to set size and application depth and water might need to be switched at such short intervals that the irrigator would not have efficient blocks of time for other chores if gates had to be opened and closed manually. Mechanization without full automation is more readily justified with more normal frequencies and larger flow rates, like the system mentioned above. A low-cost system widely used in New Zealand consists mainly of dropgate structures tripped by manually preset timers (Stoker, 1978). Under favorable topographical and soil conditions, as much as 100 ha can be irrigated per hour of labor with this system. The on-farm water delivery system for both the automated and semiautomated systems usually consists of closed or semi-closed pipelines (Merriam, 1973, 1987a,b) or a level-top canal with automatic controls (Merriam, 1977; Burt, 1987). Water should be supplied from a flexible source like a reservoir if delivery is to operate effectively. Closed pipeline systems are easy to operate but line pressure changes as the water level in the reservoir fluctuates and as flow rates are changed. The latter causes head variations due to increased or decreased friction loss. These variations require changes in the field settings for high efficiencies. Outlet controls must be capable of regulating flows at the high static pressure that occurs when flow is stopped, as well as the lower pressure when the pipeline is flowing at capacity. When there are outlets to several laterals that can be used simultaneously, the pipe sizes may be reduced in the downstream direction.

364

Chapter 11 Delivery and Distribution Systems

Semi-closed pipeline systems have the pressure broken into steps of 6.0 m or less by downstream water level float-controlled valves. Other water level control systems exist but are more sophisticated and expensive and are seldom justified for on-farm distribution systems. The semi-closed systems may be used to automatically provide variable flow, to downstream rate requirements, and still have small (0.1 to 0.3 m) head variations at turnouts. The float valves eliminate water hammer, and head loss at each float stand is about 1.5 times the velocity head for full flow. Level-top ditches can deliver water effectively to automatic or semiautomatic onfarm application systems. Such ditches permit a no-flow condition to exist. The noflow or variable flow condition is controlled by float-actuated gates maintaining constant downstream water levels, or by more sophisticated devices better adapted to larger canals (Burt, 1987). These systems are actuated by variations in the water level downstream created by the variable flow needs in the field. The bottom of the ditch may be parallel to the level top or have some slope. In either case, the water surface level decreases at a turnout, creating a negative wave at critical velocity back to the gate, which then supplies the desired flow rate. The initial withdrawal pending the gate operation is taken from canal storage as the water level drops. The canal’s cross-section usually corresponds to a design velocity of about 0.3 to 1 m/s. The level reaches can be a kilometer long for typical on-farm conditions, and longer on large canals. The length of these reaches must be shortened on steep gradients. Mechanical gates for controlling flows in such ditches are available from the French company ALSTHOM with their NEYRTEC Design equipment. Their AVIS gates are suitable for canal drops of up to about 1 m (Goussard, 1987). Their AVIO gates for reservoir outlets can control heads up to about 3 m. For greater heads, it may be feasible to use a float valve system like the Harris float valve described by Merriam (1973, 1987b). Canals controlled by the NEYRTEC gates have small variations in flow depth and thus require only a small freeboard. Because of the reduced freeboard over normal operations, canal costs are only a little greater than for low-gradient canal systems less than about 1 km long. The additional cost for gates makes possible the automatic operation of the delivery systems. Flow measurements need to be made by totalizing meters because flow rates are expected to vary. Microprocessors show promise of economical application to canal flow controls and can be expected to improve canal operational capabilities. Labor, water, and energy savings are possible when additional capital is invested to provide automation or semiautomation of the application system and to modify the delivery system to provide for the needed flexibility in frequency, rate, and duration (see Section 11.2.1). 11.2.4 Reservoirs in Farm Water Delivery Systems Under many circumstances, small reservoirs are essential elements in a complete farm water delivery system. They may be built and owned by the using farm or may be designed and built by the irrigation district for canal regulation. They may be a practical means for upgrading the scheduling system that can be offered by a district. For example, to upgrade a fixed-amount, fixed-frequency schedule system, which would characteristically have laterals too small to directly upgrade to one of the demand schedule systems, strategically placed, off-canal reservoirs can frequently be constructed at less cost than enlarging and expanding existing canals and right-of-way. This would be especially true if the existing canal were already lined, and relining costs were to be considered for a larger canal section.

Design and Operation of Farm Irrigation Systems

365

Basically, reservoirs are management facilities whereby a variable or steady flow can be regulated into a different variable or steady flow that is more convenient or efficient in terms of water, labor, energy, and crop production. They may be used in connection with pump or gravity systems and with all irrigation methods, and are often combined with other facilities like pipelines, wells, pumpback systems, and canal delivery systems. 11.2.4.1 Reservoir use and management. Classification of reservoirs may be done by major function: (1) long-term storage, (2) temporary storage, (3) “overnight” storage, (4) regulation, (5) tailwater reuse, and (6) water table control. Farm reservoirs are often designed to serve several of the above purposes as well as providing water for frost control, fire protection, domestic use, spraying operations, or recreation. Long-term storage reservoirs are usually associated with project operation and year-to-year carryover. They have large capacity and basically are needed to provide flexibility in frequency. Their occurrence on-farm is rare though they have been economically used in conjunction with low-capacity wells where water can be pumped in the spring season, stored, and then used to meet peak summer demands. Temporary storage facilities are commonly used in the more humid areas to collect runoff from rainfall for use during occasional dry periods or for frost or temperature control. They are often used to provide a source for irrigation systems in areas where rainfed cropping is possible but where irrigation can improve crop quality and yield. “Overnight” reservoirs are common. They accumulate pump or canal deliveries for 24 h, more or less, to permit applications that are variable in rate and duration. They may have a capacity as small as that needed to hold 12 to 16 h of flow from a well or small surface source with the irrigation applied for only part of the 24-h period. However, this small size lacks reserve for unusual conditions and a larger capacity for up to 48-h storage is frequently more desirable, particularly in conjunction with a semiautomated surface system. They often facilitate the continuous pumping of a well or the 24-h delivery flow from a water agency. The economical size is a function of the duration of irrigation, the area to be covered in a set, the capacity of the distribution system, the economics of labor, and the water supply rate and duration of flow. Regulating reservoirs are usually small. Their function is to permit relatively small variations in flow and/or duration. They may be needed to accept flow from a water agency or a well to absorb small variations in flow to sprinklers as the length of lateral changes or time to move reduces flow needs for short periods. One may be needed adjacent to a well supplying a sprinkler system so that variations in pumping rate due to a falling water table can be equalized by the booster system to provide a steady rate and a constant pressure. One may conveniently be used to accumulate enough water to provide the larger initial stream desired for furrows. When the reservoir is empty, the cutback stream size would be satisfied by the basic source such as a well, and the reservoir could be used to accumulate the tailwater to be ready for the next set. Extra capacity over the minimum volume is good design. Tailwater reuse systems require a sump or reservoir at the lower side of a field. The capacity needed may vary from complete storage of the runoff from one or more fields to a small sump for a recycling system. If a cycling pump from a small sump is used it should pump to a larger reservoir. If the pumped water is added to the irrigation stream, the runoff that is already occurring will be aggravated. A reservoir constructed to help maintain a water table is sometimes used in subirri-

366

Chapter 11 Delivery and Distribution Systems

gation systems or with ponded crops such as rice. These have no specific storage requirement. Evaporation may cause significant water loss if the surface area is large. Classification of reservoirs may also be done by location: gravity reservoirs serve the irrigated area by gravity flow while low level reservoirs require pumped delivery. Preferably, an irrigation reservoir is located at a high point on the farm so that all flow can be taken out by gravity, particularly for surface irrigation systems. When water supply agencies deliver water at an elevation below the desired water surface of a reservoir, a small pump to fill the reservoir may be desirable. If the reservoir supplies sprinklers, it might be most economical to let it fill by gravity. Then a single pumping system can empty the reservoir and pressurize the sprinklers. The cost of pumping may be compensated by the value of the irrigation of additional areas or by the labor and water saved through more efficient application. On large farms where topography permits, it may be practical to have two reservoirs at the same elevation at the upslope corners (more or less), interconnected by a pipeline or a level-top ditch designed to permit flow in either direction. This permits combining deliveries to the middle to obtain large flows for efficient labor utilization from smaller, less expensive conveyance. A similar condition using flow to the center from two ends exists when the reservoir is at the high side and a well is located at an alternate side. A gravity reservoir, in conjunction with large capacity closed or semi-closed pipelines or level-top ditches, provides a source for deliveries that are flexible in rate and duration and provides large stream sizes that greatly reduce labor costs. Low-level reservoirs may permit the utilization of low swampy areas, may be part of a tailwater reuse system, or may accept gravity delivery from low-head supply canals or groundwater or surface flows. In shallow water table areas, they may eliminate the need for lining to control seepage losses. The use of a pump to deliver the water from the reservoir restricts the flexibility of irrigation stream size, but permits a wider choice of delivery location. When a low or swampy area otherwise unusable for agriculture is converted to a reservoir, the reservoir can sometimes be managed to enhance wildlife habitat, albeit this may be at the expense of wildlife otherwise supported by the swamp. Depending upon the head, pumps used for surface irrigation deliveries from farm reservoirs are usually low lift, high volume propeller pumps or centrifugal pumps. They may feed directly into pipe conveyance systems which in turn may feed pipelines or ditches for water distribution. Higher pressure centrifugal pumps, including close-coupled vertical turbines, are commonly used to supply sprinkler systems directly from reservoirs. When variable flows are desired in the field, using internal combustion engines to power pumps, rather than electric motors, can provide such flexibility. To avoid having to frequently change pump speed to closely match field demand, it is often practical to have an overflow bypass back into the reservoir, designed to maintain the needed head on the pipeline for delivery. In operation, the pump is set to supply the desired field flow with a small excess overflowed through the bypass standpipe, to maintain a constant head. In semiautomated systems with closed or semi-enclosed pipelines or level-top canals, a bypass stand beside the pump and reservoir is often used. Without a bypass, the irrigator must take the full unregulated flow pumped and this almost always results in less-efficient utilization of water and labor. With a bypass

Design and Operation of Farm Irrigation Systems

367

a fine readjustment of the flow rate in the field to match the precise need will cause the small rejected flow to go directly back into the reservoir rather than being wasted in the field or continuing to a tailwater reservoir. The extra pumping cost for the small head on the rejected flow is justified by the conserved water in the field and is less than the cost of repumping tailwater. In level areas, the reservoir can consist of an extensive level-top canal system to provide storage, distribution, and sometimes drainage, depending on the drainage water quality. The reservoir may then border on several fields, and gravity outlets or portable pumps, if necessary, can be located adjacent to the field for which water is being withdrawn, minimizing conveyance piping. When the groundwater table is about the same elevation as the canal, there is little or no seepage loss. Where elevation differences preclude a single-level canal system, several steps in the system can be created by the use of mechanical gates, such as the NEYRTEC AVIS automatic constant downstream level canal gates referred to in Section 11.2.3 (Merriam, 1977). It is practical to use such depressed canals for project transmission canals. Pumping costs may be compensated by the small seepage loss from an unlined canal; and, more importantly, by the elimination of the need for on-farm reservoirs to provide flexibility of rate and duration. Pumps, portable or permanent, from the canal may serve small elevated on-farm canals or pipelines. The pumps should have variable flow and bypass capabilities. An indirect way of obtaining on-farm flexibility in rate and duration is through the use of reservoirs in the project supply system. The management and the project canal system must be able to permit variable flows and duration at farm turnouts, which are under the control of the irrigator. The farm turnouts can be automated by the use of float valves to pipeline systems or NEYRTEC AVIO gates to canals. The irrigator can then use the project system as his reservoir. With this variable outtake, the canal flows vary. To overcome these variations, the canal system may be automated, or a sloping canal section can terminate in a reservoir for re-regulation. The next sections below the reservoir must be treated the same way, and the final section should serve the farms through a system of closed or semi-closed pipelines or level canals. In this way, the project canals and reservoirs can provide the needed on-farm flexibility in rate and duration to permit a limited-flow demand schedule. Flexibility of frequency should still be under the control of the canal operation to match canal capacities. Increased canal capacities may be obtained by canal automation because closer control will reduce freeboard requirements. This permits the limited-flow demand type of schedule. 11.2.4.2 Investigations and design. In many states, laws and regulations may control, limit, or even require the installation of farm reservoirs. Water rights may be affected or public safety endangered by their construction. All applicable state statutes and regulations must be met, and often permits are required before construction is authorized. For such reasons, it is usually best that a reservoir be designed and constructed by qualified professionals. Design services may be provided by appropriate federal agencies or consulting engineering firms. The following checklist includes some considerations for irrigation reservoir design: ƒ The purpose for the reservoir must be clearly defined. ƒ Adequate consideration must be given to prevention of danger to life or damage to property, both in event of reservoir embankment failures and in the course of

368

Chapter 11 Delivery and Distribution Systems

day-to-day usage. ƒ Adequate capacity should be available for sediment inflows. ƒ Capacity provides for potential seepage and evaporation losses. ƒ Excessive inflows can be safely bypassed. ƒ Outlet capacity is adequate to meet irrigation demands. 11.2.4.3 Evaporation and seepage losses. The net depth of the water lost through evaporation during short periods in shallow ponds can be satisfactorily estimated by assuming it to be equal to the evapotranspiration minus precipitation. Because the volume of water lost is dependent upon the surface area, evaporation losses can be minimized by providing a maximum water depth and minimum surface area consistent with an economical balance between construction, maintenance costs, and water loss. Evaporation losses from storage reservoirs that are full year around may exceed 2500 mm per year under adverse climatic conditions and little compensating rainfall. At the other extreme, a regulating reservoir covering a small area may lose 4 to 6 mm per day of operation and up to 10 mm under very adverse climatic conditions. This may represent a very small percentage of the water controlled; e.g., 0.2 ha surface × 5 mm evaporated per day per 5000 m3 of water controlled per day is about 0.2% loss. Even if the reservoir is kept full, though used only one-third of the time, it is still a negligible loss. An overnight storage reservoir with large surface area and a moderately severe climatic condition might lose about 2% of its controlled water. Seepage losses are more difficult to estimate than evaporation losses, but in most instances an adequate appraisal can be based on the permeability of the foundation materials. Local experience with existing reservoirs built in similar soils is perhaps the best guide as to the need for pond sealing. The method of seepage control, for given site conditions and operation requirements, can be determined during the design process and implemented during reservoir construction. Options include compaction of existing earthen materials, treatment with bentonite or soil dispersants, or installation of membrane, concrete, asphalt, or imported earth liners. It may also be practical to delay the installation of seepage control methods until a period of evaluation has determined a need. Seepage losses can seldom be completely eliminated and control methods vary greatly in their effectiveness and cost. Draining an earth-covered membrane-lined reservoir for a few days between uses, to limit evaporation or seepage, will save very little water over retaining it partially full. This is because the water in the earth covering above the membrane will evaporate. Uncovered linings such as soil cement may be emptied but may develop cracks due to temperature changes. Most linings should be kept covered with water to reduce maintenance problems, even though evaporation will be greater than if the reservoir is emptied. Where the economic justification for the reservoir is principally in labor saving because of the larger stream of water made available, or the seepage loss returns to a shallow groundwater basin for repumping, it may not be practical to reduce seepage. Increased irrigation efficiency made possible by the reservoir may easily save more water than the seepage loss. A 200-mm per day seepage from 0.5 ha reservoir area is less than 7% of 150 mm applied on a 10 ha field. However, if the economic value of the water is high in the terms of its productivity or direct cost, very little seepage can be tolerated. In general, the value of the water conserved by lining the reservoir, or of preventing a high water table problem, whether for economic, aesthetic, or environmental reasons, must exceed the cost of the lining and its maintenance.

Design and Operation of Farm Irrigation Systems

369

11.2.5 Mosquito Control in Farm Delivery Systems Irrigation systems can frequently increase existing mosquito problems or create a new problem unless special control measures are exercised. Mosquito populations can be increased if irrigation practices cause the free water surface conditions needed for mosquitoes to complete their life cycle of egg, larval, and pupal stages. Conditions favorable for mosquito production include reservoirs with floating vegetation and debris, water standing for several days on fields or in tailwater ditches, and seepage areas along canals or through leaky dams. Mosquito control planning should be coordinated with other basic irrigated farm needs such as soil and water conservation, flood control, wildlife management, recreation, and, occasionally, hydroelectric power. Proper water management and maintenance of facilities can achieve reduction or elimination of manmade mosquito sources by applying knowledge of both irrigation principles and mosquito biology. Further elaboration on details of mosquito control in irrigation projects is presented in Engineering Practice EP267.6 (ASAE, 2005a).

11.3 FARM WATER DISTRIBUTION SYSTEMS 11.3.1 Introduction This section discusses lined open ditches and pipelines to convey irrigation water on the farm. When buried pipe is used instead of ditches, the conduit can follow the most direct route from the water supply to outlet points rather than following field contours. Weed problems and loss of productive land are eliminated because crops can be planted up to or over the buried pipeline and the pipe does not interfere with planting, cultivating, and harvesting. Seepage and evaporation losses are eliminated when water is transmitted in a well-constructed pipeline. Pipelines present fewer safety hazards than open ditch systems. Pipelines may not be desirable, however, if the irrigation supply contains large amounts of sediment and the flow conditions in the line are such that the sediment would tend to settle out and reduce the carrying capacity of the line. Portable pipe systems laid on the soil surface have many of the advantages of buried pipe systems. If adequate labor is available, they can be removed from fields to allow unimpeded cultural operations. Seepage losses in unlined farm ditches often range from one-fourth to one-third of the total water diverted. In extremely sandy or gravelly ditches, half the irrigation supply can be lost in seepage on the farm. Unlined ditches, because of their potential for high seepage and maintenance expense, should only be considered for temporary use or in cases where money is not available for lined ditches or pipelines. Reducing seepage by using better conveyance facilities may increase water available for crop needs, allow irrigation of additional land, and prevent waterlogging or salination of soils below the ditches. In areas with a history of failure or rapid deterioration of pipelines or lined ditches, the guidelines in this chapter should be evaluated and modified on the basis of local experience. The design of farm pipelines with high pressure ratings is discussed in Chapter 15 and 16. Information on design and construction of unlined ditches can be found in older irrigation engineering texts. 11.3.1.1 Capacity. The necessary capacity of farm conveyance conduits depends upon the water available and design irrigation efficiencies, which in turn depend on soil, topography, method of irrigation, crop, type of irrigation system, and farm management practices. If water is delivered to a field or farm at a fixed rate throughout the season, the conveyance capacity can be less than if water is delivered periodically in

370

Chapter 11 Delivery and Distribution Systems

larger streams. Larger available stream sizes often allow for more efficient use of water and labor. Irrigation with borders or basins and with high intake rate soils requires larger streams than furrow irrigation and low intake rate soils. When designing conduit size, keep in mind that even with good maintenance, sediment, trash deposits, or vegetative growth may reduce a ditch’s capacity. Pipeline roughness may be increased and cross-sectional area reduced by encrustations or deposits. Conduits normally are designed for a capacity equal to the crop water requirements during peak demand plus irrigation and operational losses. Water losses vary with the irrigation distribution system, method of irrigation, soil, crops grown, farm management practices, and other factors, which are discussed in other chapters. For wellmanaged, conventional furrow and border irrigation systems, conduit design should allow for water application efficiencies of 60% or more. Level basins and surface systems with tailwater recovery are capable of 80% to 90% efficiencies. For reasons of economy, conduits should not be oversized. However, they should be large enough to provide adequate water for crop requirements under existing and anticipated irrigation methods. The amount of surface water intercepted from precipitation and irrigation waste must also be considered in determining capacity. 11.3.2 Lined Ditches and Structures Lining ditches is an effective way to prevent ditch erosion, control rodent damage, and reduce seepage at reasonable cost. Lining also reduces maintenance requirements, controls weed growth, and ensures more dependable water deliveries. The seepage reduction helps protect neighboring land from waterlogging and salt accumulations. Linings must be properly designed and installed to avoid damage from vegetative growth, fluctuating water tables, livestock traffic, and freezing and thawing. 11.3.2.1 Ditch capacity and design. Principles of open channel flow hydraulics especially applicable to farm irrigation ditches are presented by Replogle et al. (1990: Section 10.1.2). After the necessary capacity has been determined, ditch size can be designed using the Manning equation: C ν = u R 2 / 3S1 / 2 (11.4) n where V = mean velocity, m/s or ft/s n = Manning roughness coefficient (see Table 11.3) R = the hydraulic radius, m or ft S = the energy-line slope Cu = 1.0 for SI units and 1.486 for foot-pound (customary) units. Efficient methods to design ditches to convey a given flow rate can be found in standard hydraulics texts (Chow, 1959: Chapter 6; Schwab et al., 1993: Chapter 13). Table 11.3. Appropriate n values for use in Equation 11.4 for common farm ditch linings. Material Manning n for Capacity Slip-formed concrete 0.015 Brick 0.017 Shotcrete, concrete panels 0.016 Sheet metal, flex membrane 0.013 Compacted earth Same as unlined ditches; see Chow (1959)

Design and Operation of Farm Irrigation Systems

371

11.3.2.2 Lining materials. Selection of the lining material should be governed by availability of the material and installation equipment, ditch size, climate, foundation conditions, and type of irrigation stream, whether continuous or intermittent. Concrete is probably the most popular lining material, but asphaltic materials, bricks, membranes, metals, chemical sealants, and impermeable earth materials are also used. All of these materials make good linings if they are properly selected for the site conditions and are installed correctly. Lined ditches can be installed in any soil, but many soils have limitations that must be overcome before a durable, low-maintenance lining can be installed. Rigid linings, such as concrete, should be installed on drained soils, particularly in colder climates. For soils that have poor internal drainage or low bearing capacity, the concrete must be thicker or steel reinforcement must be added. 11.3.2.3 Elevation and slope. Lined ditches should be designed so the water surface at field turnout points is high enough to provide the required flow onto the field surface. If ditch checks or other control structures are used to provide the necessary head, the backwater effect must be considered in computing freeboard requirements. The required elevation of the water surface above the field surface will vary with the type of turnout structure used and the amount of water to be delivered. A minimum head of 0.12 m above the field water surface should be provided. The required freeboard varies with the size of ditch, the velocity of the water, the horizontal and vertical alignment, the amount of storm or wastewater that may be intercepted, and the change in the water surface elevation that may occur when control structures are operating. The minimum freeboard for any lined ditch should be 0.075 m. For concrete lining, design velocities in excess of 1.7 times the critical velocity should be restricted to straight reaches that discharge into a section or structure designed to reduce the velocity to less than critical velocity. This will avoid unstable surge flows. The maximum velocity in these straight reaches should be 4.5 m/s. Maximum allowable velocities in ditches with sprayed-on membrane linings should be limited to 1 m/s. Buried membrane installations are only effective where the flow velocity is limited to that allowable for the covering earth material. 11.3.2.4 Concrete lining and construction. Concrete linings made from Portland cement can withstand high stream velocities and therefore are particularly suitable for controlling erosion as well as preventing seepage. They are more resistant to mechanical damage than most other linings. However, their use is best suited to sites having non-expanding soils with low sulphate concentrations and good internal drainage. Local consultants or NRCS offices may be able to provide information on concrete mixes and lining thickness. Special mixes may be required where soils are high in sulphate or climatic conditions are severe. An irrigator who wishes to excavate and line his own ditches can use concrete panel-formed lining. This lining requires only semiskilled labor and a minimum of equipment, and, for small jobs, it may cost less than slip forming. After excavating the ditch to grade, guide forms are set about 3 m apart. Concrete is then poured in alternate 3 m sections. When the concrete is set, the forms are removed and the skipped panels are poured. The bottom in each section is poured first, and then additional concrete is screeded up both sides. 11.3.2.5 Other lining materials. Other linings that may prevent erosion and reduce seepage in farm ditches include masonry, pneumatically applied concrete, prefabri-

372

Chapter 11 Delivery and Distribution Systems

cated sheet metal, various plastic and asphaltic compositions, compacted earth, clay and buried membranes. Many of these are not available or are not practical in all irrigated areas. 11.3.2.6 Structures. Canal, lateral, and reservoir turnout structures that deliver water to farm ditches are beyond the scope of this monograph. A wide variety of such structures are described by Kraatz and Mahajan (1975). These are permanent structures, considered part of the canal delivery or supply system, and should include provisions for controlling and measuring flow rates. Other structures for irrigation delivery and distribution are considered below. Control structures. One of the most important factors in the efficiency and ease of operation of irrigation systems is the character of the control structures used. Surface irrigation characteristically has a high labor requirement, but good control structures can reduce labor and simplify irrigation. Use of mechanized structures is increasing. Combining two or more functions in a single structure can often minimize cost. For example, checks, drops, turnouts, divisors, and measuring structures can be used in various combinations. Flow measuring structures suitable for use in farm ditches are described in Section 11.2.2 of this chapter. If straight, suitably lined ditch sections are used to transmit water at supercritical velocity, this velocity must be reduced before approaching bends, distribution reaches, or through turnout structures into erosive channels. Energy-dissipating structures for such channels must be individually designed. The general configuration of a suitable structure and results of laboratory tests for discharges of 0.05 to 0.60 m3/s (1.6 to 20 ft3/s) have been published by the Bureau of Reclamation (USBR, 1963). Sediment deposited in irrigation ditches and structures necessitates frequent ditch cleaning and often results in inaccurate flow measurement. Trash in irrigation water is a source of weed infestation on the farm and clogs irrigation structures. When trash is a problem, special structures are often needed to remove it from the irrigation water supply. Drop structures. Drop structures are not needed for velocity control in lined ditches unless the estimated velocities will be supercritical. If ditch water needs to be carried across a road, a drop can be combined with a culvert, as shown in Figure 11.5. Concrete pipe (or PVC, if placed in a suitable trench, carefully bedded, and covered with 0.75 to 1.25 m of well-compacted fill) might be used in place of the corrugated metal pipe shown. More care with the bedding and backfill is needed to prevent crushing of the PVC, or joint leakage with the concrete pipe, than with the corrugated metal pipe. This drop is rather easily plugged by trash. Check structures. A check is any structure used to maintain or increase the water level above the normal flow depth in an open channel. The structures in Figure 11.6 can be fitted with flash board slots and can be used as combination checks/drops. Other designs are given by Skogerboe et al. (1971), Booher (1974), Kraatz and Mahajan (1975), and USDA.1 Commercial prefabricated and modular metal, concrete, and masonry structures are commonly used. 1

USDA Natural Resources Conservation Service (NRCS) Conservation Standards can be located at http://www.nrcs.usda.gov/technical/standars/nhcp.html. NRCS warns that these standards should not be used to plan, design, or install a conservation practice. A user must have the conservation practice standard developed by the state in which the practice is to be used to insure that all state and local criteria, which may be more restrictive than national criteria, are met.

:1 1.5

CMP

373

1:1

Design and Operation of Farm Irrigation Systems

D CMP

PLAN

ISOMETRIC VIEW OF CONCRETE SLAB

Top of Ditch Bank

Top of Ditch Bank

Concrete Slab Ditch Bottom

D

Corrugatied Metal Pipe

D Level Line

Watertight Welded Joint

Ditch Bottom

SECTIONAL ELEVATION ON CENTER LINE

NOTE: Capacities range from 65 L/s for 254 mm pipe and 0.3m drop, to 229 L/s for 381 mm pipe and 0.9 m drop. Figure 11.5. Corrugated metal pipe drop (USDA-SCS, 1978).

W

A

Crest H D+C

Stilling Basin L C

Capacity of Ditch Ft3/s L/s 57 170 227 283 396

Width of Opening, W

H

C

A

m

m

m

m

0.30 0.60 0.75 0.90 1.05

0.30 0.30 0.38 0.45 0.45

0.15 0.15 0.15 0.20 0.20

0.60 0.60 0.60 0.75 0.90

2 6 8 10 14

Drop, D

Apron Length, L

m

m

0.30 0.45 0.60 0.90

0.75 0.90 1.25 1.80

Suggested Drop Structure Figure 11.6. Common types of drop structures used in farm irrigation ditches.

374

Chapter 11 Delivery and Distribution Systems

Long-crested weirs. To maintain a constant upstream water level for uniform diversion of discharge into another channel or onto a field, an overflow weir-type check should be used. The flow over such a check may be estimated from the general weir equation, Equation 11.2, which in the general form is: n

Q = CLy

(11.5)

This equation is satisfactory for design purposes but not for exact water measurement through the opening. The value of n for most overflow checks is approximately 1.5, and C varies from 1.55 to 1.82 for SI units and 2.8 to 3.3 for L and y in feet (values of 1.66 m or 3.0 ft are frequently used for design purposes). When the crest length, L, is large, variations in discharge result in relatively small changes in the upstream water level. For a channel of given width, the fluctuation of upstream water surface as flow rates vary may be greater than desirable, even with a straight weir crest perpendicular to the channel center line. The water level fluctuation can be reduced even further by lengthening the weir crest, either with a diagonal weir (a straight weir crest constructed diagonally to the channel center line) or a duckbill weir (Figure 11.7). The effective length of a duckbill weir can be several times the channel width. Kraatz and Mahajan (1975) and Walker (1987) illustrate several types of diagonal and duckbill weirs, and give design information. In about 1990, engineers at California Polytechnic State University started modifications to early designs of canal flap gates, which dated from the 1940s in The Netherlands (ITRC, 2001; Burt et al., 2001). Raemy and Hager (1998) published a version that they claim is stable and uses a shock absorber to reduce oscillations (Figure 11.8). The IERC design also currently uses a shock absorber to reduce oscillations. The flap gate is a simple, inexpensive hydraulic gate for automatic upstream water level control. It is basically a rectangular plate placed at the end of a section of canal in which a constant upstream depth is desired. In its simplest form this is at an overfall that allows no backwater against the downstream side of the plate. The plate pivots about a hinge at its top that is offset upstream from the plane of the gate. Thus, with properly designed offset length, the weight of the gate and additional weights in the plane of the gate are able to maintain a constant upstream depth, passing excess flows for uses downstream. Gates for automatically controlling the water level in canals can also be used in some farm ditches (Kraatz and Mahajan, 1975). Semiautomatic timer-controlled checks have been used to a limited extent in farm ditches. They are normally used with timer-controlled outlets for releasing flow onto a field or into another ditch (Calder and Weston, 1966; Humpherys, 1969, 1986; Hart and Borrelli, 1970; Robinson, 1972; Evans, 1977). Automation of farm water control structures can reduce the irrigation labor requirement and improve the efficiency of surface irrigation.

Figure 11.7. Plan view of diagonal weir and duckbill weir.

Design and Operation of Farm Irrigation Systems

375

Figure 11.8. Side view of counterweighted flap gate, from Raemy and Hager (1998).

Figure 11.9. Precast concrete farm turnout (from USBR, 1951).

Outlets and discharge controls. Field turnouts are used to control the release of water from a farm ditch into basins, borders, furrows, or another irrigation ditch. The turnout may be a fixed opening in the side of a ditch or one equipped with check boards, gates, or other devices to adjust the opening area. If only a portion of the total flow is to be delivered through a given turnout, discharge through an orifice-type de-

376

Chapter 11 Delivery and Distribution Systems

vice is more constant than through an overflow or weir-type structure. One of the most frequently used turnouts is a concrete, plastic, or metal pipe with a slide gate on the inlet, as shown in Figure 11.9. For lined ditches the headwall and slide gate are usually constructed flush against the side of the ditch. The capacity of short pipe turnouts can be determined from Equation 11.1. For free flow, the head differential is the upstream head measured from the center of the orifice opening; for submerged flow, it is the difference between upstream and downstream water levels. The coefficient of discharge Cd varies widely, depending on the position of the orifice in relation to the sides and bottom of the structure and the type of orifice opening, i.e., sharp edged, rounded, short tube, etc. (Brater and King, 1976). Humpherys (1978) developed timer-controlled gates for border turnouts where the turnout must first be automatically opened to begin irrigation and then closed to terminate irrigation while the water level in the head ditch remains constant. It is often necessary to divide water from a farm lateral into two or more ditches for distribution to different parts of the farm or to other farms. This may be accomplished with a divisor at the ditch junction. For accurate flow division it is best to measure the flow in each channel. Some divisors are designed to give a fixed proportional flow division, whereas others have a movable splitter to change the flow proportions. Divisors that give accurate proportions often divide the flow at a control section where supercritical flow exists, such as at the nappe of a free overfall (Figure 11.10). Flow can be divided accurately without creating supercritical flow if: (a) the approach channel is long and straight for at least 5 to 10 m upstream so that the water flow approaches the divisor in parallel paths without crosscurrents, (b) there is no backwater effect that would favor one side or the other, and (c) the flow section of the structure is of uniform roughness. Detailed divisor designs are described by Bos (1989) and Kraatz and Mahajan (1975). Debris and weed seed control structures. Trash and weed seed in the water cause problems in irrigation systems. Besides spreading weeds, trash and debris plug siphon tubes, pipe gates, and sprinkler nozzles. Trash racks are needed at entrances to pipe

Figure 11.10. Simple fixed proportional flow divisor.

Design and Operation of Farm Irrigation Systems

377

Inlet Box Screen Frame Frame Box Outlet

Figure 11.11. Structure with screen for trapping debris in irrigation streams (adapted from Bergstrom, 1961).

and underground crossings. Various types of screens and trash racks have been developed for screening irrigation water (Couthard et al., 1956; Bergstrom, 1961; Pugh and Evans, 1966). One of the simplest devices is a screen placed below a weir or drop, as in Figure 11.11. If the screen is taut and the water drops from a sufficient height, say at least 0.2 m, the vibrations of the screen will move the trash far enough away from the overfall to allow the water to pass through. Screen mesh size will depend on the type of debris in the irrigation stream and the susceptibility to plugging of downstream water control devices. Very fine screens may require more frequent cleaning, but may also have advantages, such as removing weed seed. A different configuration of a screen using much the same principle has been developed in Idaho. The “turbulent fountain” screen is circular, surrounding the opening of a vertical pipe with flow discharging from the pipe’s open top. The screened water is collected below the fountain screen, while the turbulence of the water pushes trash off its sides (Kemper and Bondurant, 1982; Bondurant and Kemper, 1985). Screens that use water wheels or electric motors to power brushes that pass repeatedly over the screen surface, moving trash to one side, are commercially available. 11.3.3 Low-Pressure Pipe Systems 11.3.3.1 Applications. Low-pressure pipe systems for irrigation water distribution have been used extensively since the 1950s. The availability of relatively low-cost, lightweight, rigid plastic pipe has made buried pipe systems especially popular. Advantages and disadvantage of buried pipelines are summarized in Section 11.3.1. Pipelines with large diameter risers may require safety precautions, either adequate height above the ground or a screen or cover to keep children from entering. There are three general types of on-farm irrigation pipelines. The first is the completely portable surface system where water enters the line at the supply (a well, reservoir, or ditch turnout) and the water is applied to the field from the open end of the pipeline or from gated outlets along the line. The second system is a combination of

378

Chapter 11 Delivery and Distribution Systems

buried and surface pipe where buried permanent line is used to transmit water from the source to the field to be irrigated. Then water is supplied to gated surface pipe through one or more risers. The third system, generally used for border dike or basin irrigation, eliminates the need for surface pipe. Water is released onto the portion of the field to be irrigated from risers on the buried line (see also Section 11.2.3). 11.3.3.2 Location. The irrigator can exercise much flexibility in locating surface pipelines between the source of water supply and the point of water application. The lines should be located to provide a minimum of interference with traffic on farm roads or with field cultural operations. A pipe network should be designed so that the shortest possible length of pipe is used to deliver water to all fields to be irrigated. More care is needed in locating buried systems. If possible, the line should be located where it can be easily buried with uniform trench depth and the minimum necessary cover. It is best to avoid routes that will be crossed by heavy surface traffic. Also, because air relief must be provided at high points in a buried pipeline, the number of these high points should be minimized. Trenching across fields to be irrigated should be avoided if possible. Irrigation water may enter poorly compacted backfill in the trench and cause piping along the line, simultaneously causing incomplete irrigation of the field being watered. 11.3.3.3 Pipeline capacity. Pipeline capacity must be great enough so that adequate water can be delivered to meet crop requirements. It generally will cost less to use large pipe with the available head than to use smaller pipe and booster pumps for such systems. When an entire irrigation system is to be designed, including pump, power unit, and piping, irrigation pipe sizes should be chosen so as to minimize total annual costs. Both fixed and operating items must be considered, including the initial cost of the piping, the number of years’ service expected from the pipe, the initial cost and expected years of operation of different sized pump units, the energy cost of pumping, and the rate of return desired on the funds invested. Keller (1975) has presented a procedure for constructing design charts from which economical pipe sizes can be selected, based on the above considerations for non-looping distribution systems having a single pump station. More complicated systems can be designed using procedures given by authors cited by Keller. The Darcy-Weisbach Formula (Brater and King, 1976) expresses head loss of turbulent flow in pipelines as: L V2 (11.6) H1 = f D 2g where Hl = the loss of head in equivalent height of water in a length of pipe L D = inside pipe diameter V = mean velocity g = gravitational acceleration f = a resistance coefficient. Equation 11.6 is dimensionally consistent and can be used with the same f values for either SI or foot-pound units. Values of f have been related to boundary roughness dimensions for certain types of pipe surfaces and determined empirically for others and are tabulated in most hydraulic handbooks. The Hazen-Williams equation (Brater and King, 1976) is used extensively for determining friction losses in irrigation pipelines. Hazen-Williams resistance coefficients are readily available, having been deter-

Design and Operation of Farm Irrigation Systems

379

mined empirically for tubing commonly used in irrigation. The equation can be written: V = CR 0.63 S 0.54

(11.7)

where R = the hydraulic radius and S = the friction slope in meters per meter or feet per foot. When R is in mm, C = 0.0109 C1; when R is in feet, C = 1.318 C1; where C1 is the Hazen-Williams resistance coefficient. Values of C1 range from 144 to 146 for aluminum tubing. C1 values for other materials are given in standard hydraulic handbooks. Energy losses through fittings and valves also have to be considered in the design of an irrigation pipeline. These so-called “local” losses are frequently estimated by applying a coefficient to the velocity head at the fitting. The sum of all local losses is then added to the estimate of pipe wall friction to give total loss in the pipeline. Local loss coefficients, K, to be used in the equation: Hf = KV 2/2g

(11.8)

are listed in Table 11.4. The coefficients in Table 11.4 have been determined for standard pipe diameters. Coefficients for special irrigation pipe dimensions and materials are not available. However, because local losses occur as a result of flow separation and changes in velocity, and are not greatly affected by the roughness of the fitting, values in Table 11.4 should be valid for irrigation pipe fittings with similar geometry. Table 11.4. Value for K for use in Equation 11.8 to estimate local head losses in pipelines. Nominal Diameter 75 mm 100 mm 125 mm 150 mm 175mm 200mm 250 mm (3 inch) (4 inch) (5 inch) (6 inch) (7 inch) (8 inch) (10 inch) Fitting or Valve Standard pipe Elbows Regular flanged 90° 0.34 0.31 0.30 0.28 0.27 0.26 0.25 Long radius flanged 90° 0.25 0.22 0.20 0.18 0.17 0.15 0.14 Regular screwed 90° 0.80 0.70 Tees Flanged line flow 0.16 0.14 0.13 0.12 0.11 0.10 0.09 Flanged branch flow 0.73 0.68 0.65 0.60 0.58 0.56 0.52 Screwed line flow 0.90 0.90 Screwed branch flow 1.20 1.10 Valves Globe flanged 7.0 6.3 6.0 5.8 5.7 5.6 5.5 Gate flanged 0.21 0.16 0.13 0.11 0.09 0.075 0.06 Swing check flanged 2.0 2.0 2.0 2.0 2.0 2.0 2.0 Foot 0.80 0.80 0.80 0.80 0.80 0.80 0.80 Strainers, basket type 1.25 1.05 0.95 0.85 0.80 0.75 0.67 Other Inlets or entrances Inward projecting 0.78 (all diameters) Sharp cornered 0.50 (all diameters) Slightly rounded 0.23 (all diameters) Bell mouth 0.04 (all diameters) 2

[

]

Sudden enlargements

K = 1 − (d12 / d s 2 ) , where d1 = diameter of the smaller pipe.

Sudden contractions

K = 0.7 1 − (d12 / d2 2 ) , where d1 = diameter of the smaller pipe.

[

2

]

380

Chapter 11 Delivery and Distribution Systems

11.3.3.4 Surface pipelines. Portable surface pipelines for low-pressure irrigation water distribution are commonly constructed of aluminum or rigid or flexible plastics. All can be used solely for transmission of water or can be provided with gated outlets for water distribution to fields. Low-pressure aluminum pipe is generally available in 127-, 152-, 203- and 254-mm (5-, 6-, 8- and 10-in.) nominal diameters. The pipe is rolled from 1.3-mm (0.051-in.) aluminum sheet and closed with a welded seam. A rolled or cast female coupling is provided on each pipe length to contain flexible, pressure sealing gaskets. Mechanical clamps are sometimes provided for positive pipe coupling. Aluminum gated pipe sections of large diameters are generally manufactured in 6m lengths, whereas those of smaller diameters are 9 m long. Standards for aluminum and PVC tubing for gated pipe are published by USDA (2006). Rigid, extruded PVC surface pipelines are also available. The PVC compound contains an ultraviolet inhibitor that allows extended exposure to sunlight without deterioration. Diameters and couplings are compatible with aluminum-gated pipe. Conduits of thin-walled, flexible vinyl tubing for water transmission and distribution to furrows became available in the mid-1970s. This tubing is sold in lengths of several hundred meters, packaged in compact rolls. Wall thickness is 0.25 mm or less and diameters from 250 mm to 500 mm are available. A tubing installation implement, which mounts on the three-point hitch of a farm tractor, can plow a shallow groove to provide a bed for the tubing and unreel and lay the tubing simultaneously. Plastic fittings are available for couplings, tees, and connections to the water supply. Gates are inserted in the field, at the exact locations necessary to best serve furrows or corrugations, using a simple insertion tool. Flow capacity of the flexible vinyl tubing is about one-fourth less than aluminum tubing of the same diameter, because at the low allowable operating pressures the tubing is not fully round. When gated tubing is laid on sloping land, and the water source is from the upslope end, the upslope end may be at such low pressure that adequate streams will not flow from the gates. As a practical solution, some irrigators shovel soil on the tubing at intervals in order to restrict through-flow and maintain adequate pressure in the upper reaches. Initial cost of flexible vinyl tubing is a fraction of that for comparable aluminum or rigid plastic pipe. However, the material is not durable and usually has to be replaced for each irrigation season. Some portable pipelines, especially aluminum lines that have low resistance to crushing, can be flattened by the pressure differential created if sub-atmospheric pressure occurs in the line. Sub-atmospheric pressures might occur if a deep well pump that supplies a pipeline without check valves is suddenly stopped and the open valves on the pipeline are inadequate for vacuum relief. Rapid closing of an upstream valve in a line carrying high velocity flow may also lower pressures below the valve. Appropriately placed vacuum relief valves and careful operating procedures can minimize the chance of such damage occurring. Also, pressure surges in the line can cause joints to separate or pipe sections to rupture. Valves controlling flow into surface pipelines should be opened carefully so that the lines fill slowly before they are brought up to full pressure. If water is to be released from the pipeline through gates, one set of gates should be open at the time the line is filled. Likewise, shutting off the flow to the pipe before all gates are closed will help prevent a vacuum from developing in the line. If the pipeline is supplied directly

Design and Operation of Farm Irrigation Systems

381

Table 11.5. Friction loss in aluminum pipe as estimated by the Hazen-William formula.[a] Joint losses are included[b]; assumes joints at 9-m (30-ft)[c] intervals (adapted from USDA-SCS, 1978). Nominal Outside Diameter 102 mm 127 mm 152 mm 203 mm 254 mm (4 in.) (5 in.) (6 in.) (8 in.) (10 in.) NG G NG G NG G NG G NG Discharge G[d] (L/s) (ft3/s) Head Loss (m/100 m or ft/100 ft) 10 0.35 1.89 1.80 0.65 0.62 0.27 0.26 0.07 0.07 15 0.53 4.07 3.86 1.40 1.31 0.59 0.55 0.15 0.14 20 0.71 6.94 6.57 2.41 2.26 1.01 0.94 0.26 0.24 0.09 0.09 30 1.06 14.86 14.02 5.16 4.82 2.18 2.01 0.57 0.52 0.20 0.18 40 1.41 8.87 8.26 3.74 3.45 0.98 0.89 0.35 0.31 50 1.77 5.71 5.25 1.50 1.35 0.54 0.48 75 2.65 12.24 11.23 3.24 2.92 1.16 1.03 100 3.53 5.57 5.00 2.01 1.77 125 8.54 7.65 3.08 2.71 150 12.02 10.74 4.35 3.82 [a] [b] [c]

[d]

Hazen-Williams coefficients: C1 = 145 for 102- to 152-mm pipe; C1 = 144 for 203- to 254-mm pipe. Local loss coefficient, Kc = 0.3 for gated pipe; 0.2 for pipe without gates. For 6-m (20-ft) lengths, increase loss by 10% for gated pipe and by 7% for pipe without gates; for 12-m (40-ft) lengths, decrease loss by 5% for gated pipe and by 4% for pipe without gates. G = gated, NG = not gated.

from a pump, a standpipe near the pump outlet, with diameter at least half the pipe diameter, serves to even out pressure surges and to relieve vacuum as well. A check valve near the pump discharge prevents flow of water from the line back through the pump if the pump stops. Low-pressure portable pipelines are generally not installed over terrain with wide variations in elevations. However, if extreme high points exist in the line, some method of air release at these points may be needed to avoid restriction in pipe capacity. Values of friction loss for aluminum pipe with and without outlet gates are given in Table 11.5 for pipe sizes commonly available in the U.S. The head loss values in these tables include a component to account for local losses at pipe joints. Water is commonly released from surface pipelines through small gates installed in the pipeline at intervals equal to the furrow or corrugation spacing. Head loss in a section of pipe discharging water through open gates will be one-third to one-half of what it would be if the pipe were transmitting the full flow from one end of the discharging section to the other. Pipe gate discharge. Pipe gates are supplied in many sizes and styles by a number of manufacturers. Gate discharge is governed by the orifice flow equation (Equation 11.1) when there is no flow in the pipeline past the gate. When there is flow, gate discharge decreases as velocity increases. Because the discharge coefficient varies significantly for different types of gates, discharge information should be obtained directly from the manufacturer of the gates used. When gated pipe is laid on too great a slope, head is greatest on the downslope gates. These gates must be nearly closed to maintain constant furrow streams along the pipeline. With the small gate openings, velocities are large and may cause erosion damage where the streams impinge on the furrow surface. Flexible tubes or “socks” attached to each gate still the high-velocity streams and minimize erosion. Simple but-

382

Chapter 11 Delivery and Distribution Systems

terfly valves or overflow stands (similar in principle to that shown in Figure 11.12b) may be installed at intervals along the pipeline to maintain more nearly constant heads. Orifice plates can also be placed in the bell ends of the pipe to dissipate excess energy. The size and number of orifices used depends on the head to be dissipated. 11.3.3.5 Buried low-pressure systems. Non-reinforced concrete was the most economical and commonly used material for buried low-pressure irrigation pipelines before thermoplastic piping came into use. Concrete pipe sections are readily available in diameters from 150 to 600 mm. Most types of concrete pipe are designed to be joined with rubber gaskets; others are joined using a cement mortar. Either joining method is more labor intensive than for plastic pipe. Nitrogen fertilizers should be used with care in concrete pipelines. When ammonia is added to hard waters, calcium carbonate may precipitate and adhere to the walls of the pipe. Ammonium sulphate concentrations should not exceed 0.1% and the line should be flushed immediately after use. Fertilizer manufacturers should be consulted on damage prevention measures. In recent decades, the ready availability and ease of handling of lightweight, semirigid thermoplastic piping has made buried low-pressure pipe systems an attractive alternative to concrete pipelines for irrigation water transmission. The plastic compounds commonly used in constructing such pipe are polyvinyl chloride (PVC), acrylonitrile butadiene styrene (ABS), and polyethylene (PE). PVC pipe is manufactured in two general size classifications: iron pipe size (IPS) and plastic irrigation pipe size (PIP). IPS pipe has the same outside diameter as iron or steel pipe of the same nominal size. PIP sizes were developed primarily for irrigation use. PIP is smaller in actual diameter than IPS pipe of the same nominal diameter and therefore has a lower flow capacity. Difference in capacity ranges from about 18% for 100-mm pipe to 8% for 300-mm. Plastic pipe is commonly available in a range of pressure ratings. The ratio of the allowable tensile stress in the pipe wall to the maximum pressure that the fluid in the pipe can exert continuously with a high degree of certainty that the pipe will not fail is related linearly to the standard dimension ratio, SDR (the ratio of pipe diameter to minimum wall thickness). Plastic pipe commonly used for irrigation is manufactured with several SDRs. Schedule 40, 80, or 120 PVC pipe, with dimensions corresponding to steel pipe, is also available. For a given pipe material and standard temperature, pipes of all diameters with the same SDR will have the same pressure rating. For PVC and ABS the average outside diameter is used in calculating SDR. For PE the average inside diameter is used. Polyvinyl chloride pipe for low-head systems is usually labeled “low head,” “SDR = 81,” or “SDR = 64.” Operating temperature also affects pressure ratings. Pressure ratings for pipes of different materials and a range of SDRs, and rating service factors for different temperatures, are given in ASAE Standard S376 (ASAE, 2005b ). Head losses in low-head plastic pipe are determined using the Hazen-Williams equation (Equation 11.7). A resistance coefficient of C1 = 150 is commonly used. Local losses are calculated using the same K factors as listed in Table 11.4. Trenches for plastic pipe should have a relatively smooth, firm, continuous bottom and be free of rocks. Where rough rock edges cannot otherwise be avoided, the trench should be over-excavated and filled to the bedding depth with sand or finely graded soils. The trench below the top of the pipe should be just wide enough to provide room for joining the pipe and compacting the initial backfill. Minimum and maximum trench widths

Design and Operation of Farm Irrigation Systems

383

for low-head and SDR-81 pipe are listed in ASAE Standard S376 (ASAE, 2005b). The pipe trench should be excavated deeply enough so that from 0.75 to 1.20 m of cover can be placed over low-head irrigation pipes. This cover will normally protect the pipe from traffic crossing, freezing temperatures, or soil cracking, but is not so great that it will cause excessive soil loading on the pipe. Low areas can be crossed with shallow trench depths and then extra fill placed over the pipeline to provide the minimum cover depth. In such cases, the fill should have a top width of no less than 3 m and 6:1 side slopes. The pipe should be uniformly supported over its entire length on firm stable material in the trench. Solvent-welded PVC pipe is often flexible enough to be assembled alongside the trench and then lowered into the trench after the solvent welds are well set. PVC pipe with SDRs of 81.0 and lower can be obtained with solvent-weld or gasketed joints. Inexperienced workers can assemble leak-free pipelines more easily with gasketed than with solvent-weld joints. Wherever the flow direction changes (at elbows and tees), reaction forces on the piping may exist. These forces, if unrestrained, may cause the joints to separate. Concrete thrust blocking is commonly used to provide resistance to pipe movement where side thrust might occur. Concrete pads, poured around pipe fittings in the trench, allow forces on the pipe to be resisted by a large area of soil along the trench surface. ASAE Standard S376 (ASAE, 2005b) illustrates those fittings that require thrust blocks, approved configuration of the blocks, and gives tables for estimating thrust magnitude. Water packing can be used to settle the backfill around low-head pipelines. Before backfilling, the pipe is filled completely with water and all joints observed for leakage. The pipe should remain filled with water during the entire backfill operation to keep it from floating or deforming under the weight of the saturated fill. After the pipe is proven to be leak free, 0.3 to 0.45 m of backfill are placed over the pipe and water is added to the trench until the fill is thoroughly saturated. The wetted fill is then allowed to dry until firm enough to walk on before the final backfill is placed. Polyethylene pipes are joined using rigid insert fittings and clamps or bell joints with rubber gaskets. Polyethylene up to 150 mm in diameter is available for buried lines. For details of plastic pipe pressure rating, material characteristics, and installation specifications and procedures, see ASAE Standard S376 (ASAE, 2005b). Protection from vacuum and pressure surges is essential for buried thermoplastic pipelines. See Pressure Relief Valves in Section 11.3.3.6 below for details. 11.3.3.6 Structures. All buried pipelines require an inlet and one or more outlet structures. In addition, structures for air release, relief of excess pressure, or head control may be necessary (Robinson et al., 1963; Seipt, 1974). Frequently two or more functions can be combined in a single structure (Figure 11.12). Construction methods and materials are similar for any given structure, regardless of pipeline material. Reinforced concrete pipe is often used for inlet and head control risers because of its durability and resistance to impact damage. Special openings in pipe sections, beveled ends, etc., are most easily made while the concrete is green and can be specially ordered from pipe casting plants. Inlets. Where water is supplied to a buried pipeline from an open channel, a short, vertical section of pipe can be used as the basis of the inlet structure. If all downstream outlets on the pipeline can pass small pieces of trash without clogging, trash bars spaced on 8-cm centers can be used to keep large trash and animals out of the pipeline.

384

Chapter 11 Delivery and Distribution Systems

Figure 11.12. Typical standpipes used for head control, pressure relief, and water release from low-pressure pipelines. Several variations on these designs are possible for specific sites.

For smaller outlets, especially on gated pipe, and a water supply containing trash, a finer inlet screen is necessary. Finer screens usually need to be self-cleaning, or they will clog and cause the irrigation stream to overflow the inlet structure. Even selfcleaning screens require frequent, perhaps daily, visits to remove trash accumulation and check mechanical functions. A flow measuring device is usually needed near the pipeline inlet, especially where the pipeline is supplied directly from a well or an irrigation district canal. Metering allows the irrigator to determine the volume of water applied and is essential if irrigation scheduling recommendations are to be followed. Changes in well capacity and pump efficiency can also be detected if metered flow rates are compared periodically. If the pipeline is supplied from an open channel, a critical depth flume upstream of the pipe inlet may be used for flow measurement. See Section 11.2.2 for details of flow measurement with meters and flumes. Pump stands. When a pipeline receives water directly from a pump, a stand similar to that shown in Figure 11.13 is necessary. Pump stands can be constructed from reinforced concrete or steel. The inside diameter (D3) should be great enough to limit downward velocity to 0.6 m/s. The vent pipe, if of reduced size as shown, should be large enough to limit velocities to 3 m/s with the full pump outflow flowing through the vent. The connection between pump outlet and stand must prevent transmission of vibrations to the stand or pipeline.

Design and Operation of Farm Irrigation Systems

385

D1 D1 + D2 Minimum

D2 D3

Figure 11.13 Surge control/air release stand for a pump discharging into a buried pipeline.

Vents, gate stands, and turnout risers. The requirements for these various types of pipeline structures are given in ASAE S376.1 (ASAE, 2005b). Figure 11.12 shows two variations of structure that can be used for venting air, and/or controlling head in low-pressure buried pipelines. Many buried pipelines are designed to release water through risers capped with alfalfa valves, which can control releases into the atmosphere or into portable surface pipelines. Procedures and equipment have been developed for automatic timed water releases from alfalfa valves, using auxiliary pneumatic valves (sometimes in gated pipe hydrants) and low-pressure air supply. Such systems can increase irrigation efficiencies and reduce labor requirements on many fields, if they are carefully designed and if the field is otherwise prepared for efficient irrigation (Haise et al., 1965, 1978; Haise and Kruse, 1966; Fischbach and Gooding, 1970; Humpherys, 1986). Safety. All pipeline structures having openings within 2 m of the soil surface should be high enough, or they should be covered, so that small children cannot enter the pipeline. Covering also prevents small animals and wind-blown trash from entering the pipeline. Pressure relief valves. Open stands can be replaced by pressure relief valves so that the pipeline need not be open to the atmosphere. Pressure relief valves must be marked with the pressure at which they start to open. Adjustable valves should be capable of being locked or sealed so that the proper pressure setting is not easily changed. The valve should be capable of releasing the design flow of the pipeline without elevating pipeline pressure more than 50% above the permissible working head of the pipeline. The number and spacing of relief valves on a pipeline is determined by the grade and the allowable working head of the pipe. For pipelines of constant slope, one valve, at the lowest point, may be adequate. If examination of the hydraulic head line shows that allowable pressures could be exceeded, additional relief valves must be added. Some manufacturers supply pressure relief valves that also provide air and vacuum relief functions.

386

Chapter 11 Delivery and Distribution Systems

11.3.4 Operation and Maintenance 11.3.4.1 Lined ditches. If lined ditches are not properly maintained, the result can be damage to the ditch lining and structures, and erosion of the field surface, as well as substantial economic loss to growing crops. Material needed to perform maintenance should be stockpiled on the farm. The small investment required for the stockpiled materials will be more than offset by the capability to make timely repairs. The ditches should be inspected routinely and maintenance requirements assessed. The timing of the inspections depends somewhat on the type of ditch lining and the soil and climatic conditions at the site. However, as a minimum, ditches should be thoroughly inspected before every irrigation season. Where rigid linings are used in northern climates, the lined ditch should be inspected just before ground freezing to make sure that drainage is adequate. Cracks in concrete and masonry structures should be caulked to prevent water entry and minimize freeze damage. Masonry structures are particularly susceptible to cracking in cold climates. Metal structures should be painted and rubber-type seals replaced as needed. Metal gates should be lubricated and, where possible, left partially open during the winter. This prevents them from sticking or rusting closed. Ditches and structures should be drained when not in use. Water ponded in or near structures contributes to frost damage and structure deterioration. The lifespan and annual maintenance costs of irrigation ditch lining will vary depending on site conditions and the quality of the initial construction. In the absence of reliable local data, a lifespan of 15 yr and annual maintenance costs of 5% of the installation cost are recommended for nonreinforced concrete, flexible membrane, and galvanized steel ditch linings. 11.3.4.2 Pipelines. Pipelines should be inspected for leakage at least once a year. Leaks may be spotted from wet soil areas above the line that are otherwise unexplained. Small leaks in concrete pipelines can be repaired by carefully cleaning the pipe exterior surrounding the leak, then applying a patch of cement mortar grout. For larger leaks, one or more pipe sections may have to be replaced. Longevity of concrete pipelines can be increased by capping all openings during cold winter months to prevent air circulation. Small leaks in plastic pipe, except at the joints, can sometimes be repaired by pressing a gasket-like material tightly against the pipe wall around the leak and clamping it with a saddle. Leaky joints or damaged pipe sections must be replaced with a new pipe section. Special gasketed steel or plastic fittings must be used to fasten a section of replacement pipe into the line if the pipe cannot be displaced horizontally to join the original belled joints. Where water is supplied from a canal to portable surface pipe, sediment often accumulates in the pipe. This sediment should be flushed out before the pipe is moved. Otherwise, the pipe will be too heavy to be moved by hand and may be damaged if it is moved mechanically. Buried plastic pipelines can be expected to have a usable life of about 15 years if well maintained. The annual cost of maintenance can be estimated as approximately 1% of the installation cost.

Design and Operation of Farm Irrigation Systems

387

Acknowledgements Much of Section 11.3 of this chapter is based on Chapter 11, Farm water distribution systems, of Design and Operation of Farm Irrigation Systems (1980), to which A. S. Humpherys and E. J. Pope were major contributors.

REFERENCES Abt, S. R., and K. J. Staker. 1990. Rating correction for lateral settlement of Parshall flumes. J. Irrig. Drain. Eng. 116(6): 797-803. Abt, S. R., K. Thompson, and K. Staker. 1989. Discharge corrections for longitudinal settlement of Parshall flumes. Trans. ASAE 32(5): 1541-1544. Ackers, P., and A. J. M. Harrison. 1963. Critical-depth flumes for flow measurements in open channels. Hyd. Res. Paper No. 5. Wallingford, Berkshire, England: Dept. Sci. and Indust. Res., Hyd. Res. Sta. Ackers, P., W. R. White, J. A. Perkins, and A. J. M. Harrison. 1967. Weirs and Flumes for Flow Measurement. New York, N.Y.: John Wiley & Sons. 327 pp. ASAE. 2005a. EP267.6: Principles and practices for prevention of mosquito sources associated with irrigation. St. Joseph, Mich.: ASAE. ASAE. 2005b. S376.1: Design, installation and performance of underground, thermoplastic irrigation pipelines. St. Joseph, Mich.: ASAE. ASME. 1971. Fluid Meters: Their Theory and Applications. New York, N.Y.: American Soc. Civil Engineers. ASTM. 2005a. Standard specification for reinforced concrete culvert, storm drain, and sewer pipe. ASTM C76-05, Vol. 04.05. West Conshohocken, Pa.: American Society for Testing Materials. ASTM. 2005b. Standard specification for concrete pipe for irrigation and drainage. ASTM C188-05, Vol. 04.05. West Conshohocken, Pa.: American Society for Testing Materials.. Bergstrom, W. 1961. Weed seed screens for irrigation systems. Ext. Publ. Bull. 43. Moscow, Idaho: Univ. of Idaho, Pacific Northwest Coop. Agr. Ext. Serv. Bondurant, J. A., and W. D. Kemper. 1985 Self-cleaning, non-powered trash screens for small irrigation flows. Trans. ASAE 28(1): 113-117. Booher, L. J. 1974. Surface irrigation. FAO Agri. Dev. Paper No. 95. Rome, Italy: FAO. Bos, M. G., ed. 1989. Discharge Measurement Structures. Publ. No. 20. 3rd rev. ed. Wageningen, The Netherlands: Int’l. Inst. for Land Reclamation and Improvements (ILRI). Bos, M. G., J. A. Replogle, and A. J. Clemmens. 1991. Flow Measuring Flumes for Open Channel Systems. St. Joseph, Mich.: ASAE. Brater, E. F., and H. W. King. 1976. Handbook of Hydraulics. 6th ed. New York, N.Y.: McGraw-Hill Book Co. Burt, C. M. 1987. Overview of canal control concepts. In Planning, Operation, Rehabilitation and Automation of Irrigation Water Delivery Systems. Proc. Symp., ASCE, 81-108. D. D. Zimbelman, ed. New York, N.Y.: American Soc. Civil Engineers. Burt, C. M., R. Angold, M. Lehnkuhl, and S. Styles. 2001. Flap gate design for automatic upstream water level control. J. Irrig. Drain. Eng. 127(2): 84-91. Discussion/closure in 128(4): 264-265.

388

Chapter 11 Delivery and Distribution Systems

Calder, G. G., and L. H. Weston. 1966. Automatic system of farm irrigation. New Zealand J. Agr. 112(2). Chow, V. T. 1959. Open Channel Hydraulics. New York, N.Y.: McGraw-Hill Book Co. Clemmens, A. J. 1987. Delivery system schedules and required capacities. In Planning, Operation, Rehabilitation and Automation of Irrigation Water Delivery System. Proc. Symp., ASCE, 18-34. D. D. Zimbelman, ed. New York, N.Y.: American Soc. Civil Engineers. Clemmens, A. J. 1992. Feedback control of a basin irrigation system. J. Irrig. Drain. Eng. 118(3): 480-496. Clemmens, A. J., M. G. Bos, and J. A. Replogle. 1993a. FLUME Design and calibration of long-throated measuring flumes. Version 3.0, Pub. #54, WCL# 1751. Wageningen, The Netherlands: Int’l. Inst. for Land Reclamation and Improvement (ILRI). Clemmens, A. J., A. R. Dedrick, and R. J. Strand. 1993b. Basin 2.0 for the design of level-basin irrigation systems. In Proc. ASCE Nat. Conf. on Irrig. & Drain., 875882. New York, N.Y.: American Soc. Civil Engineers. Clemmens, A. J., T. L.Wahl, M. G. Bos, and J. A. Replogle. 2001. Water measurement with flumes and weirs. Publ. #58. Waginingen, The Netherlands: Int’l. Inst. for Land Reclamation and Improvement (ILRI). Clemmens, A. J., and J. A. Replogle. 1980. Constructing simple measuring flumes for irrigation canals. USDA Farmer’s Bull. No. 2268. Washington, D.C.: USDA. Coulthard, T. L., J. C. Wilcox, and H. O. Lacey. 1956. Screening irrigation water. Agr. Eng. Div. Bull. A.E. 6. Vancouver, BC, Canada: Univ. Brit. Columbia. Dedrick, A. R. 1984. Special design situations for level basins. In Proc. Twelfth Congress on Irrigation and Drainage, 1(B), Question Q39, R29(4): 465-474. New Delhi, India: ICID. Dedrick, A. R. 1986. Control requirements and field experience with mechanized level basins. Trans. ASAE 29(6): 1679-1684. Dedrick, A. R. 1989. Improvements in design and installation features for mechanized level basin systems. Appl. Eng. Agric. 5(3): 372-378. Dedrick, A. R. 1990. Level-basin irrigation: An update. In Visions of the Future, Proc. Third Nat’l. Irrigation Symp., 34-39. St. Joseph, Mich.: ASAE. Dedrick, A. R., and Y. Reinink. 1987. Water ponding on level basins caused by precipitation. Trans. ASAE 30(4): 1057-1064. Dedrick, A. R., J. A. Replogle, and L. J. Erie. 1978. On-farm level-basin irrigation: Save water and energy. Civil Eng. 48: 60-65. Dethridge, J. S. 1913. An Australian water meter for irrigation supplies. Eng. News (New York) 7(26). Evans, R. G. 1977. Improved semiautomatic gates for cut-back surface irrigation systems. Trans. ASAE 20(1): 105-108, 112. Fischbach, P. E., and R. Gooding II. 1970. An automated surface irrigation valve. Agric. Eng. 52(11): 584-585. Goussard, J. 1987. NEYRTEC automatic equipment for irrigation canals. In Planning, Operation, Rehabilitation and Automation of Irrigation Water Delivery Systems, Proc. Symp., ASCE, 121-132. D. D. Zimbelman, ed. New York, N.Y.: American Soc. Civil Engineers.

Design and Operation of Farm Irrigation Systems

389

Haise, H. R., and E. G. Kruse. 1966. Pneumatic valves for automation of irrigation systems. In Proc. 6th Congr. Int’l Comm. on Irrig. and Drain. Specialty Sessions. Report No. l: S-1 to S-8. New Delhi, India: ICID. Haise, H. R., E. G. Kruse, and N. A. Dimick. 1965. Pneumatic valves for automation of irrigation systems. USDA-ARS 41-104. Washington, D.C.: USDA. Haise, H. R., E. G. Kruse, M. L. Payne, and H. R. Duke. 1978. Automation of surface irrigation: Summary of 15 Years USDA research and development at Fort Collins, Colo. Prod. Res. Report #179. Washington, D.C.: USDA, SEA-AR. Hart, W. E., and J. Borrelli. 1970. Mechanized Surface Irrigation Systems for Rolling Lands. Contr. No. 133. Berkeley, Calif.: Univ. of Calif., Water Resources Center. Humpherys, A. S. 1969. Mechanical structures for farm irrigation. J. Irrig. Drain. Div., ASCE 95(IR4): 463-479. Humpherys, A. S. 1978. Improving farm irrigation systems by automation. Preprint, 10th Congr. Int’l Comm. on Irrig. and Drain. Rpt. 5, Q. 35: 35.90-35.98. New Delhi, India: ICID. Humpherys, A. S. 1986. Automated Farm Surface Irrigation Systems: Worldwide. New Delhi, India: Int’l Commission on Irrigation and Drainage. Humphreys, A. S., and J. A. Bondurant. 1977. Cast-in-place, 2-foot concrete trapezoidal flow-measuring flumes. Tech. Bull. 1566:43. Washington, D.C.: USDA. ITRC. 2001. Flap Gate. Report No. R 01-003.San Luis Obispo, Calif.: California Polytechnic State Univ., Irrigation Training and Research Center. Keller, J. 1975. Economic pipe size selection chart. In ASCE, Irrig. and Drain. Div. Spec. Conf. Proc., 109-121. Logan, Utah: American Soc. Civil Engineers. Kemper, W. D., and J. A. Bondurant. 1982. Turbulent flow self cleaning trash screens. In Proc. Irrigation Association Technical Conference, 75-84. Falls Church, Va.: Irrigation Assoc. Kindsvater, C. E., and R. W. Carter. 1959. Discharge characteristics of rectangular thin-plated weirs. ASAE Paper No. 3001. St. Joseph, Mich.: ASAE. Kraatz, D. B., and I. K. Mahajan. 1975. Small hydraulic structures. FAO Irrig. and Drain. Paper No. 26, Parts No. 1 and 2. Rome, Italy: FAO. Merriam, J. L. 1973. Float valve provides variable flow rate at low pressures. In Proc., ASCE Irrig. and Drain. Spec. Conf., 385-402. New York, N.Y.: American Soc. Civil Engineers. Merriam, J. L. 1977. Level-top canals for semi-automation of on-farm irrigation and supply systems. In Proc., ASCE and Drain. Spec. Conf., 217-224. New York, N.Y.: American Soc. Civil Engineers. Merriam, J. L. 1987a. Pipelines for flexible deliveries. In Planning, Operation, Rehabilitation and Automation of Irrigation Water Delivery Systems. Proc. Symp., ASCE, 208-214. D. D. Zimbelman, ed. New York, N.Y.: American Soc. Civil Engineers. Merriam, J. L. 1987b. Design of semi-closed pipeline systems. In Planning, Operation, Rehabilitation and Automation of Irrigation Water Delivery Systems. Proc. Symp., ASCE, 224-236. D. D. Zimbelman, ed. New York, N.Y.: American Soc. Civil Engineers. Parshall, R. L. 1950. Measuring water in irrigation channels with Parshall flumes and small weirs. Circular 843. (Revision of USDA 1932 and 1941 Farmer’s Bull. No. 1683.) Washington, D.C.: U.S. Soil Conservation Service.

390

Chapter 11 Delivery and Distribution Systems

Pugh, W. J., and N. A. Evans. 1966. Weed seed and trash screens for irrigation water. Bull. S-522S. Fort Collins, Colo.: Colorado State Univ. Agri. Exp. Sta. Raemy, F. and W. H. Hager. 1998. Hydraulic level control by hinged flap gate. Proc. Inst. Civil Engineers Water Maritime and Energy 130: 95-103. Replogle, J. A. 1968. Discussion of “Rectangular cutthroat flow measuring flumes” by G. V. Skogerboe and M. Leon Hyatt. J. Irrig. and Drain. Div., ASCE 94(IR3): 359362. Replogle, J. A. 1970. Flow meters for water resources management. Water Res. Bull. 6(3): 345-374. Replogle, J. A. 1977. Compensating for construction errors in critical-flow flumes and broad-crested weirs. Flow measurement in open channels and closed conduits. Natl. Bureau Stds. Spec. Publ. 484(I): 201-218. Replogle, J. A. 1984. Some environmental, engineering, and social impacts of water delivery schedules. In Proc. of 12th Congress Irrigation and Drainage, 965-978. New Delhi, India: ICID. Replogle, J. A. 1986. Some tools and concepts for better irrigation water use. In Proc. Irrigation Management in Developing Countries Current Issues and Approaches, an Invited Seminar Series, 117-148. K. C. Nobe, and R. K. Sampath, eds. Ft. Collins, Colo.: Int’l School for Agricultural and Resource Development, Colorado State Univ. Replogle, J. A. 1987. Irrigation water management with rotation scheduling policies. In Planning, Operation, Rehabilitation and Automation of Irrigation Water Delivery System, I&D Div., ASCE Proc. Symp., 35-44. D. D. Zimbelman, ed. New York, N.Y.: American Soc. Civil Engineers. Replogle, J. A. 1999. Measuring irrigation well discharges. J. Irrig. Drain. Eng. 125(4): 223-229. Replogle, J. A., and A. J. Clemmens. 1979. Broad-crested weirs for portable flow metering. Trans. ASAE 22(6): 1324-1328. Replogle, J. A., A. J. Clemmens, and M. G. Bos. 1990. Chapt. 10: Measuring irrigation water. In Management of Farm Irrigation Systems, 313-370. G. Hoffman, T. A. Howell, and K. H. Solomon, eds. St. Joseph, Mich.: ASAE. Replogle, J. A., and J. L. Merriam, 1980. Scheduling and management of irrigation water delivery systems. In Proc. 2nd Nat’l. Irrig. Symp., 112-126. St. Joseph, Mich.: ASAE. Replogle, J. A., and B. T. Wahlin. 1994. Venturi meter constructions for plastic irrigation pipelines. Appl. Eng. Agric. 10(1): 21-26. Replogle, J. A., and B. T. Wahlin. 2000. Pitot-static tube system to measure discharges from wells. J. Hydraulic Eng. ASCE 126(5): 335-346. Replogle, J. A., B. J. Fry, and A. J. Clemmens. 1987. Effects of non-level placement on the accuracy of long-throated flumes. J. Irrig. Drain. Eng. 113(4): 584-594. Robinson, A. R. 1968. Trapezoidal flumes for measuring flow in irrigation channels. ARS 41-141. Washington, D.C.: USDA-ARS. Robinson, A. R., C. W. Lauritzen, D. C. Muckel, and J. T. Phelan. 1963. Distribution, control and measurement of irrigation water on the farm. USDA Misc. Publ. #926. Washington, D.C.: USDA. Robinson, E. P. 1972. Automatic control of farm channels. 8th Congress, ICID. Question 28.2, Report No. 21: 28.2.293-28.2.296. New Delhi, India: ICID.

Design and Operation of Farm Irrigation Systems

391

Schwab, G. O., D. D. Fangmeier, W. J Elliot, and R. K. Frevert. 1993. Soil and Water Conservation Engineering. New York, N.Y.: John Wiley & Sons. Seipt, W. R. 1974. Waterhammer considerations for PVC piping in irrigation systems. Trans. ASAE 17(3): 417-423. Skogerboe, G. V., R. S. Bennett, and W. R. Walker. 1972. Generalized discharge relations for cutthroat flumes. J. Irrig. Drain. Div., ASCE 98(1R4): 569-583. Skogerboe, G. V., V. T. Somoray, and W. R. Walker. 1971. Check-drop-energy dissipator structures in irrigation systems. Water Management Tech. Report 9. Fort Collins, Colo.: Colorado State Univ. Stoker, R. 1978. Simple, efficient, automatic irrigation. Irrig. Age 12(6): 76-77. USBR. 1951. Irrigation Advisors Guide. Denver, Colo.: Bureau of Reclamation, U.S. Dept. of Interior. USBR. 1963. Hydraulic design of stilling basins and bucket energy dissipators. Eng. Mono. No. 25 (revised). Denver, Colo.: Bureau of Reclamation, U.S. Dept. of Interior. USBR. 1997. Water Measurement Manual. 3rd ed. Denver, Colo.: Bureau of Reclamation, U.S. Dept. of Interior. USDA-SCS. 1978. Furrow irrigation. Chapter 5, Section 15, in National Engineering Handbook. Washington, D.C.: USDA Soil Conservation Service. 175 pp. (These standards are being regularly revised and updated by the USDA Natural Resources Conservation Service; see http://www.info.usda.gov/CED/. Detailed irrigation structure design information applicable to a given state may be obtained from the USDA-NRCS, State Engineer. Readers outside the USA should contact USDANRCS, Washington, D.C.) USDA. 1979. Field manual for research in agricultural hydrology. In Agricultural Handbook No. 224. D. L. Brakensiek, H. B. Osborn, and W. J.Rawls, coordinators. Washington, D.C.: USDA, Science and Education Administration. USDA. 2006. Above ground, multi-outlet pipeline (ft.). Natural Resources Conservation Service. Conservation Practice Standard, Code 431. August. 3 pp. Wahl, T. L., A. J. Clemmens, J. A. Replogle, and M. G. Bos. 2005. Simplified design of flumes and weirs. Irrig. and Drain. 54: 231-247. Walker, R. E. 1987. Long crested weirs. In Planning, Operation, Rehabilitation and Automation of Irrigation Water Delivery Systems. Proc. Symp., ASCE, 110-120. D. D. Zimbelman, ed. New York, N.Y.: American Soc. Civil Engineers. Wang, C., G. Guan, W. Cui, and J. Fan. 2005. Modification for the formula of head loss for long-throated flume and the optimize of bodily form. In Int’l Commission on Irrigation and Drainage, 19th Congress, Question 52; Report 11. New Delhi, India: ICID.

CHAPTER 12

PUMPING SYSTEMS Harold R. Duke (USDA-ARS, Fort Collins, Colorado) Abstract. Pumping plants are frequently a major part of the installation and operating costs of a modern irrigation system. Whether designing a pumping plant or evaluating a proposal from a dealer, the irrigator is well served by basic knowledge of how pumps work and the necessary factors to be considered when choosing and operating the irrigation pumping plant. Operating characteristics, as detailed in literature supplied by the manufacturer, include relationships between flow rate, pumping head, rotational speed, and power required. Optimum pump selection depends on consideration of the soil type, climate, and crops to be grown. Special attention must be given to the pumping plant when the irrigation system changes, as may occur when groundwater levels decline or operating pressure changes due to changing the type of irrigation system. Such changes may result in temptation to add a “booster” pump or connect two or more pumps together into a common distribution pipeline. These changes require a careful assessment if the desired result is to be achieved. Economic operation of the entire irrigation system can depend upon proper selection of the pump, power unit, and fuel type as well as proper routine maintenance, testing, and adjustment. Keywords. Affinity equations, Centrifugal, Efficiency, Engine, Head, Headdischarge curve, Impeller, Motor, Net positive suction head, Power, Pump curve, Pump stage, Pumping rate, System head, Total dynamic head, Vertical turbine.

12.1 INTRODUCTION All pumps are mechanical devices which receive energy from a power source and convert that energy to either velocity or pressure energy of a fluid. As commonly used on farm irrigation or drainage systems, pumps lift water to a higher elevation and/or increase the pressure of the water as necessary to operate the system. 12.1.1 Types of Pumps Pumps are generally categorized into two basic types based on the method by which energy is imparted to the fluid. Positive displacement pumps displace the fluid using such mechanical devices as gears, pistons, screws, diaphragms, vanes, or rollers. The pumping rate is nearly independent of pressure. This type of pump is not commonly used for irrigation or drainage. Fertilizer injection and pesticide spray pumps, however, are often positive displacement pumps and thus provide accurate control of chemical application. Irrigation and drainage pumps, on the other hand, typically transfer energy by kinetic principles such as centrifugal or viscous forces or momentum. These are generi-

Design and Operation of Farm Irrigation Systems

393

Positive Displacement

Total Head, m

1000 Vertical Turbine and

100

Centrifugal Mixed Flow 10 Axial Flow 1

1

10

100

1000

10000

Pumping Rate, L/s

Figure 12.1. Applicability of various pump types.

cally called centrifugal pumps and include centrifugal, jet, vertical and submersible turbines, axial flow, and mixed flow, which has characteristics of both turbine and axial flow. Within these subcategories, the size, shape, speed, number of sequential pump units, and mode of operation can vary significantly. In general, there are sufficient commercially available models to permit proper selection to meet any on-farm irrigation or drainage need. This group of pumps will be discussed in this chapter. Figure 12.1 illustrates generalized applicability of various pump types as determined by total head and pumping rate required. Note that positive displacement pumps produce the highest total heads, but are usually limited to small pumping rates. Most centrifugal pumps can operate over a range of conditions. In general, as the total head increases, the pumping rate decreases as a result of fluid slippage. The most common irrigation and drainage pumps are centrifugal and vertical turbine. For high pumping rates at low heads, such as pumping from a river diversion into a canal, propeller or axial flow pumps are commonly used. Mixed flow pumps use impellers which are shaped to impart the energy by both axial and centrifugal forces. Many of the newer high volume, high efficiency turbine pumps are actually mixed flow pumps. Since both irrigation and drainage waters often contain sediment and many existing wells produce sand, it may be necessary to select a pump that is minimally affected by abrasives. Sediment will wear any pump, and it is desirable, but not always practical, to assure that pumped water is clear. Both propeller and centrifugal pumps can handle a reasonable amount of sediment, but wear caused by sediment will require periodic replacement of impellers and volute cases. Turbine pumps are more susceptible to sediment damage. 12.1.2 Scope of Chapter In this chapter, I emphasize how to select the proper pump for a specific job. The theory of pumps is treated rigorously by Karassik et al. (1976) or can be found in the books by Zipparo and Hasen (1993) or Addison (1966). These authors also discuss the theories used to design new pumps, which is of lesser importance to those charged with selecting an appropriate irrigation or drainage pump. This chapter briefly describes the types of pumps most commonly used for irrigation and drainage applications. I discuss the performance characteristics of pumps,

394

Chapter 12 Pumping Systems

considerations to determine the desired pumping rate, development of the system head curve, and finally pump selection. Information is also provided to aid in selecting a power unit and controls and to determine initial investment and operating costs. Field testing of existing systems is discussed and some suggestions are offered on pump maintenance and operation. Chapters 14, 16, and 17 describe the selection and operation of specific irrigation equipment that normally use pumps. Because of these detailed descriptions, I will not discuss piping on the discharge side of the pump. The construction of drainage wells (sometimes called relief wells) is briefly referenced in Chapter 9. A rigorous treatment of well drilling and development can be found in Ground Water and Wells (Johnson Division UOP, 1975) and is included in such texts as Todd and Mays (2004), DeWeist and Davis (1966), or McWhorter and Sunada (1977). It is important that the designer understand groundwater hydraulics to properly select a pump that will perform satisfactorily when well drawdown changes or adjacent wells interfere with one another.

12.2 PUMP COMPONENTS AND CHARACTERISTICS Texts such as Hicks (1957) or Colt Industries (1979) provide background information and nomenclature for those not familiar with pumps. These books and manufacturers’ literature typically provide cutaway drawings and descriptions of the various components. A brief description of the principle components of centrifugal and turbine pumps follows. 12.2.1 Pump Descriptions and Components 12.2.1.1 Centrifugal pumps. A centrifugal pump is normally located above the water surface or in a dry well adjacent to the water supply. It has one or more impellers fastened to a rotating shaft that turns inside a spiral-shaped volute case. Water enters the eye (center) of the rotating impeller and is forced outward along the vanes of the impeller by centrifugal force. The input energy is converted to velocity head as water is forced outward. On leaving the impeller, part of this velocity head may be converted to pressure. Commonly used centrifugal pumps include: end suction, double suction, and multistage. These descriptors refer to how water enters the eye of the impeller or how many impellers are present. If two or more impellers are installed so that the discharge from one is allowed to flow into the eye of the next, the pump is multistage. Water enters the impeller eye from only one side in an end suction pump, while double suction means that water enters the impeller from both sides. Generally, the impellers are fixed onto the drive shaft which is then coupled to the drive unit—either an electric motor or engine. Electrically powered systems may have a common shaft for the motor rotor and the pump impeller. Because they are compact, centrifugal pumps are often stocked as completely assembled units. Almost all pumps have moving parts which require lubrication to reduce wear. In some instances the bearings are lubricated and sealed at the time of manufacture. In other instances oil or grease must be added periodically or continuously. Even the water itself may be used as the lubricant. Most centrifugal pumps have a packing gland around the drive shaft where it exits the volute case. Some newer pumps use a mechanical silicate seal to prevent leakage at this point and to allow an occasional drop of water to act as a lubricant. Older models use beeswax or graphite-impregnated hemp or Teflon packing, which provides some support to the shaft, acting as a bearing but not permitting large leakage. A very

Design and Operation of Farm Irrigation Systems

395

small amount of leakage through this packing gland is desirable to act as a lubricant. Packing glands must occasionally be tightened or the packing replaced if leakage cannot be controlled. Care should be taken to not over-tighten packing glands, thus cutting off the lubricating water or increasing the torque necessary to turn the pump, as this can result in pump failure or motor burnout. 12.2.1.2 Jet pumps. Jet pumps are a variation of the centrifugal pump, with modifications to allow lifting water from deeper depths than can be realized by direct suction. The “jet” is located near the bottom of the suction pipe. A portion of the water pumped to the surface is diverted back down the well through a secondary pipe for delivery to the jet. The return water is passed through a Venturi where pressure is converted to velocity. The high-velocity water jet transfers part of its momentum to water entering the pump, which increases the pressure near the bottom of the suction line, allowing a higher suction lift than otherwise possible (see discussion of net positive suction head, Section 12.2.2.4). Deeper pumping lifts require a larger percentage of the pumped water to be returned to the jet. Because energy is required to lift and pressurize the water returned to the jet, and this water is not discharged from the pump, jet pumps tend to have low operating efficiencies. Therefore, they are generally used only in small sizes, such as for domestic water supplies, where the cost of pumping is not a major concern. Lubrication of the mechanical parts above ground is similar to that of a centrifugal pump, and should be according to manufacturer’s recommendations. 12.2.1.3 Turbine pumps. The turbine pump also operates on the principle of centrifugal force, with water discharged from the impeller nearly perpendicularly to the axis of rotation. Turbine pumps are often used in deep wells or to raise water from rivers, lakes, or sumps where space is limited or high head is required. These pumps, illustrated in Figure 12.2, operate submersed in water, thus do not require priming. Like the centrifugal, both the vertical turbine and submersible turbine impart energy to the liquid by rotating one or more impellers inside the bowls. One impeller and its bowl are referred to as a stage. When several of these bowl assemblies are coupled so that the flow from the lower impeller moves directly to the eye of the next impeller above, they form a multistage pump. For example, a four-stage pump has four bowl assemblies coupled together. The impellers are fixed on a shaft, called a line shaft. The line shaft of the vertical turbine pump extends through the eye of each impeller, inside the column pipe to the surface where it couples to a motor or gearhead. The line shaft is supported in the center of the column pipe by bearings, typically at 2- to 3-m intervals. Water discharged from the bowls also exits vertically upward and is carried within the column pipe. Oillubricated pumps also have an oil tube surrounding the line shaft to direct lubricant to line shaft bearings. The column pipe supports the bowls beneath the pump head as well as serving as a conduit to direct water to the surface. The distance from the bowls to the pump head is referred to as the pump setting. The pump head serves as an elbow to redirect the flow horizontally and to support the weight of the column pipe, bowls, impellers, and line shaft. It generally rests on a foundation and serves as a mounting bracket for the motor or gearhead. The vertical turbine pump may be either oil or water lubricated. An oil-lubricated pump should be used if sand or sediment is being pumped because the encasing oil tube protects line shaft bearings from exposure to abrasives. A lightweight turbine oil

396

Chapter 12 Pumping Systems

must be allowed to drip into the annulus between line shaft and oil tube continuously while the pump is operating. Some of this oil passes out the weep hole located in the casting of the upper stage. For irrigation, it is not critical whether the water contains small amounts of oil; however, most domestic pumps are water lubricated. The bearings in the bottom of the pump are generally lubricated and sealed when manufactured. Water-lubricated pumps generally use rubber bushings that are lubricated by water flowing up the column. When the pump setting exceeds about 13 m, it is advisable to install a water prelubrication system which will dump several liters of stored water along the line shaft just prior to starting the pump. Otherwise, the top rubber bushings may overheat and seize to the line shaft before water rises in the column pipe to provide lubrication. Adjusting Nut Thrust Bearing

Hollowshaft Motor

Discharge Head

Line Shaft

Column Pipe

Bearing Bowl

Impeller Suction Bell

Figure 12.2. Major components of a typical electrically powered vertical turbine pumping plant.

Design and Operation of Farm Irrigation Systems

397

Gearhead or motor bearings also need lubrication. Motors are available with either an oil reservoir or a grease fitting. Some larger motors and gearheads route a small part of the pumped water through the motor or gearhead castings as a coolant. The submersible turbine was developed for irrigation use in the early 1960s and is becoming quite common. Submersibles are generally cheaper for smaller sized pumps and deeper settings than the vertical turbine. They are popular where the entire pumping plant needs to be below ground for esthetic or security reasons, such as in parks and golf courses. In many instances, submersibles can be used in a crooked well where it is impossible to use a vertical turbine if the line shaft is not straight. A submersible turbine pump has a watertight electric motor coupled beneath the bowl assemblies. The entire motor-pump assembly is submerged so that the water provides cooling. Water enters the first stage from below as in the vertical turbine. The discharge exits the top stage and flows to the surface in a pipe which is unobstructed by bearings or line shaft. This pipe also serves to support the motor and pump. There are no rotating parts except for the rotor in the electric motor, the short line shaft, and the attached impellers. Some disadvantages of submersible turbines are that the motor must have adequate water flow around it to provide cooling, and the length of motor below the pump reduces the maximum pump setting and may reduce the capacity of wells with little saturated thickness. Wells that pump sand may cause bearing problems in the pump and motor. Submersible electric motors of 50 kW capacity and larger are quite expensive, and if there are any problems with the motor, the entire pump and motor must be pulled. 12.2.1.4 Axial flow pumps. Axial flow impellers have a very small component of radial flow, with most of the flow paralleling the axis of rotation. The impeller of axial flow pumps is shaped like a propeller, and is most useful where flow rates and pump speed are relatively high compared to the head delivered. Axial flow pumps are typically used for low-head, high pumping rate applications, such as lifting water from a river diversion into a canal or secondary pumping station, pumping water from a runoff recovery pit to the head of a field, and emptying animal waste storage lagoons. These pumps deliver high thrust loads on the main shaft, and often use heavy brass bearings to support the propeller and main shaft. Because of the shallow setting of most axial flow pumps, the bearings are readily accessible, and they usually require a heavy grease applied with a grease gun. 12.2.1.5 Pump impellers. Three styles of impellers, open, semi-open, and enclosed, are used in centrifugal or turbine pumps. These terms refer to whether the impeller vanes are enclosed by shrouds. An open impeller has no shrouds and is often used to pump liquids that contain organic matter, such as sewage sludge, paper pulp or water containing animal wastes. The semi-open impeller has a single shroud on the side opposite the impeller intake, and must be adjusted so the operating clearance between the open side of the impeller vanes and the bowl or volute face is within very close tolerance of about 0.1 mm. If this clearance is too great, then water simply recirculates from the outer edge to the inlet of the impeller. The closed impeller has shrouds on both sides of the impeller vanes and clearance adjustment is not as critical. Hicks (1957) shows a photographs illustrating the types of impellers and the wear rings which restrict recirculation within the bowl.

398

Chapter 12 Pumping Systems

12.2.2 Pump Characteristic Curves Pumps, because of their mechanical nature, have certain well-defined operating characteristics. These characteristics vary between types of pumps and between manufacturers and their various models. Most pump manufacturers publish information describing how each pump performs over a range of operating conditions. This information, in the form of graphs or tables, is called the pump characteristic curve. Because of minor variations in manufacturing, a new pump may not perform exactly as the characteristic curve indicates. Some manufacturers’ curves represent average performance for the initial test group, but others prepare their curves for the pump having the poorest performance. In some instances, the manufacturer may add a factor of safety to assure that all pumps will provide at least the published head at a specific pumping rate. Four characteristic curves are commonly provided by a manufacturer. The four types of curves are combined in Figure 12.3 and described below. The pump model, indicated in Figure 12.3 as 10XY, typically indicates the outside diameter (10 inches) of the bowl assembly and a specific pump design, indicated by the letters (XY). 12.2.2.1 Pumping rate variation with total dynamic head. This curve is often referred to as the TDH-Q curve and relates the head produced to the pumping rate. Generally, the head produced decreases as pumping rate increases, although some pumps have more complex curves. The most common irrigation pump curves have shapes similar to that in Figure 12.3. The total dynamic head, TDH, which a pump must impart to the liquid can be computed from Bernoulli’s equation. The specific operating point will be illustrated in Section 12.3.3. The TDH-Q curve for a single stage pump is shown in Figure 12.3. Section 12.3.5 describes multistage pumps. The curve can be used to evaluate how the pumping rate will vary due to fluctuations in the total dynamic head. If a pump is operated against a closed valve, the head generated is referred to as the shutoff head, as illustrated in Figure 12.3. The efficiency of the pump at this point is zero because it still consumes energy. The pipe on the discharge side of turbine or

15

65

70

75

80

TDH-Q 82 83 82

Shutoff head, 19.2 m

80 10

Output Power, kW

Total Dynamic Head, m

20

75

E-Q

70

5 8 7

BP-Q

6

NPSHR, m

10 5

NPSHR-Q

0 30

35

40

45

Pumping Rate, L/s

Figure 12.3. Pump characteristic curve for a 1760 rpm, three-stage, vertical turbine pump, model 10XY. Add two percentage points to efficiency for 4 or more stages, subtract two percentage points for single stage.

Design and Operation of Farm Irrigation Systems

399

centrifugal pumps must be capable of withstanding the shutoff head in case a discharge valve is inadvertently closed. 12.2.2.2 Efficiency variation with pumping rate. Efficiency is the ratio of output work delivered by the pump to input work. The relation between efficiency and pumping rate is drawn as a series of envelope curves and labeled E-Q in Figure 12.3. There is generally only one peak efficiency, which occurs at a specific pumping rate. Efficiencies vary between types of pumps, manufacturers, and models, with larger pumps generally having higher efficiencies. The efficiency also depends on the materials of construction, finish on the castings or machining, and the type and number of bearings used. For example, enameled impellers are smoother than bronze or steel and will give a higher efficiency. The output work the pump imparts to the liquid is called the water power, WP (or water horsepower in foot-pound units), and is given by: WP = Q × TDH/C

(12.1)

where Q = the pumping rate, L/s (or gpm) TDH = total dynamic head, m (or ft) C = coefficient to convert units, 102.0 (or 3960) WP = output power, kW (or hp). The pump efficiency, Epump, is determined by dividing Equation 12.1 by the input power, BP, expressed in kW (brake horsepower): Epump = WP/BP = Q × TDH/(C × BP)

(12.2)

where the terms are defined as before. Knowing the pumping rate, total pumping head, and pump efficiency, one can use Equation 12.2 to compute the input power required. Published curves for turbine pumps do not include such losses as line-shaft bearing and gearhead friction. The manufacturer’s reported efficiency for these pumps is for a specific number of stages. It is necessary to adjust the reported efficiency upward or downward, depending on the number of stages. 12.2.2.3 Input power variation with pumping rate. As mentioned previously, the input power is the brake power required to drive the pump, expressed in kilowatts, and the relationship to pumping rate is commonly called the BP-Q curve. The peak power requirement is usually at an intermediate pumping rate as shown in Figure 12.3. Other pumps may have the highest power demand at the lowest pumping rate and the required input power declines as Q increases. The shape of the BP-Q curve is a function of the TDH-Q and E-Q curves. The vertical scale for most BP-Q curves is small, which limits accuracy. Manufacturers suggest computing the BP from Equation 12.2 rather than the curves. 12.2.2.4 NPSHR variation with pumping rate. The fourth characteristic curve is the net positive suction head required, NPSHR, as a function of the pumping rate, i.e., the NPSHR-Q curve. As water is drawn into the pump impeller, its pressure is reduced. NPSHR expresses the amount of energy required to move water into the eye of the impeller. If the absolute pressure in the water drops below the vapor pressure of the water at its current temperature, then the water will vaporize, creating vapor bubbles which can collapse violently when pressure increases, a condition called cavitation. Cavitation must be avoided, as extended operation under cavitating conditions

400

Chapter 12 Pumping Systems

can physically destroy the pump. NPSHR is a function of the pump design, pump speed, impeller shape, liquid properties, and pumping rate and varies for different types of pumps, manufacturers, and models. Values are determined by the manufacturer from laboratory tests. A thorough discussion of how to use NPSHR-Q values is given in below. It is necessary to consider NPSHR only for the first stage of a multistage turbine pump. However, for other pumps operated in series, NPSHR must be checked for each pump. 12.2.3 Changing Impeller Speed Although most pumps can be operated over a range of speeds, the performance is affected by the rotational speed of the impeller. The characteristic curves described by the manufacturer are determined for a specific impeller speed. The speed is generally shown on the pump curve. Most modern electrically driven irrigation pumps are directly coupled to the motor shaft and turn at the same speed as the motor, thus once the motor is selected, operating speed is not a variable. Common nominal operating speeds for electric motors in the U.S. (60 Hz) are 1160, 1760 or 3520 rpm, and pump manufacturers publish characteristic curves for speeds near these. Many smaller pumps operate at 3500 rpm while larger pumps operate at the lower speeds. The most common speed for electrically driven well pumps is near 1750 rpm. Belt-driven electric or internal combustion engine-driven pumps can operate over a range of speeds. Speed of belt-driven electric pumps can be varied by changing pulley sizes on the motor or pump. Internal combustion engine-powered vertical turbine pumps are usually connected through a drive shaft and right-angle gearhead to transmit power to the pump line shaft. Speeds for engine-driven pumps can be varied by changing the throttle setting, changing the gearhead ratio on a vertical turbine, or a combination. Variation in pump speed can be used advantageously to adapt a single pump to a range of system head curves (i.e., changing conditions) or to provide variable design pumping rates. The necessity for a single pump to operate efficiently at two different speeds increases the complexity of pump selection several fold. Mathematical expressions known as affinity laws are described in the Colt Industries Handbook (1979). These are three equations which relate how the pumping rate, head, and input power vary with change in pump speed, and are useful to estimate change in pump performance with small changes in speed. The equations are: Q1/Q2 = RPM1/RPM2 H1/H2 = (RPM1/RPM2)

(12.3) 2

BP1/BP2 = (RPM1/RPM2)

(12.4) 3

(12.5)

where Q, H, BP, and RPM can be in any consistent units, and the subscripts 1 and 2 refer to respective values of the parameters at the two speeds. The affinity laws indicate that the pumping rate varies linearly with speed, the head as the square of the speed, and the brake power as the cube. Thus, a small increase in pump speed will produce slightly more water and a higher head but will require considerably more power. The affinity laws are valid only as long as the pump efficiency remains constant. They do not address how the pump efficiency will change with speed. NPSHR also changes rapidly with changes in speed, but cannot be determined from these equations.

Total Head, m

401

87%

80% 84%

70%

2500 RPM

120

60%

Design and Operation of Farm Irrigation Systems

110 k 95 k W W 75 k W

2000

80

1800

55 k W 40 k W

1200

40

noisy

20 k W

800

7.5 kW

0

0

40

80 120 Pumping Rate, L/s

160

Figure 12.4. Effect of pump speed on TDH-Q, efficiency, and power curves.

Manufacturers test their pumps at various operating speeds and sometimes print complex characteristic curves as a function of pump speed as shown in Figure 12.4. Note that the change of head from the TDH-Q curve as speed changes from 2000 to 1200 rpm is not linear. Also note that the efficiency envelope curves have a shape similar to a system head curve. Because of this similarity in shape, the pump efficiency changes little with small variations in pump speed. However, it is safer to utilize pump curves determined at several speeds to interpolate operating characteristics rather than the affinity laws. The input power, BP, on Figure 12.4 is plotted as a family of lines that indicate the size of motor or engine required to drive the pump, disregarding line shaft and gearhead losses. This plot is useful for estimating power requirements as speed changes, and is a safer alternative than the affinity laws. 12.2.4 Changing Impeller Diameter As mentioned earlier, the operating characteristics of a given pump are also dependent on the diameter of the impeller. It is common practice to reduce the diameter of the impeller in both centrifugal and turbine pumps by machining, thus changing the pump’s performance to better match a specific job. Figure 12.5 shows five TDH-Q and four BP-Q curves. The 28-cm curve is for a full diameter impeller, the 26-cm curve is for a small standard trim in impeller diameter, and the remaining curves are for even greater standard trims. The pump supplier can order the pump with the next larger impeller than required, and reduce the diameter in a lathe before final assembly. Three additional mathematical equations relate impeller diameter to pump characteristics. The Colt Industries (1979) handbook describes these relationships in detail and gives examples of how they can be used. The equations are: Q1/Q2 = D1/D2 H1/H2 = (D1/D2)

(12.6) 2

BP1/BP2 = (D1/D2)

(12.7) 3

(12.8)

where Q, H, and BP are defined as before and D is the impeller diameter. Note that the pumping rate varies linearly with diameter, pumping head with the square of impeller diameter ratio, and input power as the cube.

8 7%

80%

87%

80 %

Total Head, m

dia = 28 cm

70 %

Chapter 12 Pumping Systems 60

60%

402

26

40

24

45

22 20 cm

0

0

kW

30

20

40

15

kW 22 .5 kW kW

80

120

160

Pumping Rate, L/s

Figure 12.5. Effect of impeller diameter on TDH-Q curve.

Impeller diameter may only be reduced, as it is not feasible to increase the diameter by adding additional material. The applicability of the affinity laws with respect to diameter is limited to a maximum trim of about 10%. As with changes in impeller speed, changes in efficiency limit application of these equations. Pump curves often show performance with several diameter impellers as in Figure 12.5. It is common practice to interpolate required trim amounts from characteristic curves such as Figure 12.5, and using the manufacturer’s curves to determine specific trim is preferred over Equations 12.6 through 12.8. 12.2.5 Hydraulic Thrust As energy is transferred to the water, forces act upon the impeller and other components. In a vertical turbine pump, for example, these forces can be sizeable and can stretch the line shafts. Bearings must be designed into the pump, motor, or gearhead to support these forces. The line shaft must be of sufficient diameter to withstand not only the torque required to turn the impellers, but also to limit the amount the line shaft stretches as pressure builds. It is beyond the scope of this text to describe computation of thrust loads in a pump and selection of line shaft diameter. Most pump manufacturer’s catalogs contain an engineering section with the necessary graphs, tables, and sample calculations to illustrate how to compute thrust loads. Line shaft stretch will be discussed briefly in Section 12.3.5.3. 12.2.6 Pumps in Series If a single pump provides insufficient head, then two or more pumps may be connected in series. Pumps are connected so that the discharge from the first pump or stage is piped into the inlet side of the second pump. If more than two pumps are in series, they are connected so all the flow passes successively from one pump to the next. Each successive pump simply adds more energy to the water. Series operation is common with centrifugal pump systems where the operating conditions have changed and the old pump is no longer capable of delivering the required head, such as after a surface irrigation system has been converted to sprinkler. In such a case, a second pump is connected in series to serve as a booster. A multistage turbine or submersible turbine pump consists of single stages connected in series. The same pumping rate passes through all stages and each one adds additional head to the water. Unlike the centrifugal pump with a booster, such a multistage pump is usually made up of identical stages.

Design and Operation of Farm Irrigation Systems

403

B

HA

HB

A HA + H B

Total Head, m

A+B

Pumping Rate, L/s

Figure 12.6. Combination of TDH-Q curves for pumps in series.

The combined head of two pumps operating in series is the sum of the individual heads for each pump at the common pumping rate, as shown in Figure 12.6. By selecting several pumping rates and the respective heads, the combined head, HA plus HB, is obtained to develop the new HA+B-Q curve. If each pump is identical, as in a multistage turbine or submersible turbine pump, it is necessary only to multiply the head for a particular pumping rate by the number of stages to get the combined value. For pumps in series the flow from one pump is identical to the flow in the others, i.e., QA = QB = QC. The combined input power for a series of pumps is the sum of the input power for each pump for the specific Q plus any power needed to overcome friction. A combined BP-Q curve can be developed in a similar fashion to the combined TDH-Q curve. The equation to compute the combined efficiency value is: E ser =

Q × (TDH A + TDH B ) 102 × ( BPA + BPB )

(12.9)

where Q is in L/s, TDH is in m, and BP is in kW. The net positive suction head criteria must be satisfied for each pump in series, but a combined curve has no meaning. Since a pump delivers water at a higher pressure than at its inlet, for identical pumps in series (e.g., a vertical turbine), meeting the NPSH requirement of the first stage will assure that there is no problem with subsequent stages. Such is not necessarily the case with dissimilar pumps, pumps connected by long pipelines, or pumps at significantly different elevations. In such cases, the NPSHR of a downstream pump may be greater than the upstream pump is capable of delivering, especially if the downstream pump has significantly greater pumping capacity than the upstream. 12.2.7 Pumps in Parallel Two or more pumps may also be operated in parallel. A typical example is several small wells with small turbine pumps discharging into a common pipeline. Pumps are

Chapter 12 Pumping Systems

Total Head, m

404

B A

A+B

QA

QB QA + Q B Pumping Rate, L/s

Figure 12.7. Combination of TDH-Q curves for pumps in parallel.

often operated in parallel if the system requires wide variations in pumping rates at approximately the same heads. Sprinkler irrigation systems having varying capacities on different circuits, such as parks or golf courses, often require pumps in parallel. To properly select pumps to be operated in parallel, it is necessary to develop their combined operating characteristic curves. The procedure for developing a combined TDH-Q curve is illustrated in Figure 12.7. First, select a particular total head, and then determine QA and QB. The combined flow is the sum of the two and is plotted against the originally selected head. New values of head are selected and the process repeated. Only if the two pumps are identical will QA equal QB for all values of head. To determine the BP-Q relationships, it is necessary to add the input power required by each pump for the same total head, add BPA to BPB, and then plot against the combined flow, QA+B. Again, it is necessary to repeat the procedure for different heads. The efficiency calculated for two pumps in parallel is: E par =

(Q A + QB ) × TDH 102 × ( BPA + BPB )

(12.10)

The computation sequence requires selecting a specific total head; determining values for QA, QB, BPA, and BPB; calculating efficiency; and plotting against the combined pumping rate. The resulting E-Q curve may have more than one peak efficiency. For more than two pumps in parallel, it is necessary to add the pumping rates in the numerator and BP values in the denominator. If parallel pumps are not taking water from a common source, different pumping heads are generated by each pump and the analysis becomes more difficult. In practice, there are many situations where two or more improperly selected pumps operate in parallel. The consequence is invariably poor efficiency, and one pump may provide virtually all the water while the other operates very inefficiently or even with reversed flow.

12.3 SELECTING A PUMP Proper selection of a pump requires knowledge and use of the pump characteristic curves described in previous sections. An equally important aspect of pump selection is to determine the operating characteristics required of the system. The procedure for selecting a pump is to first determine the desired pumping rate and the relation be-

Design and Operation of Farm Irrigation Systems

405

tween pumping rate and head required by the water delivery system. Only then can pump manufacturers’ characteristic curves or tables be used to select a pump that will operate efficiently at or near the design pumping rate. Engineers, pump installers, and irrigation equipment salespeople have historically determined a single design pumping rate and total dynamic head to use to select the pump. However, few irrigation or drainage systems have fixed operating conditions. A more realistic approach is to select the pump based on a range of pumping rates about the design pumping rate, which will result in satisfactory performance over the expected operating conditions. 12.3.1 Pumping Rate Required The major role of pumping in crop production is to control the soil-water environment. In humid or lowland areas, this may mean pumping to improve soil aeration or drainage. Where precipitation is not reliable or sufficient, pumps are used to increase soil water. Chapters 8 and 9 detail evaluation of requirements for irrigation and drainage, respectively, and define the minimum pumping rate for certain applications. Physical, economic, and legal considerations constrain the maximum pumping rate, particularly for irrigation. The maximum rate at which a well can deliver water is limited by both the aquifer properties and the construction and development of the well. Thus, it is imperative that tests be conducted to determine the capacity of the well and the relation between pumping rate and water level prior to pump selection. The pumping rate from surface supplies may be limited by ditch or pipeline conveyance capacity. Legal rights to water use limit pumping rates in many areas. The consequences of improper selection of pumping rate are costly. Large pumping rates may result in pump surging or wells that produce sand, either of which will damage the pump. Excessive design pumping rate also results in greater than necessary capital cost, higher demand charges for electrically powered units, and may cause excessive runoff and erosion. Selection of the optimum pumping rate requires consideration of both crop water needs and soil characteristics. For drainage conditions, it is generally desirable to pump at a rate to lower the water table a preselected amount within a given time while minimizing pumping cost. Too low a pumping rate, however, results in insufficient drainage and possible crop damage or salt accumulation. 12.3.1.1 Crop water requirements. Crop water use depends on many factors, including climate, type of crop, and stage of crop growth. Methods of estimating crop water use are discussed in Chapter 8. The total seasonal water use is not to be confused with the amount of water that must be supplied by irrigation. Some precipitation occurs in most areas, and the amount that is useful for crop production depends on when it falls, how much infiltrates the soil, and how much can be stored in the root zone. Thus, the total irrigation requirement is less than the total crop requirement by an amount equal to the effective precipitation. Crop water use is not a constant throughout the growing season, but increases with increasing temperature, solar radiation, wind, and crop cover and with decreasing humidity. As discussed in the next section, it is not usually necessary to pump irrigation water at the rate of peak crop water use. 12.3.1.2 Soil water storage. One of the major functions of soil in the agronomic environment is that of a water storage medium. Chapter 6 details soil water transport, plant water uptake, and the concept of available water. All soils are capable of storing

406

Chapter 12 Pumping Systems

some amount of water until it can be used by the plant. This available water capacity may range from 0.05 cm/cm for very sandy soils to 0.2 cm/cm for clay-loam soils. The available water storage capacity in the root zone is typically from three to twenty times the maximum daily crop water use. Thus, the soil serves as a buffer against short-term extremes in crop water use rates. This means that an irrigation pump can provide satisfactory performance even when the pumping rate is somewhat less than the peak crop water use rate. To determine the risk associated with a given irrigation rate and soil type, Heermann et al. (1974) analyzed 60 years of climatic data from eastern Colorado. They present the results in terms of net application rate per unit area (L/s per ha) required to meet crop water requirements for different values of available soil water at several probability levels. It is important to note, however, that the pumping rate must be greater than the net application rate because no irrigation system distributes water with perfect efficiency or uniformity. The design pumping rate must be increased to compensate for application efficiency and lack of uniformity. Application efficiencies may range from 95% or more for drip or in-canopy spray irrigation systems to considerably less than 50% for surface irrigation on coarse-textured soils. The impact of application uniformity on net irrigation water requirements depends on management goals. If the irrigator desires that almost all the field be adequately irrigated (mean low quarter concept, see Chapter 13), then even if the application uniformity is 85%, the total pumping must be 30% greater than the net requirement. Thus, the required pumping rate may be two or more times the required net application rate. The pumping capacity must account for application inefficiency and downtime that may result from system maintenance, load control management for electrically powered pumps, and other interruptions to pumping. The gross capacity needed for such systems can be computed by: Qn Qg = (12.11) Ea (1 − Dt ) where Qg = gross pumping capacity needed Qn = net capacity needed Ea = application efficiency, expressed as a decimal fraction Dt = fraction of the pumping period that the pumping systems is inoperable (i.e., the downtime), as a decimal fraction. 12.3.1.3 Soil intake rates. The rate of water infiltration is an important consideration for any irrigation system. It is probably the most important factor for surface irrigation systems because intake rate controls advance of water across the soil surface. The influence of intake rates on surface irrigation systems is covered in detail in Chapter 13. The role of intake rates is somewhat different for sprinkler irrigation systems. Sprinkler systems are capable of high application efficiency and uniformity only if the system, rather than the soil, controls infiltration. The application rate must be sufficiently low that the water applied is either infiltrated immediately or stored in small surface depressions until it infiltrates. Otherwise, water runs across the surface to low areas, and the uniformity advantages of the sprinkler are lost. The application rate of a moving sprinkler depends on the water flow rate, the shape of the sprinkler distribution pattern, and the area wetted by the sprinkler. The area wetted by a given sprinkler head is relatively independent of the flow. Thus, the

Design and Operation of Farm Irrigation Systems

407

application rate is almost proportional to the sprinkler flow rate. For moving sprinklers, the depth of application depends on speed of movement but peak application rate does not. Thus, for sprinkler systems, too low a pumping rate will provide insufficient water for optimum crop growth, while too large a pumping rate will result in surface runoff. Excessive surface runoff leads to reduced uniformity and necessitates increased pumping to adequately irrigate the “dry spots.” 12.3.1.4 Management criteria. The pumping rate is also affected by operational decisions such as the hours of irrigation per day and irrigation days per week. Most irrigators anticipate some shut-down time to maintain the system. Most surface irrigation systems require labor to change sets, and pump schedules are often modified to coincide with labor availability. When the hours of operation are reduced to make these allowances, then the pumping rate must be increased to provide adequate water during the critical crop period. Other operating decisions that impact the design pumping rate are the cropping pattern and the area to be irrigated. Preplant irrigation to fill the root zone can reduce the design pumping rate for certain crop, climatic, and soil conditions. 12.3.2 Total Pumping Head 12.3.2.1 System head curve. The system head curve describes the total head required to move water through the pumping system as a function of water flow rate. The total head required is independent of the pump, except for the friction loss that occurs in the column pipe of a vertical turbine pump. Figure 12.8 shows a typical system head curve which illustrates the parameters contributing to the total head and how they vary as the pumping rate increases. Not all of the parameters illustrated are applicable to every system. The total head is the sum of the static lift, static discharge head, well drawdown, friction, and operating pressure head. Static lift is the vertical distance between the centerline of the pump discharge and the water source elevation when the pump is not operating. If the water source is below the pump elevation, the static lift is positive. If the pump is located below the wa-

System Head

Head, m

Pressure Head

Friction Head Well Drawdown Static Discharge Head Static Lift Pumping Rate, L/s

Figure 12.8. Components of the system head curve.

408

Chapter 12 Pumping Systems

ter, for example in a sump, the static lift is negative. Static lift is independent of the pumping rate, although it may vary with time. In that case the system head curve is also time dependent. The static pumping head is the elevation difference between the centerline of the pump and the eventual point of use. When a pump discharges directly into a canal through a level pipe, the static pumping head is zero. If, however, the pump supplies water to a distant point at a different elevation, then it is necessary to compute the static pumping head. To obtain this value, subtract the elevation of the pump from the elevation of the final point of delivery. The static discharge head is independent of the pumping rate, but may also be time dependent. For example, the static discharge head will change with time if pumping to a center pivot that moves over uneven terrain. Pumping water from an aquifer causes the water level in the well to decline. This well drawdown depends on the pumping rate, the aquifer properties, the well radius, and the time the pump has operated. The drawdown can be computed using one of several equations, depending on the conditions. Methods for calculating drawdown are given by McWhorter and Sunada (1977), Todd and Mays (2004), and DeWiest and Davis (1966), among others, and are beyond the scope of this chapter. The most satisfactory method to determine the relationship between well drawdown and pumping rate is to conduct a well test. To design a pump, the well test is best conducted by the well driller or pump installer, using the contractor’s own pump. The well is pumped at various rates, drawdown allowed to stabilize, and the drawdown is measured. The well test can be conducted most quickly by beginning pumping at the highest flow rate, allowing the drawdown to stabilize, and then successively reducing the pumping rate. Care must be taken to pump at a constant rate for sufficient time that the drawdown approaches a steady value. Friction causes loss of head when water flows through a pipe system. The friction head can be determined from the Darcy-Weisbach or Hazen-Williams equations, which are described in Chapter 15 and in most hydraulics texts. Friction tables, nomograms, or curves provided by pipe manufacturers are convenient to evaluate pipe friction. Friction tables for steel pipe can be found in the Colt Industries (1979) and Ingersoll-Rand (1977) hydraulic handbooks. Friction tables for plastic, gated aluminum, and concrete pipe can be obtained directly from the manufacturer. The pump must add sufficient energy to the water to overcome friction. The water velocity in a pipeline increases linearly with the pumping rate, but friction increases as the velocity squared. Irrigation systems with long pipelines or undersized pipe may have very significant friction. It is often economically feasible to select a larger pipe size in order to reduce the cost of overcoming friction. Friction loss must be determined on both the intake and discharge side of the pump. It is necessary to compute the friction on the suction side of centrifugal pumps, especially, to assure that there is sufficient net positive suction head available to prevent cavitation. This will be discussed more later in Section 12.3.6. The friction loss in the pump head and column pipe of a vertical turbine pump must also be computed. Manufacturer’s tables or curves report friction losses of column pipe as a function of the pumping rate and account for the presence of the line shaft and supporting bearings or bushings. Friction head in the column should be kept below 5 m/100 m. If it exceeds this amount, consideration should be given to the next larger column pipe size.

Design and Operation of Farm Irrigation Systems

409

Discharge head is the pressure head required at the point of discharge, such as the outlets to a gated irrigation pipe, the head of water above the pump outlet in a drainage ditch, or the pressure on a sprinkler. Except where water is discharged directly into an open ditch or partially filled pipeline, all irrigation pumps require some discharge head. The operating pressure and its relation to pumping rate are discussed in Chapters 16 and 17. Total head curves may be continuously increasing functions as illustrated in Figure 12.8, or they may be step functions, such as with a center pivot having an end gun which starts and stops. Proper selection of a pump for a specific system requires knowledge of how the total head changes. The total head-flow rate relationship should be evaluated over the entire range of operating conditions. If the water level in a well drops during the pumping season or a center pivot sprinkler moves over significant elevation differences, then attention must be given to select a pump that can satisfy all conditions. Most pumps will not operate at peak efficiency over wide ranges in total heads. If a particular condition is prevalent, such as center pivot operation with the corner swing arm operating, then the pump should be selected to operate efficiently for that condition yet satisfactorily for the other conditions encountered. 12.3.2.2 Variations in system head curve. The system head curve is time dependent because well drawdown, friction, operating conditions, and static water level change. Figure 12.9 illustrates two conditions which change the system head curves. Figure 12.9a illustrates that the friction increases, i.e., the slope of the curve increases, as steel or iron pipe corrodes. Steel or cast iron pipes may become partially blocked by development of oxidized ferrous nodules or the pipes may rust through and begin to leak. Vertical turbine pump column pipe is usually steel and is quite susceptible to deposition or failure due to rust. Aluminum and polyvinylchloride (PVC) pipes, which are commonly used in recent years, do not deteriorate by rusting, but the friction factor for these materials can be increased by deposits on the walls. When groundwater pumping exceeds the aquifer recharge, then the static water table is expected to decline. Static water level decline, as illustrated in Figure 12.9b, increases the system head curve. The rate of drawdown increases when the water table declines, because saturated thickness is reduced. Proper selection of a pump for these conditions requires consideration of how pumping conditions will change over the life of the pump. (b)

Old Pipe

New Pipe

Pumping Lift, m

Friction Head, m

(a)

2000 1980 Drawdown 2000 1980 Static Lift

Pumping Rate, L/s

Pumping Rate, L/s

Figure 12.9. Effects of (a) pipe deterioration and (b) declining water levels on system head curves.

410

Chapter 12 Pumping Systems

For example, if additional wells are being drilled and new land irrigated, then one might expect the rate of water level decline to accelerate. On the other hand, if pumping is limited by regulation or artificial recharge is initiated, then static water levels may stabilize or even recover. It is important to know whether there will be well interference between the depression cones of nearby wells or whether underground boundaries impact a well’s longterm performance. Failure to consider these factors can result in installation of an oversized pump, which will begin to surge as the water table declines more rapidly than expected. 12.3.3 Matching Pump Characteristics and System Head It is possible to superimpose a pump’s characteristic curves of TDH-Q and E-Q upon a system head curve, as illustrated in Figure 12.10. The point at which the TDHQ and system head curves cross is the operating point. The pump in Figure 12.10 will produce 60 L/s at a dynamic head of 17 m. In Sections 12.2.3 and 12.3.2 above, I discussed how system head curves may change with time, operating conditions, or deterioration of components. It is necessary to estimate the two extremes and the average system head curves as illustrated in Figure 12.11a. A pump should be selected to operate efficiently under the dominant operating conditions, but it must also deliver water at the required head under the extreme circumstances. As illustrated in Figure 12.11a, the pump will have to produce significantly different dynamic heads for a desired pumping rate of 60 L/s. In an earlier section I showed that pumping rate decreases as head increases for both centrifugal and turbine pumps. Referring to Figure 12.10, the pumping rate will drop from 60 L/s to 56.5 L/s when the head is increased 1 m. If the system head decreases from 17 m to 15 m, the pumping rate increases to 65 L/s. Figure 12.11b illustrates the pumping rate change due to a change in the system head curve. The curves sloping upward to the right are system head curves for spring and fall. Perhaps the spring curve is lower due to a higher static water level prior to the pumping season. The water level declines through the season to a level represented by the fall curve. TD H-Q

17 16 15 Syst

ead em H

E-Q

80 75

55

60

65

Pumping Rate, L/s

Figure 12.10. Superposition of system head curve and TDH-Q curve to determine pump operating point.

Efficiency, %

Head, m

18

Design and Operation of Farm Irrigation Systems

411

(b)

(a)

Pump B

Avg Low

60

Pumping Rate, L/s

Total Dynamic Head, m

Total Dynamic Head, m

High Fall

Pump A

Spring

50

55

60

Pumping Rate, L/s

Figure 12.11. The system head curve should be determined for expected extremes (a) so that the effect of seasonal changes on pumping rate can be determined for prospective pumps (b). The flow rate of pump B will change much less than that of pump A with seasonal changes.

The two curves labeled A and B which slope downward to the right are TDH-Q curves for two possible pumps. The spring pumping rate for pump A is 60 L/s, but will drop to 50 L/s by the end of the season because of the increased head. Pump B will also deliver 60 L/s in the early season, but will only drop to 55 L/s in the fall. Superimposing the TDH-Q and E-Q curves upon the system head curves (if all have the same vertical and horizontal scales) for several pumps being considered provides a visual means of comparing each pump. The pump with the steepest TDH-Q curve will have the least fluctuation in pumping rate when water levels or discharge heads vary. If water is provided from a well in which the pumping level varies over time, then it is desirable to select the pump with the steepest curve to maintain a nearly constant pumping rate. Conversely, the flow rate to a center pivot with a corner swing arm must vary as additional sprinkler heads turn on or off. This system will benefit from a relatively flat pump curve. 12.3.4 Efficiency Considerations Superimposing both TDH-Q and E-Q curves on the system head curve also illustrates how pump efficiency will change as the operating points change. The shape of the E-Q curve is not the same for each pump and this shape has a large influence on which pump will operate most efficiently over the range in operating conditions. Ideally, the operating point on the TDH-Q curve of the selected pump should fall slightly to the right of the peak efficiency. As the pump wears, the TDH-Q curve moves downward in a manner similar to that of a smaller diameter impeller. The operating point will move to the left along the TDH-Q curve, and the efficiency will increase. The net result is a decrease in input power required. If the new pump is selected to operate to the left of the maximum efficiency, pump wear will move the operating point to the left, further reducing the efficiency. The importance of efficiency on power consumption can be computed with Equation 12.2. A simple example using Figure 12.12 may also be informative. Assume an irrigator contacts two pump dealers to obtain a pump that will produce 60 L/s (1000

412

Chapter 12 Pumping Systems

gpm) with a TDH of 30 m (100 ft). Dealer A selects a 25 cm (10 in.) pump, which will produce 10 m of head per stage at the desired pumping rate. Thus, it will require three stages of this pump to match the system head requirement. The cost of the three-stage pump is $3000. Note the operating efficiency of this pump is just on the right-hand side of the peak and has a value of 83%. Dealer B selects a 35 cm (14 in.) pump, which will produce 15 m head at the required pumping rate. It is necessary to have only two stages to meet the required 30-m head. The pump costs only $2400 because of the one less stage, but is only 63% efficient and operates far left of the peak efficiency. Since both pumps will produce exactly the same amount of water and meet the system head required, the farmer may select the cheaper pump. Some pump dealers have also emphasized that the simpler pump (fewer stages) is also better. Until recent years, little effort has been made by pump dealers to calculate the impact of efficiency on total pumping costs. Table 12.1 compares the operating cost for the two pumps under a typical irrigation scenario. Note that the savings in power costs alone is $975 per year, which will more than offset the higher purchase price. Because of the lower efficiency of pump B, it requires a 37.5 kW motor instead of the 30 kW motor on pump A. Many electrical suppliers impose a demand charge to compensate for higher costs of servicing large loads, which adds further to operating costs.

20

55

Dealer A Pump diameter 25 cm

60 70 70

20

(a)

40

75 84 80 80 75

60

Pumping Rate, L/s

Total Head, m

Total Head, m

60

10

Dealer B Pump diameter 35 cm

20 75

80 85 80 75 70

10

70 60

60

60

80

(b)

80

100

Pumping Rate, L/s

Figure 12.12. Comparison of the characteristics of two pumps, each of which will deliver the design pumping rate and head.

Table 12.1 Comparative costs[a] for the two pumps of Figure 12.12. Item Pump A Pump B Initial investment $3000 $2400 Stages 3 2 Motor size 30 kW (40 hp) 37.5 kW (50 hp) Annual demand charge $450 $560 Energy consumed 45,000 kWh 58,900 kWh Annual power bill $3150 $4125 Average annualized cost $3750 $4805 [a]

Cost figures based on demand charge of $15 per kW, assumed 395,000 m3 (320 ac-ft) pumped annually, power cost $0.07 per kWh, and 20-year expected life of the pumping plant.

Design and Operation of Farm Irrigation Systems

413

12.3.5 Selecting Specific Pump Components 12.3.5.1 Number of stages. The previous section described a procedure to obtain characteristic curves for two or more pumps in series. Superimposing these combined curves on the system head curve, as illustrated in Figures 12.10 and 12.11b, allows evaluation of a given pump operation. Multiple pumps operated in series are necessary when the required pumping heads are greater than what a single pump or stage can develop. In this section, I discuss how to determine the number of stages or series pumps needed. The selection procedure is similar to that for a single stage pump, and requires that the system head curve and design pumping rate be known. The manufacturers’ catalogs are checked to determine pump models that will produce the design pumping rate efficiently. Two such candidate pumps are shown in Figure 12.13. Single stage TDHQ curves for pumps A and B are shown in the lower part of the figure. The design Q is 60 L/s with a required system dynamic head of 120 m. To determine the number of stages of a pump required, it is necessary to divide the 120-m head requirement of the system by the head produced by each pump for the design Q of 60 L/s. Pump A produces 20 m for a single stage, thus will require six stages. Pump B produces 60 L/s at 40 m per stage, thus only three stages are required. Once the number of stages is known, the multistage TDH-Q curve can be developed by simply summing the TDH for the same values of Q. Curves of TDH-Q for the six-stage A and three-stage B pumps are plotted on Figure 12.13. Note that the single stage pump curve for B is steeper than for A, but the multistage TDH-Q curve for the six-stage A curve is steepest. Thus, the slope of the multistage TDH-Q curve is a function of both the number of stages and the slope of the single stage curves. Division of the 120-m system head by the single stage pump head results in exactly six and three stages, respectively, for the pumps illustrated in Figure 12.13. Such is 160

A - 6 Stage

Total Dynamic Head, m

B - 3 Stage 120 d Hea em t s Sy

80 B-

40

1S

A-1

tag e

Stag e

50 60 Pumping Rate, L/s

70

Figure 12.13. Comparison of multistage pump curves for two bowl assemblies with different TDH-Q characteristic curves.

414

Chapter 12 Pumping Systems

Total Dynamic Head, m

St g

e ag St

C2

-3

160

C

C3

w/ Tr im Sta ge

120

80

ead em H Syst C- 1

Sta ge

40

0

50 60 70 Pumping Rate, L/s Figure 12.14. Illustration of impeller trim to match a multistage pump to the desired pumping rate and head.

seldom the case, however, as illustrated by pump C in Figure 12.14. Division of 120 m head required by 50-m head produced per stage of pump C indicates 2.4 stages are required. Two stages will produce 57 L/s at 112 m head, while three stages will produce 63 L/s at 130 m. A solution to the lack of proper match is to select a three-stage model C pump but reduce the diameter of the impellers. Figure 12.5 illustrates how the TDH-Q curves change with impeller diameter. Manufacturer’s curves or the affinity laws (Equation 12.6) are used to compute the required trim. The pump industry uses two different approaches to solve the problem represented by pump C in Figure 12.14. One approach is to trim all three impellers to the same diameter as described above to meet the required operating conditions. The other approach, often used with closed impeller designs, is to mix similar bowl models or impeller sizes to obtain the match. There seems to be no consensus among manufacturers as to which is best. If only one impeller is trimmed, it should be assembled as the top stage on a vertical turbine pump, to minimize the risk of cavitation. Whenever impellers are trimmed, it may be necessary to reshape the vanes of the impellers to minimize turbulence as water exits a blunt, thick vane tip. 12.3.5.2 Centrifugal pumps. The selection criteria described are also applicable for centrifugal pumps. It is common to trim impellers of these pumps to match the specific use. The decisions that must be made for pump selection are which model of pump to use, how much to trim the impeller, and what speed to turn the impeller. As described in the preceding section, manufacturer’s characteristic curves contain information needed to make these decisions. A centrifugal pump has suction and discharge fittings or flanges by which it is connected to suction and discharge pipes. Care must be taken to minimize friction loss on the suction side to avoid cavitation. Although some are capable of self-priming at small suction head, most centrifugal pumps used for irrigation or drainage require a foot valve on the intake end of the suction pipe to maintain prime when not operating.

Design and Operation of Farm Irrigation Systems

415

12.3.5.3 Vertical turbine pumps. The selection of a specific pump model, determination of the number of stages, and need for impeller trim have been discussed previously. Other components that must be selected are the lineshaft, column pipe, pump head, and drive units. Line shaft size must be selected to assure that the necessary power can be transmitted to the impellers and that the shaft will not stretch excessively under load. Manufacturers’ curves and tables account for the rotational speed, shear strength of the shaft material, and the brake power transmitted to determine the shaft size required to withstand shear stresses. The lineshaft is subjected to vertical forces that stretch the shaft. These vertical forces are called thrust load and are the sum of the weight of the impellers, the weight of the lineshaft, and the hydraulic thrust on the impellers. The hydraulic thrust is computed by multiplying a thrust constant, provided by the manufacturer for each impeller, by the total dynamic head generated by the pump. Lineshaft elongation is computed from thrust load using tables or graphs. Lineshaft stretch may not exceed the impeller clearance within the bowl, else an impeller adjusted to avoid rubbing on top of the bowl when started under no load will rub against the bottom of the bowl as pressure increases. When this occurs, it is necessary to either select a larger shaft or have a greater clearance machined into the bowls. The column pipe transmits water from the bowls to the pump discharge head. The size, length of the column, and type of pipe connection must be selected. The column pipe diameter should be large enough that the design pumping rate will result in friction losses less than 5 m/100 m. Manufacturers’ friction loss tables give the head loss per 100 m of column pipe as a function of pumping rates, and are different for water and oil lubricated pumps and for each size of lineshaft. The column pipe must be long enough to keep the bowls submerged, which requires that the depth to water be known under pumping conditions. The pumping depth is the sum of the depth to the static water level and any drawdown, as described previously. A turbine pump should have a suction pipe to direct water vertically before it enters the bottom stage. The length of suction pipe may be dictated by the depth of the well or sump and the necessity to set the pump near the bottom to permit pumping during low water conditions. Special belled inlet sections minimize head losses in the suction line, and in any case, suction pipes must not be so long that they reduce NPSH below the required value. A strainer or screen is not commonly installed on a pump operating in a well with a well screen. However, pumps operating in open sumps, rivers, or lakes should have strainers or screens on the intake side to prevent entrance of objects that might damage the impellers. These strainers or screens must be kept clean, as plugging may result in pump cavitation. The pump discharge head serves to support the weight of the column, lineshaft, and bowl assemblies, acts as an elbow to redirect the flow to a horizontal direction, and provides a mounting bracket for the motor or gearhead. The discharge head must be compatible with the size of the column and the physical dimensions of the motor or gearhead. The discharge flange or threaded coupling size is also a consideration. Discharge and column pipes are commonly the same size. The pump head should be sufficiently large that the friction loss is less than 1 m. Manufacturers’ tables show the friction loss as a function of the pumping rate, as well as the physical dimensions of several types of pump heads.

416

Chapter 12 Pumping Systems

The electric motor or gearhead must match the physical dimensions of the discharge head, and must also accept the lineshaft size selected and have sufficient bearings to carry the thrust load. A short length of lineshaft, known as a head shaft, may need to be fabricated to length to assure that the threaded sections are in the correct location to permit tightening of the lineshaft nut for impeller adjustment. Gearheads are available with a range of gear ratios to match the optimum speed of the power source to the design speed of the pump. 12.3.6 Checking NPSHA to Prevent Cavitation Energy is required to move water into the eye of the impeller. This was defined previously as the net positive suction head required, NPSHR. The pump and its suction piping must be analyzed to assure that the net positive suction head available (NPSHA) is greater than that required (NPSHR). If the NPSHA does not exceed the NPSHR, then the pump will cavitate. The positive suction head comes from either atmospheric pressure or a pressure induced on the intake side of the pump, such as by an upstream pump or a higher water level. NPSHA can be computed by NPSHA = Pb – Hs – hf – Pv (12.12) where Pb = barometric pressure Hs = pumping suction lift hf = friction head in the suction pipe Pv = the vapor pressure of the liquid being pumped at the operating temperature. All parameters are expressed in meters. References such as the Colt Industries (1979) handbook and many physics texts give values for atmospheric pressure at various altitudes. The barometric pressure head, Pb, in m of water can be estimated as: Pb = 10.33 – 0.00108 E (12.13) where E is the elevation in m. The vapor pressure and specific gravity of water as a function of temperature can be found in similar references. Vapor pressure, Pv, can be estimated from Table 12.2. °C Pv, m

Table 12.2. Water vapor pressure as a function of temperature. 0 5 10 15 20 25 30 35 40 45 0.06 0.09 0.13 0.17 0.24 0.32 0.43 0.58 0.76 0.99

50 1.28

Friction loss in the suction pipe depends on pipe length, diameter, and fittings as well as the pumping rate and must be determined accurately using friction tables or the Darcy-Weisbach or Hazen-Williams formulas. Example. Compute the NPSHA for a pump that is operating at an elevation of 1525 m (5000 ft) mean sea level (MSL) where the water temperature is 12°C (60° F). Assume the pumping rate is 65 L/s (1030 gpm) and the suction line is a schedule 40 steel pipe 20 cm (8 in.) in diameter and 200 m (656 ft) long. The centrifugal pump is located 2 m (6.56 ft) above the water surface, and has an NPSHR of 1.2 m (3.9 ft). The atmospheric pressure head, from Equation 12.13, is 8.68 m (28.5 ft). From Table 12.2, the water vapor pressure is 0.15 m (0.49 ft) and the specific gravity (Colt Industries, 1979) is 0.999. The friction loss from Table 1, page 47, of the Colt Industries handbook is 1.65 m/100 m length of pipe or 3.3 m for 200 m. Substituting these values into Equation 12.12, as shown graphically in Figure 12.15, results in a com-

Design and Operation of Farm Irrigation Systems

417

puted NPSHA of 3.23 m, which is greater than the NPSHR of 1.16 m. Therefore, the pump will operate safely without cavitation. If NPSHR exceeded NPSHA, then the pump could be lowered nearer the water surface or the suction pipe could be enlarged to reduce friction loss. Changes in the system head or in water temperature may cause a well-designed pump to cavitate, thus these calculations should be based on the maximum expected NPSHA over the system life.

Atmospheric Pressure Head

Atmospheric Pressure NPSHA, m Vapor Pressure, m Friction Head, m Suction Head, m Absolute Zero Pressure

Figure 12.15. Determination of net positive suction head available to assure pump will not cavitate.

12.3.7 Variable Pumping Rates It is not always desired that irrigation systems operate at a constant pumping rate. For example, if several center pivots are supplied from a single pumping station, the desired pumping rate depends upon which pivots are to be operated. When a center pivot corner arm extends, additional sprinkler heads are turned on, necessitating a higher pumping rate or a slower angular speed. Different zones of a solid set sprinkler or microirrigation system may require different pumping rates. Over the longer term, declining water tables during the season or increased friction caused by deteriorating pipes may decrease pumping rates. There are several methods by which pumping rates can be matched to variable flow requirements. Where water is supplied from a multiple pump station, it may be feasible to change the number of pumps operating to match the desired flow rate and pressure. Less desirable options to reduce pumping rate and pressure include valving to reduce irrigation system pressure (which increases pump discharge pressure), partial recirculation of pumped water, and applying water when not needed. Each of these alternatives results in a steeper system head curve than necessary and may result in little or no reduction of power requirements, and thus energy cost. A much more desirable alternative is to change the speed of the pump. The effects of pump speed on pumping rate, pressure developed, and power required are described in Section 12.2.3. With units powered by combustion engines, a speed change is readily accomplished simply by changing the throttle setting. However, electrically powered units operate at a nearly constant speed, dependent primarily on the frequency of the alternating current supplied. Thus, changing speed of electrically powered pumps is more complex than for combustion-powered units. Adjustable speed or variable frequency drives (VFD) allow adjustment of the frequency of the electrical supply to the pump motor, thus allowing the speed of the

418

Chapter 12 Pumping Systems

pump to be adjusted to match water flow requirements. Such VFD have a distinct advantage over other methods for adjusting the pumping rate of electrically powered units because reducing the pumping rate (by reducing speed) also reduces power requirements. From the affinity equations (12.3 through 12.5), a 20% reduction in pump speed (and pumping rate) reduces the power consumption by 40% to 50%, depending on the change in pump efficiency that accompanies the speed change. Thus, a pumping plant can be designed to deliver the maximum pumping rate and the speed can be reduced to obtain significantly reduced rates. VFD pumps make it easier to implement daily management decisions, may reduce the equipment requirements if a single VFD pump can replace several fixed capacity pumps operating in parallel, and have the potential to significantly reduce energy consumption. Variable frequency drives are not without disadvantages, however. In order to change the frequency of the power, the AC current must be rectified to DC, then an AC current of the desired frequency must be produced with an inverter. The solid state electronic devices necessary to process the very high currents needed for a large pump are quite expensive to design, purchase, and install (Poole, 1993). In rural areas, it may be difficult to find qualified service personnel since these devices are most widely used in industry. The square wave power produced (rather than the sine wave of conventional line power) may cause problems with pump motors. The solid state components of the VFD are especially vulnerable to overheating in the severe conditions to which irrigation pumps are often exposed. The ideal irrigation application for a VFD is a system in which a single pump serves a large area planted to varied crops such that pumping rate requirements vary widely (Nesbitt, 1995). Much of the operation should be expected to be at a reduced pumping rate, and the irrigation season should be long. Systems serving a small area, a single sprinkler system served by a single pump, monoculture cropping systems, or low annual operating hours (typical for irrigation) are poor candidates for VFD. The irrigator should expect little savings in electrical demand charges from installing a VFD, although the VFD does provide a “soft start,” or reduced current inrush upon starting the motor, as compared with conventional single- and multi-phase motors. One should also remember that pump efficiency is not a constant as speed is changed, so careful design and pump selection are necessary. VFD provides a potential energy savings of 20% to 50% under favorable conditions, if annual operating hours are high (an industrial plant operating around the clock will run for more than 8700 hours per year, compared to 1500 to 2000 hours typical for irrigation systems in temperate areas). Hanson et al. (1996) found that VFD reduced power requirements of five pumping plants tested by 32% to 56% during the times reduced pumping rates were needed. The pump load characteristics must be variable, and the pump must have a moderate to high nameplate power rating, particularly if it operates at more than about 75% of full load most of the time. A VFD may be particularly attractive for installations where the initial pump installation is overdesigned to allow for possible future expansion of the irrigation system. Before deciding on a VFD, the irrigator must compute energy use based on flow rate, head, efficiency, and hours of operation at each level of flow and compare the power savings of a constant speed system with the capital cost of a VFD. Cost of equipment, engineering, and installation for a VFD may range from $2,500 for a 7.5 kW unit to more than $16,000 for 110 kW (Hanson et al., 1996). The Electric

Design and Operation of Farm Irrigation Systems

419

Power Research Institute (Poole et al., 1989) has developed a computer program (ASCON-I) to evaluate the economics of VFD.

12.4 POWER UNITS Pumping plant power units convert electric energy or fuels into mechanical energy. Electricity is widely used as a power source for pumping, as are natural gas, diesel, liquified petroleum (l-p) gas, and gasoline. Irrigation and drainage pumps may be permanently located at a site, permitting engines or motors to be permanently connected to a power supply. Portable pumps are generally driven either by an engine mounted on the pumping unit or by the power take-off of a farm tractor, and utilize a temporary fuel supply. The cost of energy in the U.S. for pumping prior to the mid-1970s was, in many areas, minimal. However, regardless of the source, energy costs have risen dramatically since that time, and have become a major agricultural production cost. Thus, it is imperative that the entire pumping plant be efficient to keep pumping costs as low as possible. This requires careful selection of the pump and power plant for the specific job. In Sections 12.2.2 and 12.3.4, I discussed the power required to drive a pump. When using Equation 12.2 to compute the required output of a power unit, it is important that the TDH value used include all friction head losses. Having determined the pump input power required, the designer is ready to select a specific electric motor or combustion engine. Selecting too small a unit will overload the power unit, which may result in inadequate pump speed, reduced power unit efficiency, and reduced motor or engine life from overheating and excessive bearing wear. Oversizing the power unit results in higher than necessary investment costs and may result in lower efficiencies and higher power costs. It is generally not necessary to compute the power consumed by bearing losses for centrifugal and short-set turbine pumps. When turbine pumps are set more than 20 m, however, bearing losses can be significant and must be estimated using manufacturers’ curves or nomographs. Both lineshaft bearing and motor or gearhead thrust bearing losses must be added to the brake power computed with Equation 12.2. 12.4.1 Electric Motors Prior to 1945, most electrically driven irrigation and drainage pumps were belt driven. Most modern installations couple the motor directly to the pump drive or line shaft. The impeller of many centrifugal pumps is mounted directly to the extended rotor shaft of the motor. Vertical turbine pumps commonly use hollowshaft electric motors with a tubular rotor shaft through which the pump lineshaft or headshaft passes. The direct coupled motor and pump is much safer than belt drives because exposed rotating parts are minimized. They are also more efficient, because belt slippage and power to flex the belts are eliminated. The major advantage of belt driven pumps was that pump speeds could be easily changed by changing pulleys. 12.4.1.1 Motor selection. The size of electric motor must be large enough to deliver the required energy to the pump. In years past, it was common practice to allow overloading electric motors as much as 10% to 15% above their nameplate rating. Most motors will operate under such excess load, but overloading results in higher operating temperatures. Older motors typically had cast iron frames, copper windings, were rated at service factors of 1.15 or greater, and performed satisfactorily with moderate overloads. During the late 1970s, the electric motor industry began using lighter

420

Chapter 12 Pumping Systems

(a)

80

60 40

20

0

50

100

150

Percent of Rated Load

200

(b)

.78

M RP 00 M 18 RP 00 M 15 RP 00 12

Efficiency, percent

100

Fuel Consumption, kg/kWh

motor frames, aluminum windings, different insulation material, and typically changed the service factor to 1.0 or 1.05. This lower service factor indicates that the motor will no longer tolerate the 10% to 15% overload. In general, motors available on the market today should not be operated continuously above their rated load. Because air density is dependent on altitude, motors operating at high altitudes may not have sufficient air passing through them to prevent overheating. When motors are used at altitudes over 1100 m MSL, most suppliers recommend derating the load capacity to prevent overheating. It may also be necessary to derate motors when operating at high temperatures such as in agricultural fields during the summer. Consult the motor manufacturer’s literature for specific recommendations on how to compute the lower rated power when the motor is to be used at high altitudes or where air temperatures exceed 37°C. Whether elevated temperatures arise from overloading, lack of ventilation, or high ambient temperature, operation for significant periods of time at temperatures exceeding the design operating temperature will reduce motor life significantly. A common rule of thumb is that each 10°C above rated operating temperature will reduce motor life by 50%. Thus, motor temperature should be carefully controlled to obtain a long service life. 12.4.1.2 Motor efficiency. Manufacturers provide data on the operating efficiency of their product as a function of the load factor. Figure 12.16a is a typical motor efficiency curve which shows that the efficiency is zero at no load but rises steeply so that at 50% of rated load the efficiency is over 80%. When the load increases to 75% of the rated load, the efficiency is near its maximum and will stay at that level until it begins to drop slowly when the motor is loaded to 125% to 150% of its rated load. Thus, the efficiency is nearly constant over a fairly wide range of load. Smaller electric motors have lower operating efficiencies. The maximum efficiency of 7.5 kW or smaller motors is usually below 88%. The range in peak efficiencies for 75 kW or larger motors is from 90% to 92%. Unlike combustion engines, electric motors have few internal components to wear and reduce operating efficiencies. When motor bearings wear, heat is generated and the bearings generally seize, causing motor failure. It is safe to assume that if a motor runs at all, it operates at or near its peak efficiency if the load is between 75% and 125% of rated load.

.72 .66

80

100

120

Output Power, kW

Figure 12.16. Relation between load on power plant and (a) efficiency of electric motor or (b) fuel consumption per unit energy output of combustion engine.

Design and Operation of Farm Irrigation Systems

421

12.4.1.3 Electrical service to motors. Most electrical utilities require that motors larger than 7.5 to 18 kW (10 to 25 horsepower) be connected to three-phase service. Single phase motors draw six to ten times the run current when starting, as well as higher running current per conductor than three-phase motors. The lack of three-phase service or the prohibitive cost of adding the service may make electrical power infeasible for pumping in some areas. A new design of single-phase motor, however, has the potential to significantly relax restrictions on use of larger single-phase motors. The Written Phase motor runs 10% to 25% more efficiently than a conventional single phase, and because it only draws twice the operating current when starting, allows three to four times larger motors to operate on single phase lines. These motors are commercially available up to 45 kW, with larger sizes being designed. Proper selection of electrical conductors depends on the current requirement, voltage, location of service conductors, and the proximity of conductors to one another. Wire insulation is specifically designated for the type service, with types designed for direct burial in soil, for use specifically in conduit, and many special purpose types. The current capacity of aluminum conductors is only about 60% that of the same size copper conductor because aluminum has greater electrical resistance. If aluminum conductors are used, all connections must be certified for use with aluminum to avoid corrosion and thermal expansion problems which can lead to poor connections. Most electrical suppliers recommend allowing a line voltage drop after the meter of 5% or less. Many pump loads are sufficiently large that voltage drop cannot be accurately calculated simply using Ohm’s Law for direct current resistance. Particularly as conductor sizes increase, spacing between conductors increases, and power factors decrease, reactive losses can significantly exceed DC resistance losses. Several methods can be used to estimate reactive losses, including the Mershon diagram and drop factor tables (Croft and Summers, 1992). 12.4.1.4 Overall efficiency. Overall pumping plant efficiency, Epplant, for electrically powered plants is a product of the pump efficiency, Epump, and motor efficiency, Emotor: Epplant = EpumpEmotor = Q × TDH/C × Pm (12.14) where the other terms are described in Equations 12.1 and 12.2, except that Pm is the input power to the motor measured in kilowatts. This efficiency is commonly called wire-to-water efficiency. The maximum theoretical value of this overall efficiency ranges from 72% to 77%. For example, a pump with a maximum efficiency of 84% coupled with a 90% efficient motor has a maximum overall efficiency of 75.6%. Farmers are being advised that pumping plants should attain at least 65% efficiency, and every new pumping plant should be tested to determine efficiency. Some advisors suggest that if new pumping plants do not meet this level in field tests, then payment to pump installers should be withheld until the installation is corrected. Such conditions should be spelled out in a purchase contract. 12.4.1.5 Motor operating problems. Problems develop when motors are subjected to unbalanced voltages, high voltages, or low voltages. Any of these problems will result in motor overheating, reduced motor life, and in severe cases the motor will not operate. Generally, the electric company is responsible for correcting these problems. An exception is when wires connecting the power company’s main lines to the motor are too small or become damaged, resulting in excessive voltage drop and low voltage at the motor. Submersible turbine pumps are particularly susceptible to unbalanced

422

Chapter 12 Pumping Systems

voltages or voltage surges. The major reasons for motor failure include overheating, lightning strikes, destruction of insulation by rodents, and lack of proper lubrication. 12.4.1.6 Motor options. A number of different types of motor enclosures are available including open frame, weather protected, dust proof, and explosion proof. In addition, low-, medium-, and high-thrust motors are available, as are those with antireverse ratchets and other less important options. Generally, irrigation and drainage applications use weather-protected motors of medium to high thrust. Most motors used on turbine pumps also have an anti-reverse ratchet to prevent the lineshaft from turning backward when the motor stops. Motors which are not weather protected require construction of a pumphouse to prevent rain or other moisture from entering the motor causing a short circuit. A pump house must provide sufficient ventilation for motor cooling. At minimum, sun shields are recommended to prevent sunlight from directly striking the motor and control panels; this also provides some protection from rain or water from overhead sprinklers. Some large motors are available with a water cooling option. A small part of the pumped water is circulated through the motor casting, thus dissipating motor heat. Either the motor or control panel should have a thermal breaker, which stops the motor if it becomes overheated or draws current in excess of the nameplate rating for extended periods. 12.4.2 Internal Combustion Engines Internal combustion engines, and even steam engines, were used to power pumps as early as the 1920s. The annual Irrigation Survey, published in the Irrigation Journal (2001), documents the types of power used in the U.S. to pump irrigation water. In some states, more than 90% of the pumps are internal combustion engine driven. In others, such as California, Oregon, and Washington where electricity has been readily available, the pumps are almost all electrically driven. Natural gas, where available, was long a cheap source of power for pumping. Large numbers of pumps in Texas, Oklahoma, Kansas, Nebraska, Colorado, New Mexico, and Arizona relied on this source of power for a half century. Deregulation in the 1980s resulted in large increases in natural gas cost and reduced availability. Diesel, l-p gas, and, to a lesser extent, gasoline have been used as fuel sources in remote areas. Price structure has also made diesel competitive with electricity in some areas. Because they are convenient to transport, these liquid fuels are used to power most portable pumps. Direct-coupled, engine-powered centrifugal pumps have the pump impeller mounted on the engine shaft or a shaft coupled to the engine by a clutch. Most turbine pumps powered by engines use a right-angle gearhead and drive shaft to transmit power from the engine to the pump line shaft. 12.4.2.1 Engine selection. Many sizes, makes, and models of engines are available. Each has a set of operating characteristics which describe how it will perform over a wide range of conditions, and each has its most efficient zone of operation. Figure 12.16b illustrates how fuel consumption and output power are related to engine speed. Note that the engine is most efficient at a particular speed and power output. The engine in Figure 12.16, when operating at 1500 RPM, will operate most efficiently producing about 100 kW with a diesel fuel consumption of about 0.60 kg/kWh (0.36 lb/BP-h). Engine speed is controlled by the throttle setting and the load, and may be adjusted automatically by a governor. A detailed description of the process for matching an internal combustion engine and gearhead to a turbine pump is beyond the scope of this

Design and Operation of Farm Irrigation Systems

423

chapter. The necessary steps include (1) selecting the pump that will operate efficiently while producing the desired pumping rate against the system’s total dynamic head; (2) computing the required brake power to drive the pump using Equation 12.2 and then adding bearing losses; (3) consulting engine manufacturers’ catalogs to select an engine that will produce the desired driveshaft power while operating efficiently; (4) determining the pump speed and the speed of the selected engine at the most efficient point of operation; then (5) selecting a gearhead with gear ratios to couple the two units together with each turning at the proper speed. Both lightweight and heavy duty engines are available. Engines may also be classed as low, medium, and high compression. The performance, cost, efficiency, and expected life of these engines vary significantly. Lightweight industrial or automotive engines are usually least expensive, have a short lifetime, are low compression, and thus have low efficiencies. Heavy duty engines may initially cost three to five times as much as lightweight engines, but they have high compression ratios which permit more efficient burning of the fuel. This is particularly important when natural gas or lp gas are used as fuels, because these fuels have a very high octane rating and can utilize much higher compression ratios than automotive type engines. Heavy duty engines generally run much slower, and thus have a longer life. Heavy duty engines are designed to be overhauled repeatedly, but it is often economical to scrap a lightweight engine rather than overhaul it because components wear to the point that repair is more costly than buying a new lightweight engine. Manufacturers’ engine specifications are generally for operating conditions at sea level and with air temperature below 30°C. Engine performance must be derated for both altitude and temperature where field applications exceed those values. Engine derating is much more important than derating electric motors, because the density of air decreases rapidly with altitude and temperature, and engine output is directly proportional to the mass of oxygen in the air combined with fuel. To increase air density at high elevations and temperatures, turbochargers are often used. The manufacturer should be consulted for specific information on how each engine will perform. Care must be taken to assure that continuous operating curves are used for engine selection. Irrigation pumps may operate for extended periods, which requires derating curves which describe intermittent output, which are commonly available for light service engines. Continuous rated power output may be 15% to 20% below the rated intermittent output. 12.4.2.2 Engine efficiency. Engine efficiency is the ratio of the output power to the input power. The ordinate of Figure 12.16b shows the amount of fuel consumption expected as a function of power delivered. Thus, the low point on the horseshoe curve is the point of best efficiency at each speed. This figure illustrates that even when the engine is new there can be a wide range in efficiencies depending on the operating point. Ignition timing and fuel injection adjustments also significantly affect efficiency. Fuel consumption typically increases, driveshaft power decreases, and efficiency drops as engine components wear. Continued maintenance of engines is essential to maintain high efficiency. This includes changing spark plugs and air filters and regular oil and oil filter changes as well as cleaning fuel injectors. Engine efficiency is determined in the field by installing a torque meter in the driveshaft and measuring engine speed to determine power transmitted to the pump, while also measuring fuel consumption. Such measurements are needed to determine

424

Chapter 12 Pumping Systems

Table 12.3. Representative energy content of fuels. Fuel Approximate Energy Content Diesel 39,000 kJ/L (140,000 Btu/gal) L-p gas 26,300 kJ/L (94,500 Btu/gal) Gasoline 34,600 kJ/L (124,000 Btu/gal) (975,000 Btu/1000 ft3) Natural gas 36,300 kJ/m3 Electricity 3600 kJ/kWh (3410 Btu/kWh)

when equipment changes or engine overhauls should be undertaken. Although overall pumping plant efficiency can be determined without the torque meter, use of the torque meter permits separating pump and engine efficiencies. Pumping plant efficiency is a measure of the overall performance of an engine powered pumping plant. Pumping plant efficiency is the product of efficiencies of the individual components, which for a turbine pump is: Epplant = Epump × Egearhead × Edriveshaft × Eengine

(12.15)

and for a direct coupled centrifugal pump, Epplant = Epump × Eengine

(12.16)

The pumping plant efficiency may also be evaluated by dividing the energy transferred to the water by the input energy (IE) to the engine, Epplant = C ×Q ×TDH/IE

(12.17)

where C is a conversion coefficient, 35.30 when energy consumed by the engine, IE, is in kJ/h (33.46 when Q and TDH are in foot-pound units and IE is in Btu/h). For electrically powered pumps, C is 0.00981 for Q in L/s, TDH in m, and IE in kW (1.89 × 10-5 for Q in gal/min, TDH in ft). Liquid fuel consumption for diesel, butane, propane, and gasoline is measured in liters per hour (gallons per hour). Natural gas consumption is measured in cubic meters per hour (thousands of cubic feet per hour) at a standard temperature and pressure. The energy content of fuel varies from dealer to dealer and with time. Representative values are given in Table 12.3. 12.4.3 Wind, Solar, and Other Sources of Power In some countries, both animals and humans are used to power a variety of irrigation and drainage pumps. The importance of these types of pumps is recognized, but detailed descriptions of their use are beyond the scope of this chapter. Wind has been widely used as a direct power source to provide both domestic and livestock water, and in some places for irrigation and drainage. Wind power is gaining favor for electrical generation, and is becoming economically competitive with other methods. Most wind generation systems in the U.S. provide power to the electrical grid for general distribution. Studies by the U.S. Department of Agriculture near Amarillo, Texas, demonstrated that wind can provide a significant portion of the energy required to irrigate crops in that area, but a back-up power source is needed. This system uses a clutching mechanism to mechanically transfer wind energy to a vertical turbine pump when wind speeds are adequate. When wind energy is lacking, the clutching mechanism will shift the power source to an electric motor or combustion engine. During some periods of operation, energy is derived from both wind and the standby power supply.

Design and Operation of Farm Irrigation Systems

425

The sun’s energy has been harnessed to power irrigation pumps in several experimental installations in the U.S. Both photovoltaic cell arrays to charge batteries and parabolic reflector arrays to concentrate the sun’s energy have been studied. Standby power is still needed, but technical feasibility has been demonstrated. The costs associated with utilization of solar energy remain considerably greater than for other sources, except for small remote installations.

12.5 PUMP CONTROLS The subject of irrigation controls is covered in considerable detail in a companion publication by ASAE (Duke et al., 1990) on management of irrigation systems. Discussion here is confined to generalized controls for larger pumping plants, particularly those aspects that relate to protection of equipment and safety of personnel. 12.5.1 Safety Controls Because pumps often run unattended, combustion engines require automated protection against conditions that are most likely to damage the engine. Control units are available which either ground the ignition system (for spark ignition engines) or cut off the fuel supply (for compression ignition engines). Sensors connected to this control unit may include those to assure safe levels of coolant temperature, exhaust temperature, lubricant pressure, manifold pressure, engine speed, fluid levels, and vibration. Additional sensors may protect other components, such as pump lubricant sensors. Combustion engines provide many opportunities for personnel injury from contact with rotating parts. It is particularly important that drive shafts be shielded to prevent entanglement of clothing or hair, as such accidents are frequently fatal. Other areas needing shielding are cooling fans, accessory drive belts, and exhaust systems. Electric motors also require protection from damage. First, with all but the smallest installations, the equipment used to start and stop the electric motor must be power rated to equal or exceed the power rating of the motor(s) to be controlled. Motor current must be started and stopped with magnetically controlled contacts which make or break the circuit. In addition, motor controls must provide automatic emergency interruption of electrical service in case of short circuit or other severe overload. This emergency interruption is provided with fuses. Fuses must be sized to pass the starting current required for the motors served, but to fail if the current significantly exceeds that required for normal operation over an extended period. Because starting current is considerably greater than running current, fuses of size adequate to allow normal starting provide no protection against motor overload. However, because an overloaded motor draws higher current than normal, heater elements which activate thermal overload devices can provide protection against high current for longer than required to start the motor. Dual element fuses sized to 125% of motor current rating will provide short circuit protection and will also provide motor overload protection. Even more important than protection of motors is protection of personnel. Because the local environment of pumping plants is often wet, ground path distances may be large, and voltages of 240, 480, or even greater are common, the potential for injury or fatal accident is significant. A ground conductor must be provided from the meter point to the point of service. Because distances are frequently large, ground path resistance, even with a good ground rod, can be several ohms (enough to allow lethal current to pass through the operator). This ground conductor must not be fused, because a blown fuse here can leave the entire pump and control system at line voltage.

426

Chapter 12 Pumping Systems

Ground-fault circuit devices, which detect current leakage of a few milliamperes, are seeing use on certain irrigation equipment. However, the high currents often involved with electric powered pumps and other irrigation equipment sometimes induce sufficient capacitive coupling and other forms of leakage to cause nuisance tripping of the devices. Controls should be located so that the operator does not have to climb or kneel to operate them. Disconnects must be between 1 and 2 m from the ground or operator platform. Belt-driven pumps, common with older units, must have belt guards in place to prevent personnel contact with moving belts and pulleys. 12.5.2 Automatic and Integrated System Controls Electrically powered pumping plants are particularly adapted to automatic controls which can start or stop pumping plants. These controls present a particular hazard to service personnel who may not be aware of their existence or intended times of operation. Thus, it is especially important that provisions be made for temporary lockout devices to prevent power from being applied when servicing of these units is necessary. Automatic start devices allow the pump to be started by some other device. They are commonly used with self-propelled sprinklers to allow both sprinkler and pump to be controlled by the sprinkler panel. Auto-restart devices are sometimes installed to restore pump operation without intervention after a power outage. Time controls are very common on solid set sprinkler irrigation systems, such as those used for parks and golf courses. These controls allow preselection of irrigation timing and sequence, and may include sensors to initiate irrigation when soil water becomes depleted, or to cease irrigation if significant precipitation occurs. Time-of-day controls are sometimes used on large irrigation pumps to aid the electrical supplier in spreading the electrical demand more evenly over time. Typical uses are to encourage pumping during off-peak times (perhaps at night after industrial plants have closed) by switching from one meter billed at a low rate to a second meter billed at a higher rate during peak demand periods. More sophisticated peak load control systems utilize coded signals sent by radio or over transmission lines to addressable receivers at individual pumps, which can interrupt power to the pump whenever the power supplier needs to reduce demand. Most manufacturers of self-propelled irrigation systems provide computerized controls which allow very sophisticated monitoring and control of the irrigation systems and the pumps that serve them. These control systems utilize radio, telephone, cellular phone, and even satellite data transmission to control the pump and irrigation systems. They are capable of remote operation from virtually anywhere in the world, so systems equipped with this type control are especially subject to unexpected operation.

12.6 ECONOMIC CONSIDERATIONS The decision to install an irrigation or drainage system should include an economic analysis. The returns from increased crop sales, value of landscape, or drainage benefits drainage must be large enough to repay both investment and operating costs. Pumping cost includes both fixed and variable components. Fixed costs do not depend on the annual operating hours, and include such items as initial investment, taxes, insurance, and annual connect charges for electric or natural gas service. Variable or operating costs include those items which depend on the amount of pumping each year. Variable costs are dominated by energy cost, but include lubricants, labor, repair parts, and maintenance costs.

Design and Operation of Farm Irrigation Systems

427

When energy costs were low, annual investment costs were typically larger than variable costs in the U.S. Although pumping equipment costs have increased significantly over the past two decades, power cost has increased several fold. Because operating costs typically exceed amortized investment costs, more expensive pumping equipment may have significantly lower overall costs due to higher operating efficiencies. Power costs for electricity can be very complicated and typically depend on the amount of energy used, time of use, and the connected load. 12.6.1 Investment Costs The total investment cost can be obtained by simply adding costs of the individual components, but annualized investment cost is more informative. Annualized investment costs can be computed by several methods. A simple approach is to determine the average annual cost by dividing the cost of a particular item by the expected life. However, this method disregards the impact of interest on the investment. Computing the amortized cost, considering the expected life of the item and the interest rate, is a more appropriate method. The capital recovery factor, CRF, can be used to compute the annual amortized cost. CRF is calculated by CRF =

I × ( I + 1) n ( I + 1) n − 1

(12.18)

where I is the current annual interest rate (fraction), and n is the anticipated life of the equipment (years). Multiplying the CRF by the initial cost results in the annual amortized cost. The investment cost should be calculated on the actual installed cost of all durable equipment, which may include the well, pump, gearhead, motor or engine, drive shaft, and fuel tank, gas pipeline, or electric power lines. The total annualized cost of a pumping plant is the sum of the annualized costs of each component. The expected life of various components will vary depending on the amount of use, protection from weather provided, frequency of maintenance, and the quality of the original components. If the manufacturer does not suggest an expected life, then it must be estimated. The expected life for engines, gearheads, and pumps may be expressed as hours of operation. For example, automotive engines may have an expected life as short as 3000 hours while some heavy duty engines are expected to operate for 15,000 to 20,000 hours with periodic overhaul. Table 12.4 shows life expectancy for various pumping plant components. Quality of the water being pumped can signifiTable 12.4 Estimated useful life of pumping plant components (years). Annual Hours of Operation Component 500 1000 2000 3000 Well 25 25 25 25 Pump 15 15 15 10 Gearhead 15 15 15 10 Drive shaft 15 15 7 5 Engine (heavy duty) 15 15 10 7 Engine (automotive) 5 3 2 1 Gas pipeline 25 25 25 25 Engine foundation 25 25 25 25 Electric motor 25 25 25 25 Electric controls and wiring 25 25 25 25

428

Chapter 12 Pumping Systems

cantly impact pump and well life by corrosion of ferrous components or deposition on well screens and pump impellers, even when unused. Annualized cost may also be calculated using a graduated depreciation schedule. This method is often used to accelerate credit of investment costs for tax purposes. 12.6.2 Other Fixed Costs Taxes, insurance, and power demand charges are paid annually and are independent of pump operation. The taxes are generally assessed on the value of the equipment, but may also be based on the volume of water pumped, which is a variable cost. Liability insurance and insurance against fire, vandalism, and acts of nature (e.g., lightning), protect the owner. Both electric and gas utilities may charge for the privilege of connecting. These may be billed as a one-time hookup fee (which should be annualized), divided over a minimal recovery period, or as annual connect or demand charges. 12.6.3 Variable Costs Power costs usually dominate annual operating or variable costs. Annual power costs can be determined by estimating the energy consumption per hour of operation (kW for electricity, m3/h for natural gas, or L/h for liquid fuels). Energy demand can be estimated from one of Equations 12.15 through 12.17. This method requires that the pumping rate, total dynamic head, and estimated pumping plant efficiency be determined. Typical ranges of pumping plant efficiency are given in Table 12.5. Multiplying the hourly consumption by estimated annual hours of operation and by unit cost of energy gives the annual cost. Natural gas and electric computations may require application of a complex rate structure rather than a single unit price. Other variable costs, such as lubrication and labor, can be estimated as a cost per hour of operation. Only the labor required to service the pumping plant is included. Determine the manufacturer’s recommended lubrication and maintenance interval, then multiply by the per unit cost and the number of maintenance periods annually. Costs for minor repairs and overhauls should be estimated as a function of hours of operation. Engine overhaul costs are generally prorated over the years between the overhauls. Table 12.5. Representative pumping plant efficiencies. Power Source Maximum Acceptable Average Observed Electric 72–77 65 45–55 Diesel 20–25 18 13–15 Natural gas 18–24 15–18 9–13 L-P gas 18–24 15–18 9–13 Gasoline 18–23 14–16 9–12

12.6.4 Total Pumping Costs Total pumping costs can be expressed either as annual cost or as total life cycle cost. The annual cost is the sum of the annualized investment cost, other annual fixed costs, and variable costs for a year. Life cycle costs are the sum of the initial investment costs and the other fixed and operating costs for each year, including interest on the money invested. Whatever the method of estimating pumping costs, it is important that pumping plant selection be made on the basis of total cost, not solely on initial purchase price. 12.6.5 Minimizing Pumping Costs The cost of pumping is dependent not only on the input power requirement of the pumping plant, but also on the total amount of water pumped, and the costs associated

Design and Operation of Farm Irrigation Systems

429

with supplying energy for pumping. Input power requirements have been shown to depend on the efficiency of the power plant, drive mechanism, and pump, and on the pumping lift, pumping rate, friction losses in the delivery system, and operating pressure head. Therefore, optimizing these components will minimize pumping costs. Pumping lift depends on the static water level from which the pump operates, and for wells, on drawdown within the well. Although these factors are often beyond the farmer’s control, proper development of a well and appropriate well screen when installed can reduce the resistance to water flow into the well bore, reducing drawdown. Periodic measurement of static lift will determine whether changes are occurring in regional water levels, which may eventually require modification to the pump. Measuring pumping lift will help detect changes, such as incrustation of well screens or clogging by microorganisms that can be remediated by chemical or mechanical treatment to reduce drawdown. Pumping rate can be reduced in some circumstances. A reduced rate will allow a smaller drive motor to be used, which may reduce energy demand or hookup charges. Reducing pumping rate may require pumping for a longer period, which may allow the irrigator to take advantage of the reduced power rates characteristic of many rate schedules which are based on energy consumption per unit of power connected. Before committing to pumping plant changes to reduce pumping rates, the farmer must realize that such changes may increase the risk of crop damage from inadequate irrigation or insufficient drainage. Reduced pumping rates also make the reliability of the pumping plant a more critical factor, as less time is available for preventive maintenance or repairs before crop stress begins. Friction losses can be minimized by initial selection of larger diameter pump column pipes and delivery piping. Minimizing pipe fittings such as elbows, valves, and abrupt changes in pipe diameter, will keep friction loss at a minimum. Use of smooth materials, such as polyvinyl chloride (PVC) or epoxy-coated pipes can significantly reduce friction loss. Periodic interior scaling and coating to restore interior dimension and smooth the surface of aging pipe can be a cost-effective maintenance procedure, particularly for large pipelines. Total head is the remaining factor of power costs that can be controlled. Any valving of the discharge piping system simply reduces the pressure at the point of discharge, but increases the head against which the pump operates. Therefore, reducing pressure or flow rate by valving is not an economically viable alternative. Sprinkler and microirrigation system manufacturers have developed many alternatives in recent years to allow use of lower pressures. Low angle trajectory impact sprinkler heads with unique orifice designs operate at lower pressures than conventional impact heads. These heads reduce the wind distortion of sprinkler patterns and minimize soil compaction problems associated with large water droplets. However, the reduced diameter of the spray pattern increases the water application rate and may result in water runoff and poor uniformity. Spray heads have largely replaced impact heads on self-propelled sprinkler irrigation systems. The first of these spray heads had very small pattern diameters, with high application rates and fine water droplets subject to considerable wind drift. New spray head designs increase the water pattern diameter, produce moderate droplet sizes, and function at very low pressure. With careful attention to soil characteristics, use of low pressure sprinkler heads or microirrigation emitters can significantly reduce total head and possibly pumping

430

Chapter 12 Pumping Systems

costs. Low pressure devices are most effective where elevation differences across the irrigated area are small because the flow rate is more influenced by changes in elevation than for devices operating at higher pressure (see Chapter 16). Poor uniformity resulting from low-pressure operation can increase the amount of water that must be pumped to adequately irrigate a field. The cost of this increased pumping can easily exceed savings realized by reducing pressure. Operating head of irrigation systems requiring filtration can also be minimized by careful maintenance of filters. Frequent cleaning of screens or back-flushing of filters will reduce the head loss associated with moving water through the devices. When changes are made that impact the total dynamic head of a pumping plant, such as larger piping, low pressure sprinkler heads, or more efficient filtration systems, the pumping plant must be reevaluated to determine whether these changes warrant pump modifications. One of the most effective ways to reduce pumping costs is to reduce the amount of water pumped. Low water application uniformity requires pumping more water than required by the crop to provide adequate water over all the field. Improvements to the irrigation system, such as adaptation of surge irrigation or level basins for surface irrigation or renozzling a center pivot to attain greater uniformity can significantly reduce the necessary pumping (Chapters 14, 16, and 17). Avoiding excess irrigation by applying only the amount needed at the time needed (Chapter 22) will directly reduce pumping costs. Other management decisions may impact total energy charges, if not total energy consumption. Where electrical demand charges are based on motor nameplate power, use of undersized motors may reduce energy costs, but at the risk of shortened motor life. Rate structures are often stepped, which tends to favor reduced pumping rates (smaller motors) which in turn may require increased hours of pumping. Where electrical demand charges vary with time, it may be advantageous to increase pumping capacity to allow the pump to be off during times of high cost energy or to participate in shutdown programs to control the supplier’s peak demand.

12.7 IMPACT OF PUMPING SYSTEM MODIFICATION Any modification to the irrigation or pumping system, such as declining groundwater levels or converting from open ditch to gated pipe delivery, will change the system head curve. In Section 12.3.3 I showed the relation between the system head curve and the pump TDH-Q curve and how the two determine the operating head and pumping rate. Thus, changes in the irrigation system require evaluating their impact on pumping plant performance. 12.7.1 Changing Operating Conditions Many factors can change the operating conditions of a pump. The system head curve may be changed intentionally as irrigators seek new methods to contain costs. Areas historically surface irrigated have been converted to sprinkler or trickle irrigation to increase the efficiency of irrigation. Sprinkler heads may be replaced by models that operate at lower pressures to reduce total heads. Other changes to the system head may be unintentional, such as deterioration of pipe walls, which may increase friction and even reduce pipe internal diameters. Severe cases may result in rupture of pipes resulting in loss of part of the water pumped or recirculation of water within a well casing. Each of these conditions will change the total head of the pump, thus shifting the operating point along the pump curve, with increased head generally reducing pumping rates. Whatever the direction of change, the operator should be aware

Design and Operation of Farm Irrigation Systems

431

of the potential impact of changes on the efficiency of the pumping plant. Any significant change to an irrigation system must include evaluation of the impact of that change on pumping costs. Other changes may be beyond the direct control of the irrigator. Declining water tables will likewise increase the system head curve. Sediment, whether from improper well development, deterioration of well screens, or pumping of sediment-laden surface waters, will accelerate wear of pump components, producing an effect similar to reducing impeller diameter, thereby reducing both operating pressure and pumping rate. If water levels decline, well screens become encrusted, or other conditions cause the water level at the pump suction to decline, air may be introduced directly into the pump, a condition referred to as breaking suction. This will reduce the pumping rate, which may allow the water level to recover and the pump to begin operating again. With centrifugal pumps having a positive suction lift, breaking suction will often cause pumping to cease until the operator intervenes to evacuate the air from the suction side (i.e., prime the pump). Turbine and submersible pumps may recover from this condition, but such surging of the flow can severely damage the pumping plant. Declining water levels may also result in cavitation, which can quickly destroy a pump. Similar conditions are created when water cascades into a well bore or sump. Air bubbles may be entrained with the water, and these bubbles expand and collapse rapidly as pressures change within the pump impeller. 12.7.2 Pump Modifications Whenever pumping conditions change sufficiently to warrant modification to the pumping plant, a number of options may exist. Pumps powered by combustion engines can sometimes be adjusted to account for minor changes in operating conditions by simply changing the engine throttle setting. In fact, deterioration in combustion powered pumps often goes unnoticed until severe because the engine throttle can be adjusted to give the desired pumping rate and pressure. In Section 12.2.3 I described the relation between the pump head-pumping rate curve and pump speed. It is important to keep in mind that changing pump speed not only changes the pumping rate and discharge pressure, but it also changes the power requirement. As seen from Equation 12.5, the input power changes as the third power of pump speed. Thus, increasing pump speed by only 10% will increase the power requirement by one third. Additionally, significant changes in pump speed will undoubtedly change the efficiency of the pump, and may result in significantly greater impact on power requirement than indicated by Equation 12.5. The overload resulting from increasing pump speed may well lead to severely shortened engine life, and the need to increase throttle setting should be a signal to the irrigator to evaluate the pumping plant. Those electrically powered pumps that utilize belt drives may sometimes be modified in a similar manner by changing pulley sizes to change pump speed, but the electrical load must be carefully checked to assure against overheating the motor. In general, significant changes to the pumping or irrigation systems will necessitate some modification to the pump to assure peak performance. The amortized cost of making such changes must be compared with the consequences of leaving the pumping plant unmodified. Pump modification may be as minor as adjustment of the bowls to compensate for wear, as discussed in the next section. However, if significant changes have been made to the pumping or irrigation system, more extensive pump modification may be warranted.

432

Chapter 12 Pumping Systems

A pump test, described in the following section, will determine whether pump repairs are needed. If the system head has been reduced, such as through conversion of a center pivot from a high-pressure to lower-pressure sprinkler package, and the pump is in otherwise good condition, it may be possible to modify the existing pump so that it will work efficiently under the new conditions. Such modification requires evaluation of the pump to determine the suitability of the existing bowl for the new application. If the pump can be matched to the system head curve with reasonable efficiency by either removing bowls, trimming impellers, or a combination, that may be an economical alternative. A decision to modify an existing pump for new conditions should always be accompanied by a comparison of repair or replacement costs with long-term power costs.

12.8 MAINTENANCE AND TESTING 12.8.1 Routine Maintenance As with any mechanical equipment, a good pumping plant maintenance program is necessary to maintain pump efficiency, minimize power costs, improve dependability, reduce operating costs, and extend the life of the equipment. The major item of pumping plant maintenance is proper lubrication of both the pump and drive unit. The line shaft of a turbine pump must be lubricated, by water or oil as designed, prior to starting. Pump head bearings and electric motor bearings must be greased or oiled as the manufacturer recommends. Other maintenance items include periodic adjustment of impeller clearance in turbine pumps and adjustment of packing glands or pump shaft seals. Engine oil, coolant, and filters must be changed regularly. Regular ignition and fuel system adjustments will help maintain engine efficiency. The motor or engine, pump head, and gearhead should be cleaned periodically to help dissipate heat. Any unusual noise or vibration is cause for immediate action. Pumps must be drained after the irrigation season in cold regions or else be enclosed in heated shelters if they must operate during periods when freezing is possible. Protective screens must be installed on electric motors to prevent rodents from damaging windings and these screens must be kept clean to permit air circulation. 12.8.2 Pump Adjustments The primary adjustment to be made on centrifugal pumps is to periodically tighten the packing gland that seals the input shaft. Although excessive leakage of water from this seal may cause inconvenience, it should be remembered that water leakage provides the only seal lubrication for many pumps. In any case, the seal must never be tightened so that the shaft is difficult to rotate by hand. Vertical turbine pumps also have a line shaft seal at the point the shaft exits the pump head. In addition, the vertical turbine pump allows for adjustment of the clearance between impellers and bowls. For open and semi-open impeller types, adjustment of this clearance is critical to attaining the maximum pump efficiency. The line shaft is threaded on the top, with a key to assure that the line shaft rotates with the motor or gearhead. The line shaft nut has a setscrew or similar locking device to hold the adjustment once it is set. The line shaft nut is backed off until the impeller just drags on the bowl and the line shaft cannot be turned by hand. Then the nut is tightened to lift the impeller from the bowl. The number of turns will depend upon the pump setting, line shaft diameter, pump thrust characteristics, and total head, all of which influence the amount of line shaft stretch when operating. Consult the manufacturer’s instructions for the appropriate adjustment. In any case, the line shaft nut should not be tightened so far that the impeller drags on the top of the bowl.

Design and Operation of Farm Irrigation Systems

433

Adjustment for installations with electric motors can be checked using an ammeter. If the pump is operable prior to adjustment, use a clamp-on ammeter to determine the current in each of the conductors. After adjustment, again measure the current in each conductor. Although the current may change because pump performance is affected, a large change in current should arouse suspicion that the impellers are dragging, and the pump should be stopped immediately for professional service. Dragging impellers may also cause unusual noise, vibration, or obvious overloading of an engine. Some farmers with open or semi-open impeller pumps are inclined to increase impeller clearance to reduce the pumping rate and discharge head when the irrigation or drainage system is modified. Although such adjustments may effectively change pumping rate and discharge head, the efficiency of the pump is invariably affected. If operation is to continue with increased impeller clearance, a pumping plant evaluation should be conducted to determine whether the efficiency has been sufficiently changed to warrant pump modification. 12.8.3 Evaluating Pumping Plant Performance Field testing of both new and old pumping plants is recommended to evaluate performance. New plants should be tested to confirm that the pump is operating as expected, and to rectify problems while the equipment is under warranty. The pump installer should be responsible to insure that the pump performs as designed. Every pumping plant should be performance tested after several years to develop a history of performance change. These tests will identify pumps with low efficiency or those producing less flow or head than expected. Updated efficiency values should be used to conduct a cost analysis to determine whether the pumping plant should be repaired. A field test should determine the static lift, pumping rate, drawdown, discharge head, and input power. Using a torque meter on engine-driven plants will allow separate evaluation of engine and pump performance. As mentioned earlier, the efficiency of electric motors can be assumed to remain constant so long as they run without excessive noise or vibration. Equipment required to conduct a field test includes an accurate water flow meter, a stopwatch, a device to measure the water level if pumping from a well, accurate pressure gauges when the pump discharge is pressurized, a fuel flow meter and torque meter for engine driven units, and current and voltage meters for electrical units. Because this equipment is expensive and must be well-maintained and regularly calibrated, few farmers can justify their own. In most areas, pumping plant tests are available for nominal cost from consultants, pump installers, gas suppliers, or electrical utility companies. 12.8.4 Making Repair Decisions Many repair decisions are not so apparent as those to replace broken parts that disable the pumping plant. Gradual deterioration of the pumping plant may not be noticeable until quite severe. At this point, the cost of pumping a unit of water may be several times what it could be under optimum conditions. For this reason, the periodic pumping plant test is a good economic investment. Following the pumping plant test, one must make the same calculations of operational cost as were made to select the proper components on installation. Comparing the present cost of operation with the cost of a high-performance pumping plant, and the potential savings with the cost of making necessary repairs, is the most realistic method of deciding whether to repair.

434

Chapter 12 Pumping Systems

Acknowledgements The author wishes to express appreciation to Mr. Robert A. Longenbaugh, Consultant Water Engineer, retired from the Office of the Colorado State Engineer of Arvada, Colorado. Mr. Longenbaugh was senior author of this chapter in the first edition, and much of the current chapter is taken from that edition. Mr. Quintin E. Nesbitt, Agricultural Engineer, Idaho Power Company, Boise, Idaho, provided materials from which the information on variable speed pumps was prepared.

LIST OF SYMBOLS BP C CRF D Dt E Ea Edriveshaft Eengine Egearhead Emotor Epplant Epar Epump Eser hf Hp I NPSH NPSHA NPSHR RPM TDH WP n Q Qg Qn Pb Hs Pv IE Pm

brake input power to pump, kW coefficient to convert energy units, capitol recovery factor pump impeller diameter, m downtime as a decimal fraction of pumping period elevation above mean sea level, m application efficiency, decimal fraction driveshaft efficiency, decimal fraction engine efficiency. decimal fraction gearhead efficiency, decimal fraction motor efficiency, decimal fraction overall pumping plant efficiency, efficiency of pumps in parallel, pump efficiency efficiency of pumps in series friction head head imparted by pump interest rate net positive suction head, m net positive suction head available, m net positive suction head required by pump, m pump speed, revolutions per minute total dynamic head, m water power delivered by pump, kW anticipated life of equipment, years pumping rate, L/s gross pumping capacity, ,L/s net pumping capacity, L/s barometric pressure, m pumping suction lift, m vapor pressure of pumped fluid, m input energy to engine or motor, kJ/hr or kW input power to motor, kW

REFERENCES Addison, H. 1966. Centrifugal and Other Rotodynamic Pumps. London, U.K.: Chapman and Hall. Colt Industries. 1979. Hydraulic Handbook. Kansas City, Kans.: Colt IndustriesFairbanks Morse Pump Division.

Design and Operation of Farm Irrigation Systems

435

Croft, T., and W. I. Summers. 1992. American Electrician’s Handbook. 12th ed. New York, N.Y.: McGraw-Hill, Inc. DeWiest, R. J., and S. N. Davis. 1966. Hydrogeology. John Wiley and Sons, Inc. Duke, H. R., L. E. Stetson, and N. C. Ciancaglini. 1990. Chapter 9: Irrigation system controls. In Management of Farm Irrigation Systems, 265-312. G. J. Hoffman, T. A. Howell, and K. H. Solomon, eds. St. Joseph, Mich.: ASAE. Hanson, B., C. Weigand, and S. Orloff. 1996. Performance of electric irrigation pumping plants using variable frequency drives. J. Irrig. Drain. Div. 122(IR3): 179-182. Hicks, T. G. 1957. Pump Selection and Application. New York, N.Y.: McGraw-Hill. Heermann, D. F., H. H. Shull, and R. H. Michelson. 1974. Center pivot design capacities in eastern Colorado. J. Irrig. Drain. Div. 100(IR2): 127-141. Ingersoll-Rand. 1977. Cameron Hydraulic Data. Woodsliff Lake, N.J.: Ingersoll-Rand. Irrigation Journal. 2001. Irrigation survey, 2000. Irrig. J. Jan/Feb. Johnson Division UOP. 1975. Groundwater and Wells. St. Paul, Minn.: Johnson Div.. UOP. Karassik, I. J., W. C. Krutzsch, W. H. Fraser, and J. P. Messina. 1976. Pump Handbook. New York, N.Y.: McGraw-Hill, Inc. McWhorter, D. B., and D. K. Sunada. 1977. Groundwater Hydrology and Hydraulics. Highlands Ranch, Colo.: Water Resources Publications Nesbitt, Q. 1995. Variable frequency drives and irrigation pumps. In USDA-NRCS Western States Irrigation Workshop. Washington, D.C.: USDA-NRCS. Poole, J. N. 1993. Economics of adjustable speed drives. In Motor/Adjustable Speed Drive Users Workshop, Northwest Power Quality Service Center, Electric Power Research Institute. Palo Alto, Calif.: EPRI. Poole, J. N., D. L. Seitzinger, and G. R. Stengl. 1989. Adjustable speed drives analysis (ASCON-I) energy savings and return on investment. Raleigh, N.C.: CRS Sirrine, Inc., User’s Manual for Electric Power Research Institute, Palo Alto, Calif. Todd, D. K., and L. W. Mays. 2004. Groundwater Hydrology. John Wiley & Sons, Inc. Zipparro, V. J., and H. Hasen. 1993. Davis’ Handbook of Applied Hydraulics. 4th ed. New York, N.Y.: McGraw-Hill, Inc.

CHAPTER 13

HYDRAULICS OF SURFACE SYSTEMS Theodor S. Strelkoff (USDA-ARS, Maricopa, Arizona) Albert J. Clemmens (USDA-ARS, Maricopa, Arizona) Abstract. The goals of surface irrigation are defined and the physical processes involved are described. Field variables such as slope, infiltration, and roughness; physical design variables like furrow or border cross-section and length of run; and management variables such as inflow and outflow control are discussed. The role of modeling as a tool for subsequent design and management is emphasized, as is the pivotal role of physical laws such as conservation of mass and momentum. Hydrologic, or lumped-parameter, modeling is described in detail as a relatively simple technique sufficiently accurate for many purposes. The errors that do arise are explained as a consequence of some of the assumptions made to achieve the simplification. The hydrodynamic, or distributed-parameter, approach, itself amenable to a series of simplifications, is described with sufficient detail to afford a basic understanding of the assumptions and solution techniques. A user-friendly software package allowing easy data entry and simulation graphical output is introduced. The chapter closes with a discussion of techniques for estimating infiltration and roughness in the field for entry into simulation and design software. Keywords. Efficiency, Hydrodynamics, Modeling, Parameter estimation, Simulation, Surface irrigation, Volume balance.

13.1 INTRODUCTION 13.1.1 History and Current Status of Surface Irrigation Surface irrigation is accomplished by introducing water at some point in a field and allowing it to flow over the soil surface to reach the receiving plants. It has been practiced in many areas of the world since ancient times. At least 4000 years ago, farmers in Egypt, China, India, and countries of the Middle East irrigated agricultural lands. As early as 200 B.C. the Hohokam farmers diverted water from what is now known as the Salt River through 200 km of hand-dug canals in and around the site of modern Phoenix, Arizona (Salt River Project, 1960). The long history, marked by successes and failures, is valuable. Evidences of sustained agriculture in arid lands illustrate the contribution that successful irrigation can make to society. Evidences of salted soils and abandoned systems emphasize the need to irrigate wisely or face similar consequences.

Design and Operation of Farm Irrigation Systems

437

Shortly after the close of World War II, technology, energy, and industrial support combined to promote a rapid rise in sprinkling; this was followed by micro- and LEPA systems. Still, in spite of the popularity of pressurized systems, the USGS estimates that of the roughly 25 million ha irrigated in the U.S., 48% is served by surface means (Hutson et al., 2004). The percentage increases to well over 90% in other parts of the world. It is in most cases the cheapest method of applying water to crops, and usually requires the least expenditure of energy, but it is often viewed as inherently wasteful of water. It is also considered more labor intensive and less convenient than pressurized systems. Yet, under appropriate field conditions, proper design and management of surface systems can lead to high application efficiencies, on a par with other methods. There is potential for automation and reducing labor requirements. Furthermore, in many areas of the world, labor is relatively inexpensive, while availability of water and capital are low. 13.1.2 Irrigation Scheduling vs. Uniformity of Application A key aspect of efficient irrigation is application of water to the soil in the amounts needed at the time needed, all over the irrigated field. Determination of the plant water requirements is examined at length in Chapter 8. Still, once the requirements and timing are established, actually placing that amount at the plant roots is not a trivial matter and is the subject of hydraulic studies of the on-farm system that distributes water over the cropped area. It is the variation of pressures in a pressurized system or water surface elevations in a surface system, in response to and influencing the flows prevailing at the various locations in the system, that determines the uniformity of distribution of water to the individual plants. With a complete understanding of the system hydraulics and the ability to influence it, rational decisions can be made balancing system costs against the monetary and other costs of local under- and overirrigation. 13.1.3 The Role of Simulation in Design and Operation Design of surface systems has traditionally been based on experience with successful projects, intuition, empirical compilations, and not very much theory. Now, however, many of the analytical problems associated with the prediction of surface irrigation flows have been solved, and it is now possible to develop designs and operating recommendations based on simulations, i.e., solutions of the pertinent hydraulic equations with trial values of the design parameters in the search for an optimum. These equations express basic laws of physics: conservation of mass and momentum. In view of our limited ability to predict theoretically the interaction of the soil and plants in the system with the hydraulics of the irrigation stream, some empirical input is still necessary. However, by limiting the empirical aspects of the study, we can hope for greater universality of the results than is possible with more purely empirical attempts at predicting stream behavior or irrigation performance. 13.1.4 Scope of the Chapter This chapter provides an overview of the behavior of irrigation streams as influenced by soil and crop hydraulic parameters, physical design parameters, and irrigation management parameters. These determine the outcome and merit of the irrigation. The focus is on simulations of surface irrigation flow for use in development of design and management recommendations. Simulations are viewed from both a hydrologic and hydrodynamic point of view. The hydrologic, or lumped, approach makes an assumption for the shape of the irrigation-stream profile. One or another approximation of momentum conservation estab-

438

Chapter 13 Hydraulics of Surface Systems

lishes the depth at some point in the stream, typically the upstream end. The principle of mass conservation then leads to ordinary differential equations (or an equivalent convolution integral) in stream length as a function of time. The hydrodynamic, or distributed, approach makes no shape assumptions utilizing, instead, the principles of mass and momentum conservation to solve for the variation of depth and discharge within the stream. This leads to partial differential equations in distance and time, or their equivalent. The equations governing the flow are presented, along with initial and boundary conditions. Simulation software based on numerical solution of the equations is described. The governing equations are also applied to an inverse problem: estimation of field parameters from observed behavior of an irrigation stream. When needed, surface irrigation hydraulics are presented in dimensionless form to allow a maximum amount of information to be transmitted with the least computational effort and volume of data; databases of simulations for design are typically calculated and stored dimensionless. The goal is the development of surface irrigation systems that can be operated in an efficient, effective, sustainable, and environmentally sound way.

13.2 BASIC CONCEPTS OF SURFACE IRRIGATION HYDRAULICS 13.2.1 Characteristics of Surface Irrigation In surface irrigation, water is conveyed to the upper, or inlet, end of a horizontal or sloping field and is released there to flow over the field by gravity. An obvious, overriding characteristic of surface irrigation is that the same flow channels which convey water to their distal ends are also used all along the way to draw off water for plant needs. Another clear characteristic is that the cross-section of the flow is not constrained by fixed boundaries such as a pipe wall; the depth of flow is free to seek its own level. In a pipe, even if it is not full at the time irrigation starts, water moves quickly to the end, and the time of operation of a lateral port at the distal end is much the same as at the near end. But in surface irrigation, water typically travels much more slowly, sometimes requiring hours to traverse a single furrow. Irrigation streams, when confined by furrows, flow in specific, known directions. The transverse dimensions of such streams are much smaller than their longitudinal extents, and transverse variations in water-surface elevation and transverse components of velocity are not significant. Furthermore, in well-constructed basins or border strips, cross slope is negligible, and once flow is introduced uniformly across the head end, it advances in a sheet with little transverse variation. In these cases a onedimensional analysis of the flow is adequate. Hydraulic variables like flow velocity and depth can be viewed as functions of distance along the channel and time, i.e., in one space dimension. However, in some surface irrigation systems, the conditions for an adequate onedimensional representation are not met. It is customary in large level basins, for example, to introduce water from essentially a point source, at a corner or in the middle of a side. The unconfined flow spreads out from the source in all possible directions. No matter what direction is selected for a one-dimensional simulation, the results can be considered only a very rough approximation; a two-dimensional analysis is required. Furthermore, in an imperfectly leveled or graded field, unless the stream size is so

Design and Operation of Farm Irrigation Systems

439

large that water depths are much greater than the variations in the field surface elevations, water will concentrate in the low-lying areas, both in advance and during recession, with consequent reductions in uniformity of application over the field. Twodimensional simulations can predict the resulting distribution of infiltrated water. Whereas in pressurized flow, the relationship between line pressure and lateral outflow through a nozzle or emitter is relatively well known, the prediction of lateral outflow by soil infiltration is far more poorly grounded. Though the theory of infiltration is well established (Chapter 6), the results are dependent on the physical and chemical characteristics of the soil, including the geometry of the soil particles and their configuration in the matrix and the ambient water content within. Furthermore, in furrows, a varying depth of flow can influence the wetted area through which infiltration into the surrounding soil is effected. Characterizing all of these factors as they vary from place to place along the irrigation stream can be a formidable task, and major simplifying assumptions are commonly made. In pressurized lines, resistance to flow is afforded by smooth or slightly rough walls and is relatively easy to characterize. The relationships between flow rate and pressure drop in a pipe, while empirical, are well known and measurable with good repeatability. In surface irrigation, on the other hand, resistance to flow can be exerted by bare soils, sometimes smoothened by previous irrigations, or by clods which are large compared to the depth of the stream, by mulch at the bottom of furrows, or by plant stems and leaves growing through the irrigation stream. The form drag exhibited by these elements is difficult to characterize, even if their shapes, sizes, and density of occurrence were known. As in the case of infiltration, major simplifying assumptions are commonly introduced into the characterization of flow resistance in the simulation equations. 13.2.2 The Surface Irrigation Process A complete irrigation takes place in phases, starting with an advance phase, in which the water moves out over the field from the point of entry. Advance can slow and stop spontaneously before completion, at some intermediate point along its run. This condition will occur if the intake rate of the soil is too great for the stream size or duration. Otherwise, advance ends when the stream front encounters the field boundaries. In a basin, runoff is prevented by a berm, and water ponds behind it; otherwise, flow runs off at the end. After advance is complete, during a soaking phase, inflow continues until sufficient water has been introduced to satisfy the requirement in the field. After inflow cutoff, in the depletion phase, the surface stream is reduced in depth as infiltration and, possibly, runoff continues. The soil surface below the stream is eventually exposed as the recession phase begins, usually from the upstream end. With a sufficiently small slope, recession can start from the downstream end; the stream can also contract from both ends. The time interval between stream arrival and departure at a point is known as the infiltration-opportunity time there. Advance can be characterized as aggressive if the profile of the stream ends in a blunt front advancing relatively rapidly, with only gradual reduction in velocity. This is often true as the stream starts out. From the standpoint of simulation, this is a wellposed mathematical problem: small changes in any of the postulated conditions lead to small changes in calculated advance. But as more and more of the flow is drawn off by infiltration—as the length of the stream increases, especially with small stream sizes and relatively permeable soils—the profile of the front tends more to a feathered edge,

440

Chapter 13 Hydraulics of Surface Systems

advancing very slowly (at perhaps 0.2 m/min or less) or in starts and stops. This regime is badly posed: small changes in inflow, infiltration characteristics, soil-surface slope, or roughness can produce large differences in the time to advance to a given point. Simulation of this condition is not very robust1, with the possibility of spurious calculation of negative depths and premature front-end recession. Recession can also physically occur at the leading edge of the water stream, either prior to completion of advance or after. Any reduction in inflow rate in the course of system operation can lead to front-end recession. While this occurs most often after cutoff in horizontal or nearly horizontal fields, it can also arise in sloping fields as well, and before inflow is cut off. It is only necessary that the inflow rate, Q0, be exceeded by the rate of growth of the total infiltrated volume, Vz, i.e., dVz/dt be so much greater than Q0 that the disparity cannot be made up by decreases in surface-water depth alone. In theory, an infiltration rate increasing with wetting time can also cause such behavior. Typically, infiltration rates decrease with time, or remain constant; however, increases have been documented (Trout and Johnson, 1989). A more common cause for front-end recession before cutoff is cutback operation of a furrow system. In this mode of operation, in order to prevent excessive runoff from the end of a field, the inflow is reduced when the streams reach field end. If the remaining inflow is insufficient to sustain infiltration over the entire length of the stream, not only will runoff halt, but the stream will recede from the field end, the length contracting until the inflow rate balances the total rate of infiltration from the furrow. If infiltration rates decrease sufficiently prior to cutoff, the stream eventually begins again to advance. In theory, if the infiltration rate of the soil and inflow rate to the border or furrow were constant, advance would continue, i.e., stream length would increase until the total infiltration along the length just matched the inflow rate; then advance would halt, and stream length would remain constant until conditions changed. When inflow is cut off before the stream has reached the end of the field, advance usually continues for a time, if the stream is deep enough, the slope great enough, and the infiltration, roughness, and vegetative drag small enough. This is a common operating condition for high-efficiency irrigation in level basins and in sloping border strips, where it reduces the amount of runoff. But if front-end recession is initiated by premature stream cutoff, the leading edge recedes upstream, while at the same time, if the field slope is sufficient, the trailing edge recedes from the inlet end, until all surface water has disappeared into the soil. Unless a series of surges is planned, each advancing farther than its predecessor, this is clearly not a normal operating condition, for a portion of the field is left entirely dry, and other areas receive very little infiltrated water. Appropriate mathematical models of the surface irrigation process, such as SIRMOD (ISED, 1989) and SRFR (Strelkoff et al., 1998), can predict the onset of front-end recession. The entire phenomenon represents a case of unsteady, nonuniform, spatially varied flow over a porous bed, and is similar to other cases of overland flow. As such, it can be described either through a succession of water-surface profiles, or through hydrographs depicting variation of surface discharge rate or depth with time at a series of stations along the length. 1 A model is said to be robust when a wide range of conditions can be modeled without user intervention to prevent an aborted simulation.

Design and Operation of Farm Irrigation Systems

441

13.2.3 Irrigation Performance The merit of an irrigation event defines how well the water is applied. Several descriptors are in common use. The uniformity of the post-irrigation areal distribution of infiltrated water (longitudinal distribution in one dimension) is often expressed through a uniformity coefficient, the ratio of minimum depth in the distribution to average depth of infiltration, d DU min = min (13.1) Dinf or the ratio of the average of the low-quarter of the depths to the average depth, ~ d LQ DU LQ = (13.2) Dinf Application efficiency is the percentage of the applied water volume contributing to the plant requirement (Burt et al., 1997). When the minimum in the distribution just equals the target (required) depth, the efficiency is known as the potential application efficiency of the minimum, DREQ PAEmin = × 100% (13.3) DQ with the volumes expressed in terms of field-wide equivalent depths. When the requirement is just matched by the average of the low-quarter depths, the application efficiency is known as the low-quarter potential application efficiency, PAELQ. Depending upon circumstances, other terms of interest in describing merit are volumes, equivalent depths, or percentages of runoff and deep percolation (infiltration in excess of the requirement). See also Chapter 14.

13.3 VARIABLES THAT INFLUENCE THE SURFACE IRRIGATION PROCESS The independent variables, governing stream behavior in the field or determining the solutions to the governing equations in a simulation, can be grouped into three categories: (1) prevailing field conditions—soil surface elevations, roughness, and infiltration; (2) geometry and dimensions of the system; and (3) operation of the system. Below, all geometrical aspects of the independent variables are lumped into one group, and the remaining field conditions, namely, resistance to the surface flow and infiltration into the soil, into two more. Management of the system can influence all of these factors: grading a field reduces irregularities in bottom elevations or levels the surface as well; torpedoing (dragging a heavy, smooth cylinder along the length of a furrow) or driving farm machinery in a furrow (making the wheel rows) compacts and smoothes the soil surface reducing both drag and infiltration; while applications of polyacrylamides (Trout et al., 1995) increase infiltration, as well as decrease the erodibility of the soil surface. The last category of independent variables describes the inflow hydrograph.

442

Chapter 13 Hydraulics of Surface Systems

13.3.1 System Geometry In the simplest case, the geometry of the channel conveying the irrigation stream is defined by the cross-section of the furrow, border strip, or basin and its length and slope, as well as the upstream and downstream boundary conditions, i.e., whether flow is free to enter and leave each end. At the upstream end of each flow channel, the inflow configuration determines if the discharge can be given as an independent quantity. With gated pipe and free flow out of the openings, inflow is independent of the depth in the furrow. Inflow through siphon tubes or spiles depends somewhat upon the flow depth, but is usually considered known. With a transverse head ditch feeding a set of furrows, possibly characterized by different bottom elevations—for example, with primitive construction methods—the inflows are directly dependent on the water surface elevation in the ditch. Thus, the layout of a furrow system can have a significant effect on surface irrigation performance. Traditional level basins are sometimes modified by constructing conveyance channels, large furrows around the perimeter of the basin, designed to conduct the water from the inlet quickly to the downstream end, so that the basin in effect is irrigated from both ends. Under idealized conditions, the efficiency of such a basin approaches that for a basin of just half the length. To allow light applications with good uniformity and provide an outlet for rainfall, level basins are sometimes constructed to allow both inflow and outflow from a side of the basin (level basin with drainback; Dedrick, 1991). In a recent extension of the surface-drainage concept, increasingly popular in the humid lower Mississippi Valley, a large laser-leveled field is crisscrossed by a grid of small shallow ditches serving both supply and drainage functions. Each small sub-area is quickly inundated leading to good uniformity. Excess water, likewise, freely drains off the planted area. At the downstream end of a traditional basin or set of furrows, the simplest configuration is a blocked end with zero outflows. Alternately, there may be an overfall into a drainage ditch. A transverse collector ditch or furrow, with a water level higher than the bottoms of the longitudinal furrows, can influence the outflow from those furrows, in effect coupling the flows within them. Bottom slopes can vary along the length of run, and if furrow flows are coupled, the bottom configurations of the neighboring furrows in a set will also affect the flow in any given furrow. In any event, both longitudinal changes in grade and transverse variations in bottom configuration can have an appreciable effect on field-wide distribution uniformity. Spatial variability in the bottom configuration is also likely in two-dimensional flows. In basins, even a small amount of variability can have an appreciable effect on the post-irrigation distribution pattern of infiltration. A newly laser-leveled basin typically exhibits a standard deviation in elevation of about 10 mm, while more traditional grading yields standard deviations from a plane surface of perhaps up to 20 to 25 mm—or twice that in some areas of the world (Fangmeier et al., 1999). See Chapter 5 for estimating the influence of spatial variability in soil-surface elevations on distribution uniformity in basins. The effect in sloping border strips has not been well documented, but, qualitatively, it can be expected that the significance of deviations from a plane surface will diminish with high flow rates, slopes, and roughness. If furrows are constructed by hand, there can be appreciable longitudinal variation in cross-section as well as in bottom slope. But even with machine-made furrows, if

Design and Operation of Farm Irrigation Systems

443

the irrigated soils are erodible and stream velocities are high, there can be temporal and spatial variation in cross-section as the result of erosion and deposition (Trout, 1992). Simulations of sediment transport in surface irrigation are in their infancy, but the importance of such cross-sectional variations can be studied by giving reasonable numbers as a function of space and time as input data to simulation models able to accommodate such variability. Comparisons of output data with and without variation will point up the need for movable-bed modeling of surface irrigation. It should be noted, however, that the current interest in modeling sediment transport phenomena (Strelkoff and Bjorneberg, 2001; Bjorneberg et al., 1999) stems primarily from the potential for design and management recommendations aimed at reducing soil movement on and from irrigated land. 13.3.2 Resistance to Flow The drag of the soil surface and inundated vegetation upon the irrigation stream flow is usually expressed through the friction slope Sf, defined as drag force in ratio to stream weight (each, per unit length of channel). It depends on the geometry of the resisting elements (size, shape, distribution, etc.), the cross-section of the flow, and the average flow velocity. The following definition expression for the Chézy C expresses that relationship, QQ (13.4) Sf = Ay 2 C 2 R Here, Ay is the cross-sectional area of flow, Q is the flow rate (with absolute value signs to account for flow—and drag—reversal), and R is hydraulic radius, the ratio of cross-sectional area to wetted perimeter. Alternatively, a conveyance K can be defined such that (13.5) Q= K Sf with K = Ay C R written for flow in the positive direction. 13.3.2.1 Surface drag. The Chézy C depends primarily on relative roughness, the ratio of roughness-element size to hydraulic radius. The widely used empirical Manning formula, R1 / 6 (13.6) C = cu n is characterized by the absolute roughness, n, expressed in m1/6 making the relative roughness dimensionless; the units coefficient, cu, in the metric system is 1.0 m1/2/s. To allow the same numerical value of n to be used in the foot-pound system, cu = 1.486 ft1/3m1/6/s. Devised without the modern theory of resistance to turbulent flow, the Manning formula, in order to correctly predict C when flow depths vary over a significant range, requires n to vary with R, contrary to its designation as an absolute roughness. Tinney and Bassett (1961) postulated a power-law variation of n with depth. More theoretical studies (Sayre and Albertson, 1961) in very rough open channels, following Colebrook and White’s pipe-flow analyses with equivalent sand-grain size, led to the logarithmic formula, ⎛R⎞ (13.7) C = 6.06 g log10 ⎜⎜ ⎟⎟ ⎝χ⎠

444

Chapter 13 Hydraulics of Surface Systems

in which χ is the absolute roughness, measured in m or ft; g is the ratio of weight to mass (about 9.81 m/s2 in the metric system, 32.2 ft/s2 in the foot-pound system). In a comparison study featuring large roughness elements (Strelkoff and Falvey, 1993) the small variation of the Sayre and Albertson χ compared to the large variation in n testifies both to the superiority and the imperfection of the Sayre and Albertson formula in this case. To allow estimation of χ from absolute-roughness measurements, extensive laboratory and field comparisons in the mid-1960s (Kruse et al., 1965) of measured surface micro-geometry and hydraulically determined values of χ led to the approximate result: (13.8) χ = 12.9σ 1.66 in which σ is the standard deviation of surface elevations in the micro-geometry. More information on roughness and drag can be found in the extensive bibliographic review of Maheshwari (1992). 13.3.2.2 Vegetative drag. The purely empirical, classical studies of Ree and Palmer (1949) have been supplanted by more fundamental, physically based approaches, depending on whether the vegetation is wholly submerged or protrudes through the water surface. For the submerged case, the flexibility of the plant parts, bending into evermore prone positions with increasing flow velocities, has been described by a vegetation stiffness parameter, MEI. Here, I is a representative moment of inertia of the plant-stem cross-section, E is the elastic modulus of elasticity of the plant material, and M is the number of stems per unit boundary area (Kouwen et al., 1981). Form drag on vegetation penetrating the flow is distributed over the entire channel cross-section and is the sum of the drag forces on the individual plant parts. In a border strip (Petryk and Bosmajian, 1975) with plant density ap (projected area per unit volume of flow) varying with height η above the bottom, the friction slope chargeable to vegetation in flow at velocity V and depth y is S f veg = C D

V2 1 y a p dη 2 g y ∫0

(13.9)

with CD a drag coefficient (of value, typically, around 1.0). In the especially simple, laboratory studies with vertical wires of diameter D, placed in the flow at a density of M “stems” per unit plan area, ap = MD, a constant. Evaluations of vegetative density can proceed either geometrically, or hydraulically, i.e., calculating backwards from measurements of friction slope, average velocity, and flow depth. The hydraulic measurements would necessarily include the soil-surface drag, a significant component if the vegetation is sparse, and which must be estimated to isolate ap. Characterization of variations in ap can be found in Dudley et al. (1996) and Dunn et al. (1996). 13.3.2.3 Combining soil-surface drag and plant drag. A dense, submerged growth makes the velocity distribution in the vertical more uniform, and hence, at the boundary, leads to a steeper gradient and greater shear. Furthermore, the nonuniformity in velocity distribution induced by wall shear affects the drag on the plant parts as a function of depth. But in a first approximation, the drag of the soil surface and that of the plant parts can be considered independent. Then the two components of drag can be individually calculated and simply added (note that while drag forces and Sf are additive, C, n, and ι are not).

Design and Operation of Farm Irrigation Systems

445

Neither the Sayre and Albertson χ, nor the vegetative ap, are in common use today for surface irrigation simulation or design (SRFR can accommodate values of χ and ap, constant within a cross-section). While both can be established in hydraulic experiments, there is no body of experience suggesting appropriate values a priori. The common measure of resistance is, currently, the Manning n in all cases. NRCS in its design manuals recommends values for specified border-strip conditions (see Chapter 14). 13.3.2.4 Two-dimensional considerations—Sf as a vector. With the drag force a vector whose direction is opposite the flow-velocity vector, the Sf vector is oriented, by convention, opposite the drag-force and, in the one-dimensional case, in the same direction as the flow. This is evident in Equation 13.4, valid for one-dimensional flow in either direction. In two-dimensional flow, corresponding relationships need to be expressed more generally because the irrigation stream, in flowing around high spots in the field surface can come from any direction. For an arbitrary flow direction, but for isotropic roughness, independent of the direction of flow, in a two-dimensional x1-x2 coordinate system, Equation 13.5 can be written q1 = K S f cos θ1

...

q2 = K S f cosθ 2

(13.10)

in which, with i = 1, 2, qi is the component of the vector discharge per unit width, and Sf is the magnitude of the friction-slope vector. K in this case is a positive scalar, without directional attributes. The angle θi is defined as the angle between the Sf vector and the xi-direction. A field can exhibit greater resistance in one direction than another, for example, when it has been tilled with corrugations parallel to one side, or when drilled grain is grown, with greater resistance to flow across the rows than along them. In this case, the conveyance must be viewed as a tensor whose principal values follow directly from the largest and smallest roughness values encountered by flows in various directions. When these are not the same, the friction slope will no longer be collinear with the discharge vector, but will be skewed (Strelkoff et al., 2003a). 13.3.3 Infiltration Cumulative infiltration is expressed as a depth of surface-water, z (the actual location of the wetting front below the soil surface depends upon the porosity and degree of saturation of the soil). The accumulated intake in a border strip is generally assumed to be a function of intake opportunity time alone, for any given soil at any given initial water content. In furrows, the accumulated volume infiltrated per unit length of furrow, Az, has the units of area. Local field depth z is given by z = Az/B, in which B is the furrow spacing. And because the soil-surface area through which infiltration occurs—the wetted perimeter—depends upon depth, the latter, too, in principle, affects cumulative intake. In theory, infiltration is described as a boundary condition on the Richards equation for unsaturated flow in the soil medium (see Chapter 6). In surface irrigation, prior to recession, bulk water available at the soil surface is drawn into the soil medium. A key factor in surface irrigation simulation is the infiltrated volume per unit stream length. It is possible to couple the surface-flow equations to some variant of the Richards equation (Tabuada et al., 1995; Skonard 2002; Zerihun et al., 2005), but such a theoretical approach still requires empirical soil data relating water content to piezometric

446

Chapter 13 Hydraulics of Surface Systems

head and hydraulic conductivity. The mathematical complexity the coupling entails, together with observed inhomogeneity and anisotropy of natural soils (see EncisoMedina et al., 1998, for an approximate treatment), suggests a more direct empirical approach. The major issue involved is cumulative infiltration, as a function of inundation time, wetted perimeter, and furrow spacing. It is true, even in border or basin irrigation, that depth of flow plays a role by providing some positive pressure at the soil surface, but this is small, on the order of a 2% increase in sorptivity (S, in Equation 13.13, below) per cm of depth (Philip, 1969). In addition, the effects of successive periods of dewatering and rewetting are important in surge flow and, potentially, with some other inflow-hydrograph patterns, such as cutback flow. 13.3.3.1 Cumulative infiltration as a function of opportunity time. The observed features of the cumulative-infiltration function of time are a large rate of infiltration upon initial wetting, followed by a decrease in rate, typically approaching a constant at large infiltration times. In certain circumstances (typically, with cracked vertisols) a significant volume is infiltrated as soon as the water is made available at the surface. A formula exhibiting all of these features is the following modification of the Kostiakov formula (Kostiakov, 1932; Lewis, 1937): z = kτ a + bτ + c

(13.11)

in which z is the volume infiltrated per unit infiltrating area, τ is the opportunity time, and k, a, b, and c are empirical constants. Perea et al. (2003) compared the solution of the Richards equation in a homogeneous isotropic silt loam soil to a fitted equation of the form of Equation 13.11. In this case the modified Kostiakov equation fit the theoretical results to within a half millimeter over a time span between 1 min and 20 h. The original monomial Kostiakov formulation, z = kτa, is sometimes used for cracking, swelling clay soils with a tiny value of the exponent a, but a better representation of physical reality is a nonzero value of c. Furthermore, numerical simulations of surface flow with such a highly nonlinear infiltration function are inconvenient and troublesome, as the computations attempt to track the sharp “dog-leg” in the infiltration profile. The power-law formula is often satisfactory for moderate times in loam or sandy soils, but at large times, it is characterized by an anomalous, ever-decreasing infiltration rate. An observed relatively large infiltration rate at large times is often achieved with the basic formula by postulating a large value of a. A better physical representation is achieved (Hartley, 1992) by incorporating a nonzero value of b and a moderate value of a, closer to the theoretical value, 0.5. Indeed, Philip (1969) developed a theoretical infiltration equation based on unsaturated flow in a homogeneous isotropic soil medium in the form z = Sτ 1/ 2 + Aτ (13.12) Although S and A are constants depending primarily on the soil characteristic curves, Equation 13.12 falls within the group of functions described by Equation 13.11. It is often assumed that the basic rate of infiltration is achieved near the end of the irrigation and can be found by simultaneously measuring outflow Qro and inflow Q0, i.e., Q − Qro (13.13) b= 0 LW

Design and Operation of Farm Irrigation Systems

447

in which L is the furrow length, and W is a representative wetted perimeter. It should be noted, however, the contribution of the Kostiakov power term in Equation 13.11 to infiltration rate can be of major significance even after many hours (Strelkoff and Clemmens, 2001). The Kostiakov branch function (Clemmens, 1981) describes those soils which achieve the basic infiltration rate sooner than indicated in Equation 13.11. Modified by incorporation of a constant term, this is z = c + kτ a L τ ≤ τ B 1 ⎞ 1− a

; z = c B + bτ L τ > τ B

⎛ ak ; c B = c + kτ B a − bτ B (13.14) ⎟ ⎝ b ⎠ In Equations 13.14, τB is the cross-over value of inundation time, at which the infiltration rate of the first branch matches the final, basic rate b, and cB is the intercept on the z-axis of the second branch. In some cases, the entire first (power-law) branch can be effectively assumed to take place in zero time (relative to the total time of simulation), leading to an exceptionally simple, one-branch infiltration formula consisting of a constant infiltration depth upon initial wetting followed by a constant rate (CollisGeorge, 1974). Midway in complexity between solutions of the Richards equation and the pure empiricism of the Kostiakov formulations is the Green and Ampt equation (Green and Ampt, 1911) (see also Chapter 6 and Skonard, 2002). This implicit equation for cumulative infiltration represents an approximate solution to the Richards equation, under the assumption that behind the wetting front, the soil is saturated, and that the soil water content exhibits a discontinuity across this front. The three parameters can be determined from the same kind of measurements as required for a general solution of the Richards equation, but it is also possible to fit measured infiltration data for the formula parameters. In USDA (1974), the NRCS characterized infiltration into borders and basins by membership in a family (extended to furrows in USDA, 1985). The concept implies that the curve of cumulative infiltration vs. time for any soil at any initial water content will fit within one of the regions, or families, into which the z-τ plane is divided. Each family is characterized by a function of the form z = kτa + c, passing through the middle of the region, and named for the basic infiltration rate for the soil (in inches per hour). The c-value in the SCS functional form (a constant for all families) does not represent anything physical, but was introduced to better fit empirical data. While many soils fail to fit any of the families (graphs of their cumulative infiltration vs. opportunity time intersect a number of SCS families), some are successfully incorporated within the SCS group. If experience justifies describing a soil in this way, the families allow a simple representation of infiltration, with prescribed values of k, a, and c (see also Chapter 14). It follows that k and a are related,

τB = ⎜

⎡ 0.148 ⎤ k (a ) = 60 a ⎢14088 a 45 + ⎥ (− ln a )1.652 ⎦⎥ ⎣⎢

(13.15)

as per the empirical curve fit to the NRCS-tabulated data (Valiantzas et al., 2001). Merriam and Clemmens (1985) built upon the NRCS concept of infiltration families and developed the time-rated intake families, arguing that the time required to

448

Chapter 13 Hydraulics of Surface Systems

infiltrate a target depth is the key parameter characterizing infiltration. With a large database of non-cracking soils, they found a general empirical relation between Kostiakov a and the time to infiltrate 100 mm, (13.16) a = 0.675 − 0.2125 log10 (t100 ) A single measurement of t100, the time in hours to infiltrate 100 mm, provides the entire Kostiakov cumulative infiltration function, since, with k = 100/t100a, ⎛ 0.675 − a ⎞ 2− a⎜ ⎟ = 10 ⎝ 0.2125 ⎠

k (13.17) (see also Chapter 14). 13.3.3.2 Influence of wetted perimeter on infiltration. In borders and basins, the wetted perimeter is essentially independent of flow depth. In furrows the infiltrating surface depends on depth and varies with both distance and time. If infiltration is being calculated theoretically, e.g., by solution of the unsaturated flow equations in the porous soil (Schmitz, 1993a,b; Šimunek et al., 1999) coupled to the hydrodynamic equations in the surface stream (Enciso-Medina et al., 1998; Skonard, 2002), the wetted perimeter would be taken into account automatically. Boundaries to the lateral flux of soil water engendered by infiltration from neighboring wetted furrows are treated by postulating a no-horizontal-flow boundary midway between furrows. But infiltration in surface irrigation simulations is usually treated as an empirical boundary condition on the surface stream. For furrow infiltration, two schools of thought have emerged. Some models of the surface irrigation process expect as data input the parameters of empirical infiltration formulas directly expressing cumulative volume per unit length of furrow, AZ, as a function of time. This implies that inflow and outflow measurements had been taken in the furrow, or a similar one, to yield a time-rate of growth of infiltrated volume, thus incorporating the ramifications of whatever wetted perimeter was present when the flows were measured. Infiltrationformula parameters are generally obtained by measuring the increasing infiltrated volume in a short furrow section containing water at some constant depth and fitting the data with a formula, now, for example, of the form AZ = Kτ a + Bτ + C (in contrast with Equation 13.11), with the units of each term now comprising an area (volume per unit length). The measured infiltration is then viewed as a consequence of both soil properties (including initial soil water content) and the details of the irrigation system (furrow size, shape, spacing, flow depth, etc.) But how these data should be adjusted to apply to water at, say, a different depth, even a constant one, is not clear. In other models, the infiltration-data input is considered a function of soil properties alone, with the contribution of the irrigation system then calculated by the model. The data input consists of formula parameters for volume infiltrated per unit infiltrating area (i.e., z, cumulative depth of infiltration, as in Equation 13.11), like basin infiltration. The volume infiltrated per unit length of furrow, AZ, is then calculated for the varying flow depths and wetted perimeters of the simulated irrigation stream (NRCS, 1984; Perea et al., 2003). In any event, at present, it is essential to recognize that the parameter values in the infiltration formulas are strongly dependent on the accompanying wetted perimeter assumption. The important physical quantity is Az—for a given value of Az, the parameters will be much smaller for a wetted perimeter assumed equal to the furrow spacing rather than, say, on the smaller actual wetted perimeter, dependent on the

Design and Operation of Farm Irrigation Systems

449

stream hydraulics. Thus, for a given physical infiltration, the parameter values should be roughly inversely proportional to the average wetted perimeter assumed encountered in the simulation. It should be noted, in passing, that there is some experimental evidence that as wetted perimeter increases, infiltration increases more than in proportion (or less), suggesting that wetted perimeter be raised to a power somewhat greater than unity (Blair and Smerdon, 1985), or less (Trout, 1992). The significance of variations in wetted perimeter, along the length of an advancing stream, depends on circumstances. If the advance constitutes a significant fraction of the total application time, the variation in wetted perimeter can have a significant effect on the final distribution of infiltrated water. This effect is minimal if there is a long period of runoff. But there may also be a long-term difference in wetted perimeter between the head and tail ends of the furrow, particularly with porous soils. In a tight soil, over a long period of runoff, if the flow in the furrow is essentially uniform, the wetted perimeter will be practically constant in the furrow, nearly equal to the wetted perimeter at normal depth for the inflowing discharge. With extensively cracked soils drawing water off laterally, the variations in wetted perimeter with depth, again, may have little overall effect on the distribution of infiltrated water, the effective wetted perimeter basically just the furrow spacing. A review of extensive field data (Walker and Kasilingam, 2004) indicated that measured wetted perimeters failed to show any consistent variation with either time or distance during the course of irrigations, leading to doubts that incorporating theoretical changes based upon the assumption of constant cross-sections (in the dry) in the course of a simulated irrigation, provides any improvement in accuracy. 13.3.3.3 Influence of rewetting on infiltration. It has long been noted that following recession of water from a furrow, upon rewetting, the infiltration rate is smaller than it would have been had the furrow been wet continuously. Various theoretical reasons have been suggested for this behavior, including redistribution of soil water after the initial wetting, but a likely significant cause is the tightening of the soil structure near the surface under the influence of the negative soil pore pressures as water disappears from the surface. From the early eighties onward, irrigation by pulses of applied water, with recession in between, has been used to take advantage of the reduced infiltration rates for the successive surges and achieve faster advance rates and a more uniform application than with continuous inflow until cutoff. The issue of rewetting arises also in the case of cutback furrow streams, if the reduced flow rate is incapable of maintaining infiltration over the entire length of furrow previously wetted at the full flow rate. From the standpoint of simulation, some albeit approximate estimate of infiltration changes under rewetting is desirable, in lieu of totally new measured values for the constants in the infiltration formulas for each surge, or renewed advance. Theoretical solutions of the Richards equation would likely fail to provide good estimates unless some recognition of the tightened soil structure were incorporated into those solutions. Various empirical models have been proposed (e.g., Walker and Humphreys, 1983; Blair and Smerdon, 1984; Izuno and Podmore, 1985; Purkey and Wallender, 1989). More recent investigations (Palomo et al., 1996) suggest a simple, accurate (for the soils tested) modification of Izuno and Podmore (1985) for soils described by Equations 13.14. The Kostiakov formulation of the first branch, in effect for the first surge, is replaced for all subsequent surges by the constant, final (basic) infiltration rate, the second branch.

450

Chapter 13 Hydraulics of Surface Systems

13.3.4 The Inflow Hydrograph In the simplest case, inflow is held constant until cutoff. In cutback operation, the inflow is cut, usually in half, when the stream reaches the end of the run. Other fractions of initial stream size and other times to effect the cutback might prove more effective and can be studied with simulations. A survey of several different inflow patterns as they affect uniformity and efficiency (Alazba and Fangmeier, 1995) provides an example of necessary studies on the influence of hydrograph shapes on performance and how these can be achieved in a practical way, without extensive labor requirements. In surge operation, pulses of inflow (usually of equal magnitude and of increasing duration) are released, with the intent to cut off each surge as successive constant increments of total length of run are wetted. Commercial surge valves are designed with a butterfly valve to shunt water to one or the other of two sets of furrows, the on time of one set corresponding to the off time of the other. These valves, which operate on a timer, are programmed to estimate the increasing pulse durations necessary. After advance is complete, they start a more or less rapid pulsing (on the order of 30-minute cycles) between the two sides in what is called a cutback mode, until the final cutoff. The two furrow sets are not served with identical inflow hydrographs, because one side starts cutback immediately after advance is complete, while the other pauses after its last advance, corresponding to the last advance increment of its counterpart (see, e.g., sample surge simulations in Strelkoff et al., 1998). Cablegation (Kemper et al., 1981) provides a gradually decreasing inflow to a set of furrows fed from gated pipe through which a plug is pulled by a cable. A computer program (Kincaid and Stevens, 1992) modeling the resultant furrow inflow hydrograph was incorporated into simulation software (see, e.g., sample cablegation simulations in Strelkoff et al., 1998). 13.3.5 Estimation of the Independent Parameters Geometrical parameters for input to simulations or design can be measured in the field with rulers, tapes, levels, and other surveying equipment. It is worthwhile to note that any measurement of a level in a survey of field-surface (or furrow-bottom) elevations with standard surveying equipment is likely to be in error by at least 3 mm. Measurements of inflow are conveniently and accurately obtained by means of long-throated critical-depth flumes easy to construct, commercially available in many sizes, and requiring no field or laboratory calibration (Clemmens et al., 2001b). Theoretical predictions of infiltration require measurement of the soil characteristic curves (Chapter 6), but in most cases, empirical determinations are made of infiltrated depth, or volume per unit length of furrow, as a function of time. Direct measurements are possible with ring infiltrometers (preferably buffered), but their indications, at one or a few points in a field, are of limited value in predicting what the average field infiltration will be, under flowing water in particular. In general, because of the influence of running water on the soil structure, flowing-furrow infiltrometers are preferable over ponded-furrow infiltrometers, which typically underestimate infiltration during an actual irrigation. Where it is necessary to measure water depths to estimate infiltration, say in large level basins, the double-bubbler apparatus of Clemmens and Dedrick (1984) allows this to be done from a distance, accurately and conveniently. A considerable body of experience has been gained in the last twenty years in estimating infiltration and roughness from measurements of complete border and furrow

Design and Operation of Farm Irrigation Systems

451

irrigations, or during the advance period. Because these techniques often utilize one or another simulation method to accomplish the evaluation, presentation is delayed until Section 13.7, after mathematical simulation of surface irrigation has been presented.

13.4 CONSERVATION LAWS FOR MASS AND MOMENTUM 13.4.1 Water Volume Balance As irrigation water flows down its channel, part of it infiltrating into the surrounding soil and part of it advancing onward—to be infiltrated farther down the channel or to drain off the end of the field as surface runoff—a volume balance is necessarily maintained: all of the water entering the field at the inlet is accounted for at all times. At any stage of the process, some of the total inflow volume has infiltrated into the soil, some has evaporated2, some is (temporarily) stored on the surface, and some, possibly, has run off. This accounting is a manifestation of the physical principle of conservation of mass. Because water exhibits so little compressibility in the context of surface irrigation, conservation of mass translates into conservation of volume. In terms of total volumes, the balance can be expressed mathematically as follows: VQ (t ) = VY (t ) + VZ (t ) + VRO (t )

(13.18)

in which VQ is the volume of inflow, VY is the volume on the surface, VZ is the volume infiltrated, and VRO is the volume of runoff, all at any time t from the start of the irrigation. Equation 13.18 is equally valid for a one-dimensional or two-dimensional formulation. In Equation 13.18, t

VQ (t ) = ∫ Q0 dt

(13.19)

0

in which Q0 is the inflow rate. After advance is complete, runoff volume begins to accumulate, at the rate QRO(t): t

VRO (t ) = ∫ QRO dt

(13.20)

0

In contrast to inflow, runoff is usually extremely variable. Zero during the entire advance, it typically rises quickly to a shoulder, continues to rise gradually as infiltration rates decrease with time, and finally returns to zero more or less gradually after cutoff. In a one-dimensional context, the surface volume is given by x A (t )

VY (t ) =

∫ AY ( x, t ) dx

(13.21)

0

in which AY is the volume per unit length, or cross-sectional area, of the surface stream, and xA is the advance distance. After advance is complete, xA = L, the length of run. Similarly, the infiltrated volume is x A (t )

VZ (t ) =

∫ AZ ( x, t ) dx

(13.22)

0

2

Evaporation is generally ignored; the influence of air and soil temperature is much more pronounced in affecting water viscosity and infiltration rate than in hydraulically significant evaporation.

452

Chapter 13 Hydraulics of Surface Systems

in which AZ is the infiltrated volume per unit length. If infiltration is assumed a function only of wetting time (with variable hydrostatic-pressure or wetted-perimeter effects neglected), Equation 13.22 can be written x A (t )

VZ (t ) =

∫ AZ (t − t A [ x]) dx

(13.23)

0

(provided xA never decreases with t); tA is the time to advance to x. 13.4.2 Water Dynamics Of equal importance in governing the flow of the irrigation stream is the dynamic interplay between the forces of gravity and flow resistance. The gravitational forces are evident in two ways: directly, as the down-field component of fluid weight, dependent on both water depth and bottom slope, and indirectly (through the hydrostatic pressure distribution in the vertical) as the consequence of a depth gradient, present if the water surface is not parallel to the bottom. The two aspects of the gravitational influence are reflected in the water-surface slope, S0: if this is horizontal, the net downfield gravitational force is zero; the water is ponded, and the flow is zero. When the downfield gravitational pull on a stream cross-section, ρgAYS0 per unit length (with ρ the mass density), is just balanced by the drag upslope of the channel walls and vegetation, ρgAYSf per unit length, the flow is in equilibrium and uniform, at constant depth, called normal depth. In a given channel, normal depth is directly related to flow rate; with some empirical formula for the Chézy C, e.g., Equation 13.6,

Q=

cu AY R 2 / 3 S 0 n

(13.24)

in accord with Equation 13.4, with Sf = S0. If the gravitational and drag forces are not in equilibrium, there is a depth gradient, and the water accelerates or decelerates. In most cases of surface irrigation, however, flow velocities are so small that their temporal changes are negligible, and thus, the forces of fluid weight, depth gradient, and boundary and vegetative drag are essentially in equilibrium (see Section 13.6.7). Integral relationships similar to those developed for mass conservation can be written for momentum conservation in the entire stream, but the dynamic equations are usually used to describe small slices of the surface stream in a hydrodynamic approach to simulation (see Section 13.6). In addition, hydrologic modeling commonly assumes normal depth at the inlet.

13.5 HYDROLOGIC MODELING OF THE SURFACE IRRIGATION PROCESS An irrigation stream is subject to much the same analyses as in simulating canal and river flows, the major differences being the strong influence of infiltration and advance over a dry bed. But to avoid the complex mathematics of the hydrodynamic approach, reasonable assumptions on size and shape are sometimes imposed on the surface-stream or infiltration profiles, as described below. Then, application of mass conservation alone leads to the stream behavior. To the extent that the assumptions are justified, the results can be as good as those of hydrodynamic simulations, with far less computational effort and a far more robust model. Furthermore, the simplicity of the equations makes them better candidates than the hydrodynamic equations for solv-

Design and Operation of Farm Irrigation Systems

453

ing inverse problems—estimation of infiltration and resistance from flow observations, or finding design variables for a given performance level. However, if there isn’t a good physical basis for the assumptions, the hydrologic model can be grossly in error, though it remains robust and economical. In the present section, the techniques of hydrologic modeling are presented; the validity and accuracy of the assumptions made are, for the most part, deferred until hydrodynamic modeling has been discussed, as the results thereof are taken as a standard for comparison. 13.5.1 Introduction to Hydrologic Simulation—a Simple, Explicit Model As noted, conservation of mass or volume, i.e., a volume balance, lies at the heart of any rational simulation of surface irrigation flows. In a one-dimensional formulation, during advance, Equation 13.18 can be expressed ~ ~ (13.25) Q0 t = x A AY + x A AZ ~ in which xA is the advance distance of the stream up to time t, AY ( x, t ) is the average ~ stream volume per unit length (cross-sectional flow area), and AZ ( x, t ) is the average infiltrated volume per unit length; time averages are shown barred, while distance averages over the stream length are expressed through a tilde. While inflow continues, i.e., as long as the upstream depth does not fall to zero, Equation 13.25 can be written for the advance function as Q0 t xA = (13.26) rY AY (0, t ) + rZ AZ (0, t ) in which rY and rZ are shape factors for the surface and subsurface profiles, respectively, i.e., ratios of average cross-sectional area to upstream cross-sectional area. If estimates are made for the four terms in the denominator, Equation 13.26 becomes a simple, explicit formula for advance. AZ(0,t) can be calculated, for at x = 0, t = τ. In a moderately steep flow channel, the upstream depth rises quickly to normal depth, yielding AY. Given the generally convex form of the surface-stream and infiltration profiles, the range of possibilities for the shape factors is 0.5 to 1. Both surface and subsurface profiles can be assumed monomial power laws in distance back from the wave front. During advance, the wave profile is quite blunt near the front. This suggests a rather small exponent in the power law. The arbitrary choice of 1/4 for the power yields a shape factor rY = 0.8 (see Section 13.6.12 for theoretical values). With z(τ) given by the Kostiakov equation, assumption of a constant advance rate leads to the exponent a for the infiltration profile, and the corresponding shape factor is 1/(1+a) (theoretically correct for small distances from the tip of the advancing profile: Katopodes and Strelkoff, 1977a). The right-hand side of Equation 13.26 is now determined for any advance time t. As an example, these assumptions are applied to a field test in a border strip (AR-9: Roth et al., 1974). The irrigation parameters (see Figure 13.1) lead to y0 = 0.0285 m, ~y = 0.023 m, and rZ = 0.74. A plot of Equation 13.26 for this case over a range of time is shown by the solid line in Figure 13.1, with the measured advance trajectory shown as open circles. Solutions by more complex methods outlined in subsequent sections are also shown. Evidently in this case, there is not a great deal of difference with the various methods of solution, and all agree well with observation. However, the above assumptions are not always justified. Test ID-NB-12AS-8-60 (Bondurant, 1971) was performed in a border strip of very small slope and a heavy

454

Chapter 13 Hydraulics of Surface Systems

growth of alfalfa. This case (Figure 13.2) is characterized by y0 = 0.3253 m, rY = 0.8, rZ = 0.6. Solution of Equation 13.26 yields the solid curve, greatly at variance with the open circles of observed advance. The great discrepancy between calculated and observed results has prompted much of the research effort devoted to surface irrigation simulation. Certain results of that effort are shown as dashed and dot-dash curves on the figure, to be explained in subsequent sections (13.5.2, 13.6.12). Observed ry=0.8, rz=1/(1+a)

Time (minutes)

40

ry=0.8, rz varies in accord with Lewis-Milne Equation Hydrodynamic theory (zero-inertia)

30

AR-9

20

q0=2.40 Lps/m S0=0.001 n=0.035 m1/6 k=61.8 mm/hra a=0.358

10

0

0

25

50

75

100

Advance (m) Figure 13.1. Comparison of theory and experiment in a border on a moderate gradient.

Figure 13.2. Comparison of theory and experiment in a nearly level border.

Design and Operation of Farm Irrigation Systems

455

13.5.2 Lewis and Milne Equation Lewis and Milne (1938) recognized that if infiltration is a function of wetting time alone, the subsurface profile depends solely upon the infiltration and advance functions of time. Then, rZ is a function of xA(t), and with the infiltration function of time given, Equation 13.25 can be viewed as an implicit expression for the advance, eliminating some of the previous arbitrariness. Indeed, with wetting time, τ (x,t) = t – tA(x), then, t ~ dx Q0 ⋅t = AY ⋅ x A + ∫ AZ (t − t A ) dt A (13.27) dt A 0

Although Equation 13.27 is exact for the few conditions postulated, it contains too many unknowns to allow for solution of the advance function, specifically, the un~ known variation, AY (x,t) from which the distance-average, AY (t ), could be obtained. Without a rational (or arbitrary) estimate of this, or of the two factors Ay0(t) and ry(t), further progress is impossible. 13.5.3 Assumed Average Stream Section In applying Equation 13.27, in the often misnamed volume-balance approach (every rational approach to simulation incorporates a volume balance), a constant ~ value for Ay is usually assumed, based on normal depth at the inflow and known a priori—as in the simple, explicit calculation of Section 13.5.1. It was shown theoretically (Katopodes and Strelkoff, 1977a) that close to the front of a stream with negligible infiltration and Manning roughness, the exponent of a power-law surface profile is 3 /7. Tinney and Bassett (1961) found a terminal profile shape, with constant inflow and no infiltration, following a hyperbolic tangent law for its entire length, somewhat blunter than the 3/7 power law and closer to a power of 1/4. Soil with substantial infiltration, on the other hand, can lead to nearly linearly increasing depth with distance from the stream front. In any event, the greater the infiltration, the less critical is the ~ estimate of Ay. Indeed, if advance proceeds for a sufficiently long time in a sloping channel, AY0 and rY are both limited, while AZ, typically, grows without bound. In such ~ cases, advance becomes relatively insensitive to the assumption for AY . Lewis and Milne proposed solutions for several simple infiltration functions z(τ); the simplest case, for I = dz/dτ = constant had already been solved by Israelsen (1932). In this case, Equation 13.27 can be supplanted by the differential equation, ~ Q0 ⋅ dt = AY ⋅ dx A + W ⋅ I ⋅ x A ⋅ dt (13.28)

with the solution

W ⋅I t ⎛ − ~ Q0 ⎜ A xA = ⎜1 − e Y W ⋅I ⎜ ⎝

⎞ ⎟ ⎟ ⎟ ⎠

(13.29)

exhibiting an advance necessarily limited to Q0/(W· I). But more realistic infiltration like the Kostiakov z = kτ a was not incorporated for over 20 years. One of the earliest schemes, for general forms of infiltration formula was presented by Hall (1956), who dealt with the (convolution) integral in Equation 13.27 numerically.

456

Chapter 13 Hydraulics of Surface Systems

y

t2

t t 1=1

z01

2 t 2=

t

3 t 3=

x

t

t

xA4

t4

xA1

t4

t3 xA3

xA 2

=4

t1

z02 z03 z04 xA1

xA2

xA3

xA4

z

Figure 13.3. Cumulative infiltration during advance. Time lines separated by constant δt.

13.5.4 Hall Recursive Technique Figure 13.3 shows how z(x,t) is related to z(0,t). The profiles are drawn for successive instants of time, a constant interval δt apart. Once z(τ) is known, the horizontal lines labeled z01, z02,..., etc. separated by the constant time interval can be drawn. The verticals xA1, xA2,..., etc., represent successive locations of the stream front (albeit unknown), at t = δt, t = 2δt, t = 3δt, etc. It follows (Hall, 1956) that the successive subsurface profiles can be constructed by joining the opposite corners of the resulting rectangles as shown (furrow irrigation requires neglecting the effect of variable wetted perimeter upon Az or z). Indeed, with zk,i representing z(xk,ti), and the xk chosen the same as the successive positions of the advancing front xAi, then zk,i = z0,i-k. It follows that the change, in one time step, of the convolution integral can be approximated numerically, so that the increment of advance in the time step is i −1

Qi ⋅ δt − ∑

δx Ai =

k =1

AZ 0 ,i−k +1 − AZ 0 ,i−k −1

⎛ ⎝

1⎞ 2⎠

δxk + ⎜ rzT − ⎟Az0,1 δx Ai−1

2 ry Ay0 + rzT Az0 ,1

(13.30)

(Hall 1956, corrected as per Alazba and Strelkoff 1994), in which rzT is the shape factor for the tip cell only of the subsurface infiltration profile. 13.5.5 Solutions of the Lewis and Milne Equation with Constant Average Stream Section One standard approach in treating convolution integrals is with Laplace transforms (Boas, 1966; Philip and Farrell, 1964; Lenau, 1969). Amongst other efforts to solve the Lewis and Milne integral equation, Hart et al. (1968) derived a dimensionless version for the Kostiakov equation, solved by a numerical solution of controllable accuracy, while Collis-George (1974) developed a closed-form solution for the second branch of the Kostiakov-Clemmens branch function. Figure 13.4 provides an overview of solutions to the Lewis and Milne equation with constant average surface-water depth and Kostiakov infiltration formula. The

Design and Operation of Farm Irrigation Systems

457

dimensionless representation allows all solutions to be represented in a single family of graphs. In this representation, t* = t/TR, and xA* = xA/XR, in which ~ ⎞1/ a ⎛ A Q T y ⎟ ⎜ TR = X R = 0~ R (13.31) ⎜ B0 k ⎟ Ay ⎝ ⎠ Each point was obtained by numerically integrating a definite integral derived as per Lenau (1969), except for the curve labeled a = 1.0, representing a constant infiltration rate. For this, the dimensionless counterpart of Equation 13.29 was graphed. The lower limit of a, 0.0, is not shown, because in this case the reference time TR becomes undefined, and the selected reduction to dimensionless form breaks down. In dimensional terms, the advance for a = 0.0 is given by Q xA = ~ 0 Ay + B0 k

(13.32)

in which k is the instantaneous infiltration depth when water is applied (as into cracked soils over hardpan). Figure 13.5 is a similar overview, showing the corresponding variation in infiltration-profile shape factor (which would yield a more accurate estimate of advance, with Equation 13.26, than Figure 13.4). Theoretically derived limiting values of rz are (Hart et al., 1968), 1 a π (1 − a ) rz = at large time (13.33) rz = at small time 1+ a sin (aπ )

The two known limits have suggested a number of curve fitting approaches to replace the information given graphically in Figures 13.4 and 13.5 by analytic procedures. Valiantzas (1997a) took advantage of the partial collapse of the curves of Figure 13.4 towards a single curve when they are normalized and the independent variable is changed to xA*/t* (which varies in the range 1 to 0, as t* varies from 0 to ∞ ). He derived an approximate analytic fit by a simple polynomial; a few iterations prove necessary to construct a point x*(tA*) on the advance curve because of the form of the independent variable in the polynomial. Alazba (1999) sought an explicit solution by fitting a rational expression to rz(t*), which then requires solution of a quadratic equation for a point x*(tA*). Both approaches are computationally less intensive than the Lenau (1969). It is useful to consider the conditions of Figures 13.1 and 13.2 in the light of Figure 13.5 in order to judge the magnitude of the error introduced by the assumption of the small-time value for rz appearing in Equation 13.33. In Figure 13.1, the dimensionless time associated with a 40-minute advance is 10.9, which in Figure 13.5 is seen (by interpolating between the 0.3 and 0.4 a-curves) to increase the small-time value of rz of 0.74 by about 4%, to 0.77. With Az about 1.7 times Ay at that time, correcting rz would decrease xA by about 2.5%, i.e., to the dashed line, which represents the solution to the Lewis and Milne equation. Similar corrections in the case of Figure 13.2 also move the calculated advance from the solid line to the dashed line, which still exhibits a fundamental divergence from the measured values. In this case, then, the assumption of small-time conditions for rz was not significant (for a real time t = 60 min, t* = 10-5), and some other factor is responsible for the discrepancy (in this extreme case, up-

458

Chapter 13 Hydraulics of Surface Systems

10 000 0.3

1000 0.8

100.

0.7

0.6

0.5

0.4

0.2

.1 a=0

0.9 a=1.0

10.0

1.00

0.10

0.01 0.01

0.10

1.00

10.0

100.

1000.

Figure 13.4. Advance with constant average surface depth (Kostiakov infiltration). 1.00 a=0.1 0.2 0.3

0.75

0.4 0.5 0.6 0.7 0.8 0.9 a=1.0

0.50 0.01

0.10

1.00

10.0

100.

1000.

10 000

Figure 13.5. Infiltration-profile shape factor (Kostiakov infiltration).

stream depth during the hour of advance shown is very much less than normal depth; see Section 13.6.12). It cannot be generally said, however, that errors in assuming the first of Equations 13.33 throughout the advance are negligible: Figure 13.5 shows that while at low values of a, variations in rZ with time are indeed insignificant, the larger values show variations of up to 100%. The commonly used, theoretical value, a = 0.5, shows a variation of about 16% over the entire possible range of irrigation times.

Design and Operation of Farm Irrigation Systems

459

Figures similar to Figures 13.4 and 13.5 could be drawn for infiltration formulations other than the Kostiakov. For example, with the theoretical Philip formula, Sτ 1/2 + Aτ, dimensionless variables could be based on the reference time, ~ ⎞2 ⎛ A y ⎟ ⎜ TR = , instead of that in Equations 13.31, and the parameter identifying the ⎜W ⋅S ⎟ ⎝ ⎠ A curves would then be A* = 2 ~ . Infiltration formulas with more than two paS W / Ay rameters would require pages of graphs to cover the potential range. 13.5.6 Power-Law Advance The curves on logarithmic paper in Figure 13.4, solutions of the Lewis and Milne equation for advance (an expression of mass conservation) with a constant average surface water depth and power-law infiltration, exhibit straight-line (power-law) advance, xA = f t h (13.34)

⋅

at both very small times and very large times, but not in between. However, measured advance as well as hydrodynamic simulations often show a good power-law fit over moderate intervals, say from 10% of field length to field end. Then, assumption of both z = kτ a and xA = f · th leads to significant simplifications in the Lewis and Milne expression for infiltrated volume, namely, t

⋅

⋅

1

B0 ∫0 k (t − t A )a hf t A h −1dt A = B0 kt a f t h h ∫0 (1 − α )a α h −1 dα

(13.35)

in which the definite integral on the right-hand side, a constant for given a and h, independent of time, is the well-known beta function, expressible in terms of gamma functions (Abramowitz and Stegun, 1964, formulas 6.2.1, 6.2.2). The resulting shape factor for the full infiltration profile is (13.36) Γ (1 + a) Γ (1 + h) rz = Γ (1 + a + h) completely independent of time. The algebraic expression (Christiansen et al., 1966), rz =

1 + a + h − ah 1 + a + h + ah

(13.37)

is a close empirical fit over the entire range of 0 < a < 1 and 0 < h < 1. At very small advance times, when the infiltration volume is negligible compared to the surface volume at constant depth, h = 1, and rZ reduces to rZ =

Γ (1 + a ) Γ (1 + h) Γ (2) Γ (1 + a ) Γ (1 + a) 1 = = = Γ (1 + a + h) Γ (2 + a) (1 + a )Γ (1 + a) 1 + a

(13.38)

confirming the limiting values on the left of Figure 13.5. At very large advance times, the surface volume is negligible compared to the infiltrated volume, and a + h = 1 (Hart et al., 1968); then, (13.39) Γ (1 + a)Γ (2 − a) π a(1 − a ) = rZ = Γ (2) sin(π a )

460

Chapter 13 Hydraulics of Surface Systems

(derived from Abramowitz and Stegun, 1964, formula 6.1.17), a function of a alone and confirming the values approached at large times in Figure 13.5. The more general Modified Kostiakov infiltration formula (Equation 13.12) can also be coupled to power-law advance to yield

(

V z = f ⋅ t h rz1 kt a + rz 2bt + rz 3c rz1 = rz of eq. 36/37;

rz 2 =

)

1 ; 1+ h

in which : rz 3 = 1

(13.40)

The power-law approach in this section can be utilized either without any assumptions regarding the average surface-water depth, as in the inverse problem of estimating the Kostiakov parameters in Section 13.7, or as an approximation to the solution of the Lewis and Milne equation with an assumed constant average surface-water depth, as follows. 13.5.6.1 Two-point method for advance. Assumption of mass conservation with a constant average surface-water depth and approximate power-law advance provides an approximate alternative to the other hydrologic methods of estimating advance times, ~ given infiltration in the form kτa, slope, roughness, and inflow rate (leading to AY). For example, in a field of length L, if advance times to the halfway point tL/2 and endpoint tL were known, the entire advance curve, f · th, would be known. Then, simultaneous solution of three nonlinear transcendental equations for tL/2, tL, and h solves the advance problem. With Q0 and xA known at the two unknown times, the two corresponding values of VZ are known as functions of the two times; the pertinent simultaneous equations are: ⎛L⎞ G1 = B0 kt L / 2 a ⎜ ⎟ rZ (a, h) − VZ L / 2 = 0 ⎝2⎠ G2 = B0 kt L a L rZ (a, h) − VZ L = 0 G3 =

in which

(13.41)

1 h t L −t L / 2 h = 0 2 ~ L VZ L / 2 = Q0 t L / 2 − AY 2 ~ VZ L = Q0 t L − AY L

(13.42)

For solution of the system by the Newton-Raphson procedure, the required derivative of rz with respect to h is most simply found by differentiating the Christiansen expression (13.37). 13.5.7 Hydrologic Modeling —Conclusions A fundamental goal of surface irrigation simulation is to predict the distribution of infiltrated water along the length of run. If the advance and recession of the surface stream were known, the opportunity times for infiltration would be known, and with a known infiltration function independent of flow depth, the problem would be solved. For infiltration that is dependent only on wetting time and not upon flow depth, the Lewis and Milne integral equation is an exact statement of mass conservation during irrigation-stream advance, and if some way were found to estimate the variation in average surface-water depth with time, it would not lead to the kind of errors exhibited

Design and Operation of Farm Irrigation Systems

461

in Figure 13.2. Most existing methods for solution of the equation, however, have assumed that the average depth of surface water is a known constant. The inaccuracies engendered by this assumption have led to the development of methods for simulating advance by utilizing the laws of open-channel hydraulics governing the flow in the surface stream, as in the next section. It is possible to examine other phases of a surface irrigation, for example, depletion, runoff, and recession by making plausible assumptions regarding the surfacewater profile and infiltration rates (Shockley et al., 1964; Wu, 1972; Strelkoff, 1977; Singh and Yu, 1989a,b) and solving an equation of continuity for mass conservation. But with the generally increasing tenuousness of the necessary assumptions on streamprofile shape for the phases of an irrigation following advance, and the widespread availability of fast personal computers, these approaches have, by and large, been supplanted by simulations—numerical solutions of the governing hydraulics equations, which are equally applicable to all phases of an irrigation event. Still, the simplicity of the hydrologic approach suggests using it to solve the inverse of simulation, for field-parameter estimation and for design. Data for the appropriate assumptions, then, could be gleaned from parallel hydrodynamic solutions, rather than guesswork. Likely there will be a continuing interest in this oldest attempt at simulating surface irrigation.

13.6 HYDRODYNAMIC MODELING OF THE SURFACE IRRIGATION PROCESS 13.6.1 Physical Laws of Open Channel Hydraulics— One-Dimensional Solution Grid The major errors that can be introduced by an assumed shape and depth of the surface stream have led to calculating its length and depths on the basis of conservation of mass and momentum applied over a series of time steps to elements of the surface stream and infiltrated water. With this approach it is unnecessary to assume that infiltration is independent of furrow depth (wetted perimeter), for stream depth is now a solution variable. But both infiltration and drag must still be estimated empirically. Recent theoretical developments in turbulent flow calculations, and techniques for coupling unsaturated flow of soil water to flow in the surface stream (Tabuada et al., 1995; Bradford and Katopodes, 1998; Skonard, 2002; Zerihun et al., 2005) are making such purely empirical contributions unnecessary, but these methods are not yet practical or sufficiently accurate in their theoretical infiltration results for routine calculations. In a typical hydrodynamic approach, the surface stream is subdivided into enough slices, or cells, that the shape of each cell is closely approximated by a trapezoid, as in Figure 13.6 (for the tip cell, a power-law shape is often assumed); the computational

Figure 13.6. Subdivision of stream length into cells.

462

Chapter 13 Hydraulics of Surface Systems

L J

R M

Figure 13.7. Computation network in x-t plane.

field expands as the stream lengthens. The total simulation time is divided into a sufficient number of intervals that the variation of pertinent variables over a time step can also be considered linear (Figure 13.7). The distance and time steps are often program controlled, with the user supplying a target cell density, the approximate number of distance steps to cover the entire length of run. In this case the time steps constitute part of the solution for stream advance. The cell faces—nodes in the solution grid—are best fixed with time along the length of run, as the stream increases in length, in order to avoid interpolation in the infiltration profile from time step to time step3. The nodes do not need to be equidistant, but should coincide with field locations where there is a break in slope, or step change in infiltration or roughness parameters. At such locations, a double node is needed because of possible step changes there in surface or infiltrated depth. Similarly, cutback and cut-off times or the start of a surge should coincide with a solution time level. When the time step size is given during the advance phase, the advance increment constitutes part of the solution. The goal is always that the variables change continuously and gradually over a time or distance step, without breaks in gradient, for then a linear approximation within the cell is justified. 13.6.2 Mass Conservation In Figure 13.8, the solid lines represent the cell profile at the beginning of a time step, the dashed lines at the end of the time step. The cells for expressing mass conservation include both the surface water and infiltrated water, and are bounded at the top by the actual water surface, at the bottom by the depth of infiltration (averaged over the transverse direction), and at the front and rear by faces normal to the channel bottom. 3

In the mass-balance computations, interpolation errors in the calculated infiltrated volume can lead to computed negative surface depths, when these are small compared to infiltrated depths, and an aborted simulation (Strelkoff et al., 1998).

Design and Operation of Farm Irrigation Systems

463

Figure 13.8. Mass conservation in a cell of surface and infiltrated water. Solid line: beginning of time step; dashed line: end of time step.

The net inflow of mass into a cell must equal the increase in its mass. In terms of the variables in Figure 13.7 and 13.8, the mass balance for a time step δt is (with constant mass density),

[θQL + (1 − θ )QJ ]δt − [θQR + (1 − θ )QM ]δt =

{[φ y L Ay L + φ y R Ay R + φz L Az L + φz R Az R ] − [φ y J Ay J + φ y M Ay M + φz J Az J + φz M Az M ] } δxLR

(13.43)

The weighting factors θ andφ reflect any departures from a straight-line variation of the variables with time or distance (the trapezoidal rule calls for the value 1/2). The factor θ is given a value a little greater than 0.5 even when time steps are small enough that the time variation is well approximated by a straight line. This proves necessary to prevent instability of calculations proceeding over a large number of time steps. Otherwise, numerical schemes like Equations 13.43 and 13.45 (below) magnify small numerical errors of truncation and roundoff with each succeeding time step, so that ultimately, these oscillatory errors completely mask the true solution (see, e.g., Cunge et al., 1980). 13.6.3 Momentum Conservation Unlike mass, momentum can be created. The impulse, or time integral, of the resultant of all the forces acting on a water-filled cell increases momentum in the same amount. That increased momentum either builds up inside the cell, or flows out through the cell boundaries. The forces acting on the surface water in the cell in the direction of flow are pictured in Figure 13.9. These are the down-slope component of fluid weight, drag of the walls and bottom and any vegetation in the flow, and the difference in hydrostatic pressure forces on the left and right faces. In a nonprismatic furrow, the pressures on the cell faces are augmented by an inline component of the water pressures on the flaring side walls. The weight component is equal to the volume of water in the cell, multiplied by γ, the unit weight of water, and the slope S0 of the channel bottom. The drag is expressed through the friction slope Sf. The net hydrostatic pressure force on the cell water, including the contributions of nonprismatic side walls, is most simply expressed through

464

Chapter 13 Hydraulics of Surface Systems

Figure 13.9. Momentum conservation: forces acting on cell of surface water.

the average cell cross-sectional area multiplied by the difference in water depth at the two faces and the unit weight of water. In fact, the sum of pressure and gravitational forces is well described by the average cell cross-sectional area multiplied by the difference in water-surface elevations at the front and rear faces of the cell. This formulation insures that standing water in a trough of any shape will be computed with a horizontal surface. Thus, the impulse δIF of the resultant force acting on a cell of water in the direction of flow over the duration of a time step is

δIF = [θ (φyL AyL + φyR AyR)(hL – hR – Sf LR δxLR) + (1 – θ) (φyJ AyJ + φyM AyM)(hJ – hM – Sf JM δxJM)] γδt

(13.44)

in which h is water-surface elevation. Just as each element of volume possesses mass, each element of volume δV possesses momentum in the amount, ρVδV, oriented in the same direction as the vector V. Thus, momentum enters and leaves the cell through its upstream and downstream faces; the infiltrating water possesses virtually no momentum. The buildup and flux of momentum δM for the cell is given by the following equations, paralleling Equations 13.43,

δM = [θρVR QR + (1 − θ ) ρVM QM ]δt − [θρVL QL + (1 − θ ) ρV J QJ ]δt

(

)

(

)

+ φ y LVL Ay L + φ y RVR Ay R ρδx LR − φ y J V J Ay J + φ y M VM Ay M ρδx JM + δM I

(13.45)

The last, negligibly small term represents momentum carried out of the surface stream by infiltration. On the basis of energy considerations and reasonable physical assumptions (Strelkoff, 1969), the momentum flux δMI contributed by infiltration in a time step is ρ ⎛ ρ ⎞ δM I = ⎜θ VLR W ⋅ iLR δx LR − (1 − θ ) V JM W ⋅ i JM δx JM ⎟δt ≅ 0 (13.46) 2 ⎝ 2 ⎠ The first overbar in each expression is the average velocity in the cell; the second is the average volumetric rate at which water infiltrates out of the cell per unit length, consisting of a representative wetted perimeter W and a representative infiltration rate i per unit area. Ultimately, the momentum equation for each cell is

δI F = δM

(13.47)

Design and Operation of Farm Irrigation Systems

465

13.6.4 Reduction to Partial Differential Equations With specified roughness and infiltration characteristics and given inflow rate and downstream boundary condition (advance, blocked end, or free overfall), the nonlinear algebraic equations, 13.43 and 13.47, are ready for simultaneous numerical solution for a given time step, over all of the cells. To orient the reader, however, it is worth noting that they can be reduced to the momentum-conservation form of the partial differential equations of open-channel flow (Saint Venant, 1871) by dividing both equations by γδxδt and reducing δx and δt to zero. In the limit, ∂Q ∂Ay ∂Az + + =0 ∂x ∂t ∂t 1 g

⎡ ∂ ⎛ Q2 ⎢ ⎜ ⎢⎣ ∂x ⎜⎝ Ay

⎞ ∂Q V ∂Az ⎤ ⎡ ∂y ⎟+ ⎥ + Ay ⎢ − S 0 + S f + ⎟ ∂t 2 ∂t ⎥ ⎣ ∂x ⎠ ⎦

⎤ ⎥=0 ⎦

(13.48)

which reduces, with standard calculus, to the traditional non-conservative form, Ay

∂V ∂y ∂y ∂A + VB + V Ay xy + B + z = 0 ∂x ∂x ∂t ∂t

1 ∂V V ∂V V ∂Az ∂y = − S0 + S f + − g ∂t g ∂x 2 g ∂t ∂x

(13.49)

in which B is the top width at the depth y, and Ay xy is the rate of increase in Ay with distance with the depth held constant in a non-prismatic furrow; in a prismatic furrow, it is zero. 13.6.5 Initial and Boundary Conditions (Saint Venant) Initial conditions for these equations, at t = 0, are zero depth and discharge everywhere. Upstream boundary conditions are usually a given inflow hydrograph, with the solution yielding upstream depths. Alternately, upstream depths or water surface elevations in a supply ditch can be given, and the resulting inflow calculated. During advance, the downstream boundary condition, at the wave front, is zero depth. Flow velocities approach a finite limit there (Whitham, 1955; Sakkas and Strelkoff, 1974), so the discharge at the front is also zero. When advance is complete, at the downstream field boundary, if the flow is blocked, discharge is zero, and depth is calculated. For free overfall into a drainage ditch, the flow is at critical (local Froude number is unity); if the water level in the drainage ditch is above the inundated soil surface by an amount greater than critical, that depth is the boundary condition, and the discharge is computed. To avoid the need for specifying a depth, Sakkas et al. (1994) assume continuing advance into a hypothetical field extension, the resulting discharge at the actual field end constituting the runoff hydrograph. At the trailing edge in recession, depth and discharge are both zero. 13.6.6 Solution of the Complete Saint Venant Equations The acceleration terms (δM in Equations 13.47; the terms divided by g in Equations 13.48, 13.49) make the equations hyperbolic, i.e., wave equations, whose physical behavior is truly mirrored only in a solution based on intersections of two families of

466

Chapter 13 Hydraulics of Surface Systems

characteristic curves (Garabedian, 1964; Sakkas and Strelkoff, 1974; Strelkoff and Falvey, 1993). As noted, subject to minor damping and phase shift, Equations 13.43 and 13.47, written in a Preissman scheme (Cunge et al., 1980; Walker and Skogerboe, 1987), are ready for solution without the need for any characteristic equations. Solution of the St. Venant equations provides an accurate solution, sometimes numerically sensitive, yielding jagged profiles or even failing with a computation of a negative depth. Two common truncated forms of the St. Venant equations, zero-inertia approximation and kinematic-wave approximation, are used to provide a more robust simulation. The approximations find a broad range of applicability, though their limitations should be recognized. 13.6.7 Zero-Inertia Approximation The acceleration (inertial) terms in Equations 13.43 and 13.47, or 13.48 or 13.49, though relatively small in surface irrigation because of the low water velocities, can still be numerically troublesome and so reduce the robustness of a model. Justification for truncating the governing equations to a zero-acceleration or zero-inertia form is provided by quantifying the influence of the acceleration terms on surface irrigation advance, as shown in Figure 13.10. Dimensionless representation, with each physical variable viewed in ratio to a reference variable, allows a single-page overview of this influence for a wide range of Kostiakov k, Manning n, and inflow rate. In a border strip with unit inflow rate q0 and Kostiakov infiltration and Manning roughness, a convenient system is q* =

q QR

y* =

y YR

t* =

t TR

τ* =

τ TR

with

1.00

od el -in er tia m

0.10

Ze ro

Dimensionless time (t*)

a=0.5 S0* =1.0

St. Venant Model FN =0.1 FN =0.2 FN =0.3 FN =0.4 0.01 0.01

0.10

1.00

Dimensionless advance (xA*)

Figure 13.10. Effect of reference Froude number on stream advance.

Design and Operation of Farm Irrigation Systems

QR = q0

YR = y N (q0 , S 0 , n )

467

1/ a

⎛Y ⎞ TR = ⎜ R ⎟ ⎝ k ⎠

QR TR = YR X R

(13.50)

Substitution into Equations 13.48 yields dimensionless equations with just three independent parameters during advance at constant inflow: a Froude number FN at normal depth for the inflowing discharge, a dimensionless slope S0*, i.e., FN =

QR 2 g YR

and

S 0* =

S0 YR / X R

(13.51)

and the Kostiakov exponent a. Figure 13.10 (drawn from Katopodes and Strelkoff, 1977b) demonstrates the very small influence on the results if the inertial terms are discarded from the equation of motion, leaving a statement of equilibrium amongst the forces acting on the surface stream. The limiting curves, with FN = 0 were obtained with all inertial terms (those containing g, in Equations 13.48, or 13.49) simply truncated from the equations (or setting δM = 0 in Equation 13.47). The specific requirement for justification of this procedure is that the Froude number at normal depth is small enough, say less than about 0.4, a condition easily met in surface irrigation, with its small velocities. The dimensioned zero-inertia equations, then, for borders, basins, or furrows are Equation 13.43 and Equation 13.44 written for the end of a current time step and reduced to the following form, (φyL AyL + φyR AyR)(hL – hR – Sf LR δxLR) = 0

(13.52)

a simple statement of equilibrium amongst the forces acting on the water in a cell, since δM = 0. 13.6.8 Initial and Boundary Conditions (Zero Inertia) Initial and boundary conditions for the zero-inertia equations are the same as for the Saint Venant equations (Section 13.6.5), but it should be borne in mind that flows with zero inertia are infinitely subcritical, i.e., critical depth is zero (Strelkoff and Katopodes, 1977a). At a free overfall, while the theoretical velocity is infinitely great, the discharge remains bounded, and just a short distance upstream from the brink, the velocity is down to realistic levels. The common practice of specifying normal depth as the downstream boundary condition for a freely draining furrow or border is inconsistent with the zero-inertia concept. If the bottom slope is too small to warrant a kinematic-wave approach to the simulation (see Section 13.6.11), i.e., too small to assume normal depth at all points in the surface stream, then it is too small to assume it at the end. 13.6.9 Solution of the Zero-Inertia Equations With the acceleration terms truncated from the Saint Venant equations (tantamount to an infinitely great relative value of g in the partial differential equations), the characteristic directions in the x-t plane for the two families degenerate into dt/dx = ±0. This indicates that any disturbance introduced at the boundary travels instantly into the interior of the stream, felt immediately everywhere (though with ever-decreasing intensity with distance from the source, just as with the heat equation, which the zeroinertia equations mimic). Thus, the zero-inertia equations represent a two-point

468

Chapter 13 Hydraulics of Surface Systems

boundary value problem progressing with time, a natural for algebraic solution of Equations 13.43 and 13.52 at each time step (Strelkoff and Katopodes, 1977b). While most solutions of the zero-inertia equations have been numerical (Wallender and Rayej, 1985; SRFR, Strelkoff et al., 1998; SIRMOD, ISED, 1989), including the finite-element approach of Shayya, Bralts, and Segerlind (1991), a few elegant and complex analytic solutions exist (Schmitz and Seus, 1987,1990,1992). In numerical solutions, the implicit system of nonlinear algebraic equations at each time step together with the boundary conditions, constitute a closed nonlinear algebraic system. The Newton-Raphson technique is effective in obtaining a solution for y and Q at each of the interior points, plus either one or the other at upstream and downstream field boundaries, and δxA (or δt) during advance. The matrix of coefficients for the system of linear equations in the corrections to the successive approximations of the solution variables is sparse, and is effectively inverted by double-sweep techniques (e.g., Cunge et al., 1980). Either, the magnitude of the time step is given, and the advance increment is unknown, or, the advance to a given location is given, and the time step is unknown. In the latter case, the matrix of coefficients contains a column of coefficients for δt. An extension of the original double-sweep technique is available for the solution (Strelkoff, 1992). Convergence of the Newton-Raphson iterations is good most of the time, but depends to some extent on the initial guesses. Furthermore, in poorly posed regions of solution (e.g., very slow advance), the corrections developed by the technique can be too large—and negative—for the small depths, and the simulation aborts. Usually, satisfactory initial guesses are simply the values at the beginning of the time step. For advance, a more satisfactory first guess comes from approximate mass-balance considerations (SRFR, Strelkoff et al., 1998). In general, if the correction terms prove so large as to disrupt the iterations, they are arbitrarily limited to a fraction of the prescribed values. However the same fraction is applied to all the terms in the correction vector so that its (Newton-Raphson) direction remains unchanged (“backtracking,” Press et al., 2001). Convergence, even if slow, is sufficient justification for the procedure. Sometimes the nonlinear algebraic equations are linearized and the solution vector at the end of δt found in one step (Cunge et al., 1980). Theoretically, the procedure, which amounts to just the first iteration in the Newton-Raphson procedure, is justified if the solution is continuous and δt is small enough, but pursuing the successive iterations until the equations are satisfied to a prescribed tolerance ensures that continuity and momentum are indeed satisfied in each cell at every time step (SRFR, Strelkoff, 1998). 13.6.9.1 Nearly stopped advance. As noted in Section 13.2.2, very slow advance is a poorly posed problem, with premature recession a likely result in the course of the iterations. An approach taken in SRFR (Strelkoff et al., 1998) is applicable both to numerical wavering of advance and physical front-end recession stemming from a temporary insufficiency of inflow. Upon calculation of negative advance or a negative depth just behind the wave front, the computational boundary is backed up one cell. The downstream boundary condition is set to a small depth, through which outflow spills out and infiltrates onto the soil ahead. When the depth of this infiltration ahead of the front is large enough in one time step to satisfy the infiltration equation for the time step, the computational boundary is moved one cell downstream again. In this

Design and Operation of Farm Irrigation Systems

469

way, advance can halt and restart, perhaps many times, as observed in the field with nearly stopped advance. 13.6.10 Kinematic-Wave Approximation While the zero-inertia approximation is valid provided only that the Froude number of the flow is low enough, it becomes increasingly inefficient to implement as the steepness of the bottom slope increases. Then the flow profile is almost uniform along the length of the stream, and only at the end makes a sharp down turn to zero depth at the very front of the wave. In order to apply the trapezoidal rule to the numerical integration over the slices comprising the surface stream, they need to be made smaller and smaller toward the front. A typical numerical response to too coarse a grid near the wave front is a blip or spike at the front, which if sufficiently great is accompanied by a negative depth just behind, which aborts the computation. To avoid the problem with the x-nodes stationary in a full hydrodynamic or zero-inertia context, all cells must be small. At the same time, with a large bottom slope, the forces stemming from the pressure difference on the faces of a cell are dwarfed by both the component of stream weight down the slope and the drag resisting the flow. The stream is thus essentially at normal depth—for the local discharge—everywhere along its length. This is sufficient to meet the conditions for the existence of a kinematic wave (Lighthill and Whitham, 1955), viz., a unique relation between discharge and cross-sectional area at any location in the stream, specifically, in open-channel hydraulics, S f = S0 =

Q2

(13.53)

K2

in which K is the conveyance of the flow channel (Equation 13.5), a function of x and Ay, or Ay alone in a uniform, prismatic channel. Substitution of Equation 13.53 (the second of Equations 13.48 with ∂y/∂x and acceleration terms deleted) into the first of Equations 13.48 (continuity) yields the kinematic wave equation, ∂Ay ∂t

+w

∂Ay ∂x

=−

∂Az − QxAY ∂t

(13.54)

In a nonuniform channel, the relation between discharge and area is dependent on location x, so that w=

∂Q( Ay , x ) ∂Ay

while

A

Qx y =

∂Q( Ay , x ) ∂x

(13.55)

describes the nonuniformity, whether variable slope, cross-section, or roughness. In a A uniform channel, Qx y = 0, and w=

S 0 dK ( y ) dQ = dAy B( y ) dy

(13.56)

Kinematic-wave solutions have been obtained analytically or semi-analytically, depending on the infiltration function selected, for example, by Sherman and Singh (1982), Singh and Ram (1983), and Weir (1983, 1985). A theoretically correct numerical solution to Equation 13.54 would follow a single family of characteristic

470

Chapter 13 Hydraulics of Surface Systems

curves in the x-t plane, each curve with inverse slope w(x,t) (Garabedian, 1964; Smith, 1972; Strelkoff, 1985), but at the expense of some numerical attenuation (Hromadka II and DeVries, 1988), a practical numerical solution to the kinematic-wave approximation can be obtained in terms of the same rectangular grid (Figure 13.7) as with the zero-inertia and full hydrodynamic forms. Analysis of the characteristics solution discloses essential elements of kinematic-wave behavior that can be accommodated on a rectangular grid. For example, no independent downstream boundary condition is possible; conditions at the downstream end of the channel are governed entirely by circumstances upstream. The front of the wave in advance constitutes a step in depth (and cross-section area AyF), related to the discharge there, QF, through Equation 13.53. The velocity of the front, wF , advancing on a dry bed is given by wF =

QF AyF + AzF

(13.57)

with AzF non-zero only if the constant c term in the cumulative infiltration function of time, Equations 13.11 and 13.15, is not zero. In addition, in Equation 13.43, underscoring the importance of upstream conditions in the surface stream, φYL is set to unity, and φYR is set to zero. Finally, a true kinematic-wave solution, in contrast to other solutions, shows instantaneous reduction to zero depth at the upstream end upon cutoff. Downstream from this point, the depths fall gradually, leading to a feathered trailing edge. A finite-difference approximation to Equation 13.54 on an x-t grid can indicate a nonzero lag time and a gradual approach to zero depth at the upstream end, but this result is due to numerical-approximation error, and represents neither a true kinematicwave solution nor the corresponding zero-inertia or Saint Venant solution. 13.6.11 Applicability of the Kinematic-Wave Assumption Kinematic-wave theory is useless in level basins or furrows. When S0 = 0, Equation 13.53 becomes untenable, and normal depth does not exist. Instead, the depth at the inlet of constant inflow rises indefinitely, until cutoff. With nonzero, positive slopes, Figures 13.11 and 13.12 illustrate conditions under which the kinematic-wave approximation yields results close to solutions of the full Saint Venant equations, or zero-inertia approximation. The dimensionless variables depicted are the same as those in Equations 13.50 and 13.51. To orient the reader in terms of typical dimensioned variables, say in an irrigated border strip, with inflow q0 = 4 Lps/m, S0 = 0.001, n = 0.04, k = 30 mm/hra, a = 0.5, it follows that the reference slope,YR/XR = 0.0000625, and the relative slope, S0* = 16. From Figure 13.12, with XR = 671 m, the error in the time to advance to 100 m using kinematic-wave theory would be about 5% high. Figure 13.13 gives an indication of how much time must pass, as a function of relative slope S0*, before normal depth is achieved at the upstream end. Kinematic wave theory assumes that at every point in the surface stream, the depth is instantaneously at normal for the discharge there. Clearly, a long run on a steep slope is the ideal application. The full complement of curves of the type of Figure 13.12 with a in the range, 0.1 < a < 0.9, presented in Katopodes and Strelkoff (1977b) shows that a value of S0*xA* > 20 is required for small errors in advance. It follows that the ratio of bottom slope to the quotient of normal depth and stream length should exceed 20, i.e.,

Design and Operation of Farm Irrigation Systems

471

od el Ki

ne m

at tm icwa od el ve

m

a=0.5 FN =0.4 S0* =100

Sa int

m od el

Ve na n

0.10

Ze ro -in er tia

Dimensionless time (t*)

1.00

0.01 0.001

0.010

0.10

Dimensionless advance (xA*) Figure 13.11 Comparison of simulation models in stream advance.

1.0

0.10

a=0.5

10 0 10 (K in 1. em 0 at 0. ic1 wa 0. 01 ve 0.0 so 0 lu 0.0 1 tio S 001 n) 0 *= 0.0 00 01

Dimensionless time, t*

10

* S0

0.01 0.01

0.1

1.0

10.

Dimensionless advance, xA*

Figure 13.12. Influence of dimensionless bottom slope on advance: applicability of kinematic-wave model.

Chapter 13 Hydraulics of Surface Systems

Dimensionless upstream depth, y0 *

472

1.0

S0*=100 10 1.0 0.1 0.01 0.001

0.1

0.0001 S0*=0.00001

a=0.5

0.01 0.01

0.1

1.0

10.

Dimensionless time, t* Figure 13.13. Growth of upstream depth as a function of relative bottom slope.

_

Surface shape factor, ry=y/y0

1.0 a=0.5

S0*=100

0.9

10

0.8 1.0 0.1

0.7

0.01 S0*=0.00001

0.6

0.01 0.01

0.1

1.0

10.

Dimensionless advance, xA* Figure 13.14. Evolution of surface stream shape factor as a function of relative bottom slope.

y0 S 0 < x A 20

(13.58)

for reasonably accurate computations of advance. 13.6.12 Average Depth of Surface Flow in Border Strips and Basins Dimensionless forms of the zero-inertia equations, 13.43 and 13.52 (FN = 0), were solved for stream advance with a series of S0* (Katopodes and Strelkoff, 1977b). The more or less gradual rise in upstream depth with time is noted in Figure 13.13, while

Design and Operation of Farm Irrigation Systems

473

the variation in surface-profile shape factor (Equation 13.26, Section 13.5.1) with advance is shown in Figure 13.14. Such curves constitute a criterion against which the assumptions yielding an average stream cross-section (Section 13.5.3) can be measured. In particular, an examination of the conditions of Figure 13.1 and 13.2 in the light of Figure 13.13 explains the sometimes significant discrepancy between field observations and the simplified theory of 13.5.1 and 13.5.3. While Figures 13.13 and 13.14 were drawn for Kostiakov a = 0.5, and cannot be expected to represent conditions for other values, a qualitative appreciation of the variations is nonetheless possible. Likewise, similar considerations can be expected to hold for furrows. Recently, a dimensionless database of average surface-water depths in border irrigation has been developed for use with hydrologic models (Monserrat and Barragan, 1998). 13.6.13 Modeling the Two-Dimensional Surface Irrigation Process Some surface irrigation flows, for example in wide basins or a border strip with cross slope, do not possess such a dominant longitudinal direction that transverse velocities and variations in water-surface elevation are negligible. In this case, the basic physical laws on which simulations are based are still conservation of mass and momentum with empirical infiltration and flow resistance, but now, these must be expressed in terms of two lateral coordinate directions, x1 and x2, as well as time t. To limit the study to two space dimensions, the pressure distribution in the vertical is again assumed hydrostatic. The velocity distribution in the vertical is assumed essentially uniform, represented by an average flow velocity vector V (x1 , x2 ) oriented at an angle to the x1 and x2 directions. The drag vector was discussed in Section 13.3.2; infiltration is assumed solely a function of opportunity time. The pertinent equations have been solved in the context of dam-break flooding, a problem similar to surface irrigation on a greater scale, by finite-element techniques on a plane surface (Akanbi and Katopodes, 1988), and by finite differences applied to arbitrary, convenient grid cells over measured topography (Bellos et al., 1991). In direct application to surface irrigation, Playán et al. (1994) discretized the full Saint Venant equations with isotropic resistance and a level bottom in a centered leapfrog scheme. For computational convenience, to allow solution of the wave equations everywhere (in a through calculation) rather than just in the region covered by the advance of the stream, a small depth was postulated in the nominally dry areas, with the implied assumption that the advancing front constituted a hydraulic bore. In Playán et al. (1996), the scheme was modified to accommodate an irregular soil surface. Strelkoff et al. (2003a) solved zero-inertia equations on irregular topography, in good agreement with measured advance and surface-depth hydrographs (Clemmens et al., 2003). The parabolic nature of the governing equations allowed a through computation in both wet and dry regions without assumption of a leading-edge bore (analogous to heat propagation in an insulated rod, suddenly heated at one end [Carslaw and Jaeger, 1959]). To assure that water was not calculated as flowing from a higher dry point to a lower point, a potential consequence of a through computation, water levels were monitored and flow allowed only from wet cells. Likewise, infiltration was allowed only from cells with sufficient water depth not to go dry in the course of the time step.

474

Chapter 13 Hydraulics of Surface Systems

13.6.14 Non-Dimensional Overviews of Surface irrigation Simulations and Performance As noted earlier, expression of governing equations, boundary conditions, and solutions in dimensionless form led to simplifications allowing general overviews of results. Figure 13.4, on a single sheet of graphs, shows all possible advance curves in border strips, subject to the assumptions that the average surface-water depth is constant with time and that cumulative infiltration follows some power law with time. Figure 13.12 is one of just nine figures (one for each value of Kostiakov a = 0.1, 0.2, ..., 0.9), which show dimensionless pre-cutoff advance x*(t*) in sloping borders without restrictive assumptions. Surface water depth is allowed to vary with time and distance in accord with the physical laws governing its behavior and, so, is a function of relative slope S0*. Once it is recalled that pre-cutoff advance in a border strip is a function of a, k, n, q0, and S0, it is evident that to present the same information in dimensioned graphs of xA(t) with five independent parameters would require shelves of books of graphs, and the advantage of nondimensional representation is clear. There are many different possible systems of reference variables, each leading to results that point up one or another aspect of the irrigation hydraulics (see Strelkoff and Clemmens, 1994, for an extensive discussion of various schemes). The space-saving opportunities afforded by use of nondimensional variables have led to development of databases of pre-run simulations encapsulating overviews of a particular phenomenon. For example, an overview of irrigation performance in level basins in general has led to the design-aid software, BASIN (Clemmens et al., 1995a) that allows a user to see virtually instantaneously what physical dimensions and operating conditions lead to given application efficiencies. At the heart of BASIN are tabulations of DUmin (the same as PAEmin) as a function of Kostiakov a and dimensionless length and unit inflow rate based on the primary reference variables Dreq, τreq, and Manq 0 (Lps/m)

30.0

10.0 Limit line E

Completion of advance

A C

70

60

DU

80

90

1.0

mi n

a = 0.5

95 (PA

E)

50

B

100

L (m)

1000

D

L* =

L ⎛ Cu ⎞ ⎜ ⎟ ⎝ n ⎠

2

3

7

Dreq9

2

τ req3

Figure 13.15. Potential application efficiency (DU) in level basins. Dimensioned and dimensionless scales: length, unit inflow. Kostiakov a = 0.5. (a) limiting length for DU = 80%: 229 m; (b) practical length at DU = 80%: 200 m, width = 63 m; (c) practical width = 75 m; DU = 81.1%.

Design and Operation of Farm Irrigation Systems

475

ning n, field variables assumed known at the onset of the design process. Figure 13.15, drawn from Strelkoff and Clemmens (1994) illustrates the case for a = 0.5 (except the target depth in this case is set to 100 mm, the time to infiltrate that depth, 210 min, and Manning n = 0.15 m1/6; see sample level-basin design below). Every point on the contours is characterized by a minimum depth of infiltration in the basin of 100 mm. Two sets of scales are shown: dimensionless ones, constituting the permanent, hardwired database and ordinary, dimensioned ones for a particular set of field conditions. For a given field, the two sets differ by constant multipliers (reference variables XRc and QRc, Strelkoff and Clemmens, 1994). The curve labeled completion of advance represents the operating condition of cutoff at the time the stream reaches the downstream end. The curves show that for a given length of basin, performance can be enhanced by increasing the unit inflow while initiating cutoff before completion of advance, allowing the stream to “coast” to the end after cutoff. The steepening of the DU contours at high inflow rates, however, shows that there is a limit to this procedure. Clemmens and Dedrick (1982) suggested practical limitations on the process, arguing that overreliance on coasting to the end can leave that end high and dry if field conditions are not precisely known; their suggested limits are shown as the limit line. 13.6.14.1 Sample level basin design. Drawing on the example physical design (p. 4-1, Clemmens et al., 1995a), a designer seeks to achieve a target DUmin of 80% in each basin in a field with the aforementioned infiltration and roughness properties, dimensions 600 m wide and 1200 m long, and an available water supply of 230 Lps. He or she first looks for the limiting length. Point A in Figure 13.15 depicts that operating point, corresponding to a length of 229 m (the corresponding unit inflow, q0 = 3.65 Lps/m leads, with the given total available supply of 230 Lps, to a width of 63 m). With 1200 m evenly divisible by 200, that is chosen for the basin length. Point B in Figure 13.15 is the operating point for a length of 200 m and a DUmin of 80%. This corresponds to a width of 84 m. Points between D and E in Figure 13.15 (where D is an indeterminate operating point of very low DUmin, and E represents a very determinate limiting width) are available to the designer for selecting a practical width. Eight basins, each of 75 m width and 200 m length, L, evenly divide up the available field and lead to a DUmin of 81.1% (operating point C). The appropriate cutoff time, calculated from the simple volume balance, tco q0 = LDreq /DUmin, is 134 minutes. The dimensionless advance curves stored in the BASIN database yield the corresponding advance at cutoff, 176 m, a useful guide in inflow management (Clemmens, 1998). The user-friendly BASIN software allows menu-driven data input (dimensioned) for field conditions and design objectives and then transparently performs all the conversions between dimensional and dimensionless data evident in Figure 13.15, as well as the necessary interpolations and calculations to satisfy the given target depth by the minimum in the post-irrigation distribution of infiltrated depths. The user never gets to see the curves depicted in Figure 13.15; they remain in the background. 13.6.14.2 Sloping border strips with tailwater runoff. Sloping borders with tailwater runoff are governed by a larger number of input parameters than level basins, and a different strategy was employed for the dimensionless database lying at the heart of the design-aid program, BORDER (Strelkoff et al., 1996; Strelkoff and Clemmens, 1996a). The reference variables, QR, YR, XR, TR, are, respectively, the unknown unit inflow rate, normal depth at the given slope and roughness, normal depth divided by bottom slope, and XR YR /QR. Several thousand dimensionless zero-inertia simulations

476

Chapter 13 Hydraulics of Surface Systems

were run (Shatanawi and Strelkoff, 1984; Strelkoff and Shatanawi, 1984) with Kostiakov infiltration, i.e., with a range of K* = kTRa/YR, a, and dimensionless cutoff times in hypothetical borders indefinitely long, so that advance ends before reaching the border end. “Infiltrated” volume past the actual end of the border strip then closely represents actual runoff. Within the border length, a stored shape factor for the post-irrigation infiltration profile allows determination of the minimum or low-quarter-average depth. In use, e.g. to aid in physical design, a grid of border lengths and widths spanning the range of interest of the design variables is established, and successive approximations utilizing interpolations within the data base yield the grid of cutoff times that just satisfy the target depths, as well as a series of performance parameters. The software presents performance contours as a function of field conditions, physical design, and operating conditions. In this case, the user does not have to set a desired performance level a priori; the contours show what the maximum performance level can be for the given field, available water supply, and target depth, as well as how to achieve a particular level by selection of design parameters. Figure 13.16a (Strelkoff and Clemmens, 1996b) shows sample contours of potential application efficiency for a field with the following characteristics: Slope S0 = 0.001 Kostiakov k = 57.4 mm/hra a = 0.5 Target depth: Dreq = 100 mm n = 0.10 m1/6 Supply: Q = 60 Lps Manning

Figure 13.16. Contours of low-quarter potential application efficiency as a function of border length and width for different soils. Kostiakov a = 0.5, Manning n = 0.10 m1/6, bottom slope = 0.001, available water supply = 60 Lps, target depth (a-c) = 100 mm; (a) Kostiakov k = 57.4 mm/hra; (b) sandier soil, k = 80 mm/hra; (c) tighter soil, k = 4 0 mm/hra; (d) k = 40 mm/hra, target depth = 75 mm.

Design and Operation of Farm Irrigation Systems

477

A companion graph (not shown) presents the corresponding pattern of cutoff times required to achieve the performance. Additionally, contour maps of distribution uniformity, runoff, deep percolation, stream advance at the time of cutoff, and water cost per unit area can be displayed. The contour plot shows a ridge of relatively high potential application efficiency (PAE) designs, with the lower unit inflow rates (larger widths) corresponding to the shorter lengths of border. The slopes near the top of this ridge are gentle, implying that errors in evaluating field conditions or executing the design may not affect potential application efficiency very much; closely spaced contours would indicate large changes in performance with small changes in input parameters. The efficiency ridge climbs gradually as larger lengths, larger unit inflow rates, and shorter application times are considered. At the highest and lowest unit inflow rates, however, the contours near the ridge top are closer together than with more moderate unit inflow rates, say, around, 1.5 Lps/m (40-m width), indicating increased sensitivity to input conditions. Figure 13.16b shows the pattern for a sandier soil; with all other conditions remaining the same, the Kostiakov coefficient has increased, so that the one-hour cumulative infiltration depth is now 80 mm. Qualitatively, the same observations as made for the original soil hold here, but at shorter lengths. Conversely, Figure 13.16c, with Kostiakov k = 40 mm/hra, requires very long lengths to achieve comparable potential application efficiencies (note the change in scales). However, with a decision to irrigate more frequently, and lightly, say with the 75-mm target application of Figure 13.16d, the higher efficiencies return to the shorter border-strip lengths. Alternative charts (not shown) provide an overview of the effects of inflow management with a given physical design. 13.6.15 User-Friendly Simulation Software; Transport of Constituents As noted, the numerical tedium of solving Equations 13.43 to 13.47 et seq., as well as supplying the variety of initial and boundary conditions for the range of onedimensional surface irrigation scenarios encountered has been taken over by userfriendly, menu-driven software (SRFR 4.06, USDA, 1999; SIRMOD III, Walker, 2003). Figure 13.17 shows a frame of the automation depicting the stream behavior as calculated by SRFR for a case of furrow irrigation with erosion. Shown are the surface water profile, sediment transport capacity and load profiles, and infiltration profile 11/4 h into the irrigation. Figure 13.18 is a post-irrigation hydraulic summary of the event, showing the inflow hydrograph and resulting advance and recession curves, runoff hydrograph, and final distribution of infiltrated depths. The software also prepares a list of performance indicators, from which the merit of the irrigation can be judged (Figure 13.19), including, in this case, the off-site transport of sediment, GS_O. The software, as released by USDA free of charge, comes with some 15 sample data-input files, illustrating several scenarios. SIRMOD III contains some basic infiltration estimation capability as well as certain system design aids. Recent enhancements in SRFR have allowed researchers to utilize its hydraulic output to study transport of constituents in surface irrigation flows, e.g., in addition to sediment (Strelkoff and Bjorneberg, 2001), fertilizer in solution (Strelkoff and Clemmens, 2006, by advection alone; and Zerihun et al., 2005, and Perea et al., 2005, by advection and turbulent diffusion). Currently underway is a USDA project (WinSRFR) to port the current stand-alone DOS programs, SRFR, BASIN, and BORDER along with an irrigation and field-properties evaluation module into an integrated WINDOWS environment allowing data interchange between the various modules for input and output (USDA-ARS, 2006).

478

Chapter 13 Hydraulics of Surface Systems

SURFACE STREAM

Transport capacity

SEDIMENT TRANSPORT

Sediment load

INFILTRATION PROFILE Infiltration requirement

Figure 13.17. Frame of animation calculated by SRFR during advance with irrigation-induced erosion.

Figure 13.18. Graphical output of SRFR simulation. Hydraulic summary shows inflow hydrograph, advance, recession, outflow hydrograph, and post-irrigation infiltration distribution.

Design and Operation of Farm Irrigation Systems

479

Figure 13.19. Graphical output of SRFR simulation of furrow irrigation with erosion. Performance synopsis shows post-irrigation infiltration distribution and a series of performance indicators, including off-site sediment transport, GS_O.

13.7 ESTIMATION OF FIELD PARAMETERS Neither simulation nor design can rationally be undertaken without entry of field parameters such as infiltration and roughness. Infiltration in particular plays an enormous role in the outcome of a surface irrigation. With the variability found in the field and limited success of infiltrometers in predicting even local parameters for the conditions of a surface irrigation, it is reasonable to deduce these parameters from observations of an actual irrigation. In addition, it might be necessary to estimate infiltration and roughness in real time, i.e., during an irrigation—say during early advance—in order to predict how successful the irrigation will prove to be, and what to do about it. Generally, the greater the detail in measurement, the more precise the evaluation can be. On the other hand, detailed measurements, of surface depths of flow, for example, are time consuming and expensive. Thus, attempts are made to estimate field parameters from simpler measurements, for example, stream advance. In any case, whenever an estimate is made of field parameters to input into a simulation or design procedure, an additional estimate of likely error in each parameter should also be made. Then, the procedure can be executed for a range of input parameters, with the totality of results lying within a confidence band. In other words, the graphical results of simulation or design should not be thought of as a curve with the thickness of a pencil line, but as a broad brush stroke, with the true results lying somewhere within. It is the task of the engineer to predict not only the centerline behavior of the brush stroke, but also to estimate its thickness. 13.7.1 Mass Conservation—Volume Balance Explicit expression of mass conservation forms the basis for all of the direct methods of parameter estimation. During advance, the inflow volume must equal the volume infiltrated plus the volume in surface storage, as given in Equation 13.18. In one-

480

Chapter 13 Hydraulics of Surface Systems

dimensional flow, the surface storage is given by Equation 13.21, while the infiltrated volume is given by Equation 13.23, which contains both the advance function (xA(t) or tA(x), which can be measured) and the infiltration function (sought). Many of the direct parameter-estimation methods are based on a known VZ(t), stemming from measured inflow and outflow, and AY from either measured surface water depths or an estimate. 13.7.1.1 Post-irrigation determination of infiltration characteristics. A particularly simple and theoretically exact application of the volume balance to evaluation of the infiltration parameters stems from the post-irrigation analysis of measured advance and recession curves. With the irrigation completed, all surface water has either drained off or infiltrated, VY in Equation 13.18 is zero, and VZ is known from measured inflow, VQ, and outflow, VRO. Thus the average depth of infiltration is known. Indeed, following an initial suggestion by Merriam (1971), Clemmens (1981) developed equations for whole-border estimates of Kostiakov k, if the exponent a is known from other sources. If the basic Kostiakov formula, kτoa, is assumed, in which τo(x) is the infiltration opportunity time between the advance and recession curves, and if a is taken, for example, from ring data, k can be found from the equation for the total volume infiltrated per unit width, N

VZ = kW ∑τ o aj δx j

(13.59)

j =1

in which the border length has been subdivided into N segments and τoj is the average opportunity time for the jth segment; W is the effective wetted perimeter, assumed constant, so that AZ = Wz. Clemmens (1981) extended the technique of Equation 13.59 to the branch function of Equations 13.14 for the common case that the basic infiltration rate has been achieved within the smallest opportunity time measured. For example in a border strip, with the post-irrigation depth of infiltration d at any station dependent solely on the opportunity time there, d = kτ B a + b(τ o − τ B )

(13.60)

the average depth of infiltration over the border (known from measurement of inflow and outflow) is

(

d = kτ B a + b τ o − τ B

)

(13.61)

With a and b estimated from ring data, the cross-over point τB is found in terms of k from the third of Equations 13.14 and the result substituted into Equation 13.61, to yield the following (simplified, corrected) whole-border estimate of k, ⎛b⎞ k =⎜ ⎟ ⎝a⎠

a

⎛ d − bτ o ⎜ ⎜ 1− a ⎝

1− a

⎞ ⎟ ⎟ ⎠

(13.62)

If τB is better estimated from the ring data than b, b follows from b=

ad (1 − a)τ B + aτ o

(13.63)

obtained after eliminating k from Equation 13.61 and the third of Equations 13.14, which then yields k.

Design and Operation of Farm Irrigation Systems

481

The details of the infiltration function at opportunity times less than the observed minimum play a minor role in the distribution of infiltrated water over the given length of run, as long as the behavior of the function within the observed range of opportunity times is reasonably correct. The more nearly uniform are the opportunity times distributed over the length of run, the more exact will be the evaluation of the irrigation. At the same time, Clemmens et al. (2001a) caution that a narrow range of opportunity times in the parameter estimation can lead to large errors in predicted performance for the same soil but under different hydraulic conditions (e.g., slope, inflow rate, length, etc.); ultimate stream behavior is dependent on the entire infiltration function of time right up to the end of recession. It should also be possible to avoid independent assessment of a, if two borders with the same soil conditions are irrigated, and their variations in opportunity time are analyzed. In principle, both k and a can be solved by a Newton-Raphson solution of the nonlinear equation pair. Overconditioning the problem with more irrigations and equations would allow for a least-squares best fit for both k and a. This multiple-border post-irrigation approach is based on that of Bouwer (1957), which in the absence of a functional form for infiltration leads to a table of cumulative infiltration vs. time that may, like the results of Finkel and Nir (1960) described below, tend to oscillate. Independent determination of a is also unnecessary with soils that can be characterized by membership in empirically defined families (Section 13.3.3), e.g., the SCS (NRCS, 1984) families or the Merriam and Clemmens (1985) time-rated families. Both sets are characterized by an implied relationship between k and a, Equations 13.15 or 13.17. Then, a Newton-Raphson solution for k and a would utilize the partial derivatives of Equation 13.59 (restated in the form G1 = VZ – f [k,a] = 0) with respect to k and a, and those stemming from a restatement, say, of Equation 13.17, G2 = k

⎛ 0.675 − a ⎞ 2−a⎜ ⎟ − 10 ⎝ 0.2125 ⎠

=0

∂G2 =1 ∂k ⎛ 0.675− a ⎞ ⎟ 0.2125 ⎠

2−a⎜ ∂G2 = − ln 10 ⋅10 ⎝ ∂a

2a − 0.675 0.2125

(13.64)

Similar equations can be derived for Equation 13.15. 13.7.1.2 Parameter estimation requiring a closely spaced set of surface-depth hydrographs. At the other end of the spectrum of required field data, measurement of a complete set of depth hydrographs along the length of run during the irrigation yields the time variation of the volume temporarily stored on the surface, VY(t) in Equation 13.18. Thus, with inflow and outflow known, VZ(t) is also known. Together with the advance and recession curves, xA(t), xR(t), (a byproduct of the depth hydrographs) this information can be used to deduce the field infiltration parameters. This class of methods requires the most extensive data collection, but is encumbered by the fewest assumptions, and so is the most direct and physically based of all the techniques reflecting conditions during the entire irrigation. Based almost exclusively on mass conservation, the principle problem lies in extracting the infiltration parameters (and possibly roughness) from the measured data. The techniques, all of

482

Chapter 13 Hydraulics of Surface Systems

which assume that VY(t) is a known, measured function of time, differ in the assumptions made on the functional forms of the advance and infiltration functions. In an early development, Finkel and Nir (1960), making no assumptions on infiltration- or advance-function form, proposed a graphical inversion of Hall’s (1956) technique (Section 13.5.4). In principle, a tabulated infiltration function of time could be constructed step by step, a Δz for each Δt. However, unless great care is taken that the increments in surface volume are accurate, and the construction is limited to a small number of Δt, the results for Δz begin to oscillate. This instability evidently stems from the structure of the governing algebraic equation, the inversion of Hall’s. Any error in measurement or calculation (roundoff, etc.) must be absorbed by the current calculated Δz, and successive steps in the calculation magnify it. The forward calculation, advance, is stable with the Hall technique; the reverse, for infiltration, is not. In a direct, interactive computer procedure (EVALUE), Strelkoff et al. (1999) superimpose screen plots of measured VZ(t) and a theoretical VZ(t) calculated from estimates of infiltration parameters in a selected functional form. Variable wetted perimeter is not considered, and the unknown parameters are assumed constant over the entire length of run. Numerical integration for the volume under the infiltrated profile is enhanced by weighting factors based on the ratio of calculated infiltrated depths at each end of a profile segment and on an assumed power-law (with exponent a) for depth vs. distance back from the advancing front. The parameter values are modified at the keyboard until the user is satisfied with the fit of the two curves. In addition to the function of time, the final values of infiltrated volume are also plotted for matching. Various functional forms can be investigated simultaneously with the parameters in those forms. This brute-force approach is aimed at selecting those parameters best representative of the entire time span of the irrigation. A by-product of the calculations is a determination of Manning n values at the stations, determined from Equations 13.4 and 13.6 by calculating the local discharge from continuity and setting Sf to the water-surface slope given by smoothed profiles derived from the measured hydrographs. The average or median constitutes the representative value of n for the irrigation. Maheshwari et al. (1988) working with irrigation borders made in extensively cracked clay soils (k and a of Equation 13.11 were essentially zero in many cases) automated a similar procedure, by formally minimizing an objective function, Z*, N

Z * = ∑ (VO i − VC i )2

(13.65)

i =1

with the Hooke and Jeeves pattern-search technique. VOi is an observed infiltrated volume, based on Equation 13.16 with measured inflow and outflow rates and VY given by numerically integrated measured surface-water depths. VCi is calculated from current values of infiltration parameters and the measured advance curve. N is the number of times (separated by equal increments) that the comparison is made in the duration of the study, ending, typically, when recession begins at the upstream end. In application, the measured advance and infiltrated-volume functions of time are fitted with mathematical expressions by regression. A number of different functional forms for infiltration and advance were investigated. In 1997 (Esfandiari and Maheshwari, 1997), the method was extended to furrows. The modified-Kostiakov parameters for furrows, K, a, and B, were sought, yielding intake directly (see Section 13.3.3), so no explicit consideration of wetted perimeter in infiltration was undertaken.

Design and Operation of Farm Irrigation Systems

483

Monserrat (1994) devised a volume balance for the advance phase that utilized theoretical rather than measured values of average surface water depths. He minimized an objective function, based on a balance between inflow, surface, and infiltrated volumes in level or sloping borders calculated with measured data, assumed Manning n, and unknown k and a. Post-irrigation conditions were also included in the objective function. The optimization procedure allows solution for both k and a, as opposed to ~ the inversion technique of Clemmens (1991), as well as AY ; on the other hand, Monserrat’s static database of solutions limits the analyses to the conditions—broad as they are—of the pre-run simulations. Other techniques have been designed to extract the infiltration parameters by one or another manipulation of the integral in Equation 13.23. The best known of these is the Lewis and Milne equation. 13.7.2 Lewis and Milne Integral Equation As noted in 13.5.2, Lewis and Milne (1938) substituted a time integral for the distance integral in Equation 13.23 through a formal change in variable, leading to Equation 13.27. Some well-known applications thereof incorporate a simplifying assumption for the functional form of the advance curve. These will be described in Section 13.7.2.1. A technique free of arbitrary assumptions on the advance function is the linear station advance procedure of Clemmens (1982), who proposed direct evaluation of the theoretical VZ at a sequence of time levels ti in terms of increments j of the total profile length at that time, i.e., N

VZ (ti ) = ∑ ΔVZ j (ti )

(13.66)

j =1

with the volume increments calculated by assuming a (different) constant rate of advance over each increment ΔxA. The advance rate in Equation 13.27 can be taken outside the integral and expressed as a quotient of differences, so that

ΔVZ j (ti ) =

tA j

Δx j Δt A j

∫ AZ (ti − t A ) dt A

(13.67)

t A j −1

in which ΔtAj = tA(xj) – tA(xj-1). Now, with any given functional form for z(τ), the integral in Equation 13.67 can be evaluated for each j = 1,..., N, and the results summed for a theoretical Vz(ti) in terms of the infiltration-function parameters. For example, with the modified-Kostiakov formula of Equation 13.11, Equation 13.67 appears as

⎡ ⎛

{

} { a +1

− ti − t A j Δx j ⎢ ⎜ ti − t A j −1 k⎜ ΔVZ j ( ti ) = ⎢ a +1 Δt A j ⎢ ⎜

⎣ ⎝

}

a +1 ⎞

⎟ ⎟ ⎟ ⎠

(13.68) 2 2 ⎞⎤ ⎛ ⎜ ti − t A j −1 − ti − t A j ⎟⎥ + b⎜ ⎟⎥ + cΔx j 2 ⎜ ⎟⎥ ⎝ ⎠⎦ as can be followed in Figure 13.3. The first term of Equation 13.68 is identical to the result presented by Clemmens (1982).

{

} {

}

484

Chapter 13 Hydraulics of Surface Systems

With the branch function (Equations 13.14), if t – tA < τB, the first and third terms of Equation 13.68 are appropriate. For t – tA > τB, ⎛

{

} { 2

Δx j ⎜ t − t A j −1 − t − t A j

} ⎞⎟ + c 2

(13.69) ⎟ B Δx j ⎟ ⎠ ⎝ The form of Equations 13.68 and 13.69 allow for spatial variability in the infiltration parameters, but, typically, the equations would be used to estimate whole-border values. With two parameters, k and a in a power law, a minimum of two time levels is required for solution. For the four parameters in the Modified Kostiakov formulation of Equation 13.68, it would follow that four simultaneous equations of the type of Equation 13.67, at each of four time levels, would be the minimum. Typically, many more are used, with a best fit sought. In a computational simplification, Clemmens (1982) guesses a, say 0.5, allowing a direct, rather than simultaneous, solution for k at each t up to ti, with the same averaging of opportunity times leading to a point on the plot. The slope of the line, on logarithmic paper, provides a better guess for a, and the process is repeated until a converges (with two trials typically necessary). With the infiltration parameters found, Equations 13.68 or 13.69 provide the changes in subsurface storage in each segment in each time step, while the depth hydrographs at the stations yield the changes in surface storage. A volume balance between stations and time steps yield the discharges at each station, while the profiles derived from the water surface elevations yield the water surface slope. Thus Manning n can be determined at each station at each time level. With the median values at a station found over time, the median or distance average, as in Clemmens (1982), yields a representative value for the border strip. If the advance points leading to Equation 13.68 had been found at constant increments of time (either measured or interpolated) as in Figure 13.3, Equation 13.68, with the Kostiakov power law, z = kτa, leads to a numerical evaluation of rZ through the summation, a +1 N Δx ⎛ V − {i − j}a +1 ⎞⎟ j {i − j + 1} ⎜ (13.70) rZ i = aZ i = ∑ ⎟ i a (a + 1) kti x A i j =1 x A i ⎜⎝ ⎠ in which rZ is the ratio of average depth of infiltration to the upstream depth. The relationships tAj = jΔt, ti = iΔt, tAj – tAj-1 = Δt, etc. were used in the development of Equation 13.70. 13.7.2.1 Power-law advance. Significant simplification in evaluating the Lewis and Milne integral in Equation 13.27 follows assumption of a functional form for advance as well as for infiltration, as discussed in Section 13.5.6 for power-law advance. In fact methods for determining infiltration use the same concepts as were used for determining advance (Sections 13.5.2 to 13.5.6). The two-point method (Elliot and Walker, 1982)—In a popular variation of the Christiansen et al. (1966) approach, advance time is measured only twice, typically halfway to the end of the field and at field end. The method assumes power law advance ~ and, rather than base AY on the degree to which VZ follows a power law, makes an inde~ pendent estimate for AY based on assumed normal depth for the given cross-section, bottom slope, Manning n, and inflow rate and an assumed constant shape factor rY, as

ΔVZ j = b

⎜

Δt A j ⎜

2

Design and Operation of Farm Irrigation Systems

485

~ AY = rY AY 0

(13.71)

Then VY in Equation 13.18 is known at both xA. The two-point method shares with similar methods any inaccuracies stemming from estimation of the average crosssectional area of the surface stream (Strelkoff et al., 2003b). It also shares the small violation of mass conservation stemming from assumption of power laws for both infiltration and advance along with a constant average cross-sectional flow area during advance. Concern that one or both of the two advance measurements may be faulty can be alleviated by plotting a line of best fit to the advance curve on logarithmic paper and determining the two points from that. The two-point method is easily extended to accommodate non-zero b values in Equation 13.11, once these are determined independently. Note, however, the caveat regarding Equation 13.13. The outcome of the method, for the two points at full length, L, and half length, L/2, are the Kostiakov k and a, as follows: log 2 L h= f = h (13.72) ⎛ t ⎞ tL log ⎜⎜ L ⎟⎟ ⎝ tL / 2 ⎠ ⎛ V ⎞ log ⎜⎜ Z L ⎟⎟ V ⎝ Z L/2 ⎠ a+h = ⎛ t ⎞ log ⎜⎜ L ⎟⎟ ⎝ tL/ 2 ⎠

and

k=

g f rZ (a, h)

g=

VZ L t L a+h

(13.73)

(13.74)

Analogous equations yield K if desired for a direct empirical determination of infiltrated volume per unit length. Figures 13.20a,b illustrate the effect of an incorrect estimate of average surfacestream cross-section for two different bottom slopes, (a) S0 = 0.0005, (b) S0 = 0.005. The two-point method was applied to a simulated border irrigation. The infiltration function input to the simulation (a time-rated family member requiring 4 h to infiltrate 100 mm) is labeled SRFR; the simulation then yielded both an advance curve (t2 = tL shown) and rY, slowly varying with time. A representative value for rY was selected for use in the two-point method, and the resulting infiltration function is labeled “rY correct”. Then the representative value was changed by ±10% with the results shown. As can be expected, with the surface volume at the small slope comprising a greater fraction of the total inflow than at the larger slope, an error in rY leads to greater errors in estimated infiltration with the small slope than with the large slope. In both cases the deviations from the correct infiltration increase significantly for times extrapolated beyond t2 (Strelkoff et al., 2003b). In the event of a level field, normal depth has no meaning, but an extension of the method to this case has been proposed (Zerihun et al., 2004). Instead of assuming AY0 constant in Equation 13.71, it is allowed to grow with xA in accord with Equations 13.4 and 13.6 and the approximation that Sf = y0/xA. Then, for a given furrow crosssectional shape, AY0 is implicit in the relationship,

486

Chapter 13 Hydraulics of Surface Systems

AY 0 2 R0 4 / 3 y0 = x A Q0 2 n 2

(13.75)

with the right-hand side known at both advance distances (selected in Zerihun et al., ~ 2004, as L/3 and 2L/3). Then AY follows from Equation 13.71. To extend the method to non-zero B requires depth measurements as the ponded water surface falls after cutoff. B would be given by the measured rate of fall and the assumption that at this point in time infiltration rate does not vary with location in the basin. Alvarez (2003) utilized the assumptions of the two-point method to predict the advance and Kostiakov K for furrow discharges other than the particular one at which the field test was run, to account for the different wetted perimeter. The technique is based on the further assumption, supported by field observations, that with different inflow rates, the exponent in the power law for advance varies very little.

ry correct Infiltration depth (mm)

100.

(a)

ry 10% high

Border strip S0=0.0050

SRFR

t2

50.

ry 10% low

0. 0.00

1.00

2.00

3.00

4.00

Time (hours)

Infiltration depth (mm)

100.

(b)

Border strip S0=0.0005

SRFR

ry correct 50.

ry 10% low t2

ry 10% high 0. 0.00

1.00

2.00

3.00

4.00

Time (hours)

Figure 13.20. Effect on infiltration prediction of errors in assumed shape factor rY for surface water profile. Two-point method. t2: advance time; SRFR: z(τ) in simulation. (a): S0 = 0.0005; (b): S0 = 0.005.

Design and Operation of Farm Irrigation Systems

487

One-point methods—One-point methods require measurement of the time to advance to a single point, typically field end. It is also assumed in this method, as usually applied, that the flow channel is sufficiently steep to achieve normal depth at the upstream end by the end of advance, and that rY is known. This allows solution of the Lewis and Milne (1938) integral equation. And, while two measured advance points are generally required to yield independent values of the two Kostiakov parameters, additional assumptions can yield the pertinent parameters with just one advance point. In the original one-point method (Shepard et al., 1993), the reasonable assumption of a Philip infiltration function, z = Sτ1/2 + Aτ (Philip, 1969), and a quite restrictive advance function, x = f·t1/2, lead to both constants, S and A, from a measurement of the time to reach furrow end. The concurrent assumptions, instead, of NRCS (1984) (SCS) infiltration families and power-law advance, x = f·th, with f and h site-specific constants, lead approximately to the pertinent family, once the time for the stream to reach a specified point is given (Valiantzas et al., 2001). This is made possible by the fact that each family is defined by a particular value of k and a particular value of a; thus k is a function of a (Equation 13.15). In an alternate method, introduced because some non-cracking soils are fit better by the time-rated families of Merriam and Clemmens (1985), a similar development (see Strelkoff et al., 2003b) also leads to the Kostiakov k and a from a single advance point. In this connection, the empirical relationship between k and a (Equation 13.17) depends on the actual wetted area, and consequently, in furrows k should be based on wetted perimeter rather than on furrow spacing. The effects of errors in surface volume estimation are also discussed. 13.7.3 Direct Inversion of the Hydrodynamic Flow Equations Clemmens (1991) was able to invert the hydrodynamic unsteady-flow equations governing the irrigation-stream flow to the extent of one unknown field parameter. With measured time and distance increments given in the double-sweep algorithm for solving the linearized equations of the Newton-Raphson scheme, the column of coefficients stemming from the gradients of depth and discharge with respect to time (see Section 13.6.9) is replaced by the gradients with respect to some field parameter (Clemmens, 1991). Then that field parameter comes out in the solution at each time step. The method requires a priori specification of all the remaining field parameters, usually, Kostiakov a and Manning n. At each time step, the resultant k is used to calculate an opportunity time for the computed average depth of infiltration. Then the resulting plot is used to fit a new Kostiakov function, allowing recalculation with the more current value of a. The original paper provides a wealth of detailed comparisons of results from various combinations of assumptions in the method. 13.7.4 Parameter Optimization through Repeated Simulation When the unknown field parameters are so embedded in the pertinent equations that the latter cannot be inverted to yield them, they can be sought in a program of successive approximation. This typically involves repeated simulations, with changing values of the parameters, in a formal search procedure known as optimization, aimed at minimizing the discrepancies between simulated and measured values of selected quantities, such as advance, water depths, runoff hydrographs, etc. General concerns and approaches to securing an optimum can be found in Press et al. (2001), which explains the concepts in the following discussion with more detail and background information. In particular, they first discuss the problems encountered and techniques available in a one-dimensional search for a minimum. The global, or

488

Chapter 13 Hydraulics of Surface Systems

true, minimum of a general objective function F(x) of one independent variable x can certainly be found, within a pre-specified permissible range of the independent variable, by brute force, i.e., calculating F over sufficiently small increments, δx, that a dip lower than all the others will not be missed. The fundamental problem in practical one-dimensional optimization is in devising a technique that will lead to the global minimum via a reasonable number of calculations. In multi-dimensional optimization, in which there are a number of independent parameters, described, say by the vector p, the additional problem to be confronted is in the selection of directions in which to change p from a current estimate, as well as in how much to change it, so as to reach a global minimum, all within the permissible range of p. In general, a good optimization algorithm is robust, i.e., will inexorably converge to the global minimum, bypassing local minima, regardless of the starting guesses for the independent parameters or the nature of the objective function. It will be economical of computation and will not require excessive computer storage. In the gradient methods, the search proceeds in a series of one-dimensional steps, with the direction in each step selected in response to the gradient of F (calculated, for example, by simulation at two neighboring p). In the step from iteration i to i + 1, p( i +1 ) = p( i ) − d = p i − α ( i )R ( i )g( i )

(13.76)

the change in p is made in the direction of the vector d, given by the matrix R premultiplying the gradient g, and of length given by the scalar α. R = I, the identity matrix, corresponds to the method of steepest descent, but this can be extremely inefficient (Press et al., 2001), and other choices, twisting the direction of correction away from the gradient, prove more suitable in the search for the bottom of a long, narrow valley with spurious local minima representing the objective function (as visualized with two free p components). A logical selection for R in the search for the minimum of the F surface is the inverse H of the Hessian matrix for F, −1

⎛ ∂2F ⎞ ⎟ (13.77) H=⎜ ⎜ ∂pi ∂p j ⎟ ⎝ ⎠ for that is an indication of the local rate at which the gradient is approaching zero, i.e., where F is at a minimum. The expectation is that, except for points at the boundaries of the permissible domain of independent variables, the minimum of the objective function corresponds to a zero gradient there. Direct calculation of H is inconvenient, partly because of the difficulty in determining the second derivatives and partly because of the need for inverting a potentially large matrix. A number of numerical alternatives have been proposed, as noted in subsequent sections. 13.7.4.1 Optimization to match advance. The issue of how much information about the field parameters can be deduced from measurements of advance alone (observability of advance) was examined theoretically, in the context of a linearized, zero-inertia model, by Katopodes (1990). He concluded that resistance and infiltration have indistinguishable effects on the rate of advance, but that if one is given, the other can be found. With three field parameters controlling advance, say, Kostiakov k and a and Manning n, if any two are known, the third can unequivocally be deduced from the observed advance. Two parameters can be deduced if some measured surface depths are available for comparison. The optimization can be difficult: in the case of an objective

Design and Operation of Farm Irrigation Systems

489

function based on varying n and a, several local minima are observed in a very narrow valley, each of which is higher than the global minimum. All three parameters can be estimated with measurements of a single depth profile (Yost and Katopodes, 1998). A series of methods, also, in a sense, constituting an inversion of the advance problem, consists of repeated simulations with a mathematical model controlled by one or another search procedure. Walker and Busman (1990), with the aim of evaluating realtime infiltration, i.e., while an irrigation event is in progress, minimized an objective function, Y, based on measured and calculated advance, Y=

m

∑ (T A j − t A j ) 2

(13.78)

j =1

in which the TAj are measurements of advance time to specified locations xj, while the tAj are simulated times with specific values for the infiltration parameters, in a current estimation. The roughness parameter is assumed known. The number of comparison points, m, increases as the irrigation progresses, providing ever more data on which to base the infiltration parameters. At any step of the irrigation m, the objective function is minimized (the infiltration parameters are optimized) in successive approximations via the downhill simplex method (Press et al., 2001), not requiring calculation of the gradient of the objective function, but only of the objective function itself, at N + 1 estimates of the N infiltration parameters at each step of the iterative process, i.e., m

Yi = ∑ (TA j − t Ai j )2 j =1

i = 1,..., N +1

(13.79)

With two infiltration parameters k and a (N = 2), there are three vertices to the simplex, in general a polyhedron with N + 1 vertices in N-dimensional parameter space, reducing, in the case of Kostiakov parameters, to a triangle. With three, in a search for best values of k, a, and b, the simplex is a tetrahedron in three-dimensional space. Each vertex is characterized by a set of values for the N desired parameters. The simplex method consists in a strategy (Press et al., 2001) for changing the locations of the vertices in a manner designed to bring the optimum point into the interior of a very small polyhedron. Azevedo (1992) found the multidimensional modified Powell method for determining successive directions for solution-vector change coupled to the one-dimensional line minimization of Brent at each direction change, together with minimum bracketing (all in Press et al., 2001) much faster and more reliable than the downhill simplex method. He also found that smaller minimums of the objective function prevailed when all three Kostiakov-Lewis parameters, k, a, and b, were sought in the search— field determinations of b, for example, were often in considerable error. Of note, the solutions proved not unique, with different ultimate parameter values resulting from different starting values. At the same time, advance calculated with the different sets did not differ much. Bautista and Wallender (1993) used the Marquardt method (Press et al., 2001), a standard technique for non-linear least-squares curve fitting, to minimize the objective function,

490

Chapter 13 Hydraulics of Surface Systems

⎛ T A j − t A (p ) j ⎜ ⎜ σj j =1 ⎝ m

F (p ) = ∑

⎞ ⎟ ⎟ ⎠

2

(13.80)

in which m is the number of comparison stations, TAj is measured advance to a given station xj, while tAj is simulated, with field-parameter vector p having components, k, a, and b; σ is a weighting factor, generally set to unity, to account for uncertainty in a measurement. Roughness was assumed known, and ultimately, b also (field-measured steady state infiltration rate), as the Marquardt algorithm exhibited difficulty in converging to optimum values of three parameters. An alternate objective function, couched in terms of advance velocity, proved more useful in identifying spatially varying infiltration, but ultimately, those results were inconclusive. All of these methods provide more reliable infiltration data if advance data from the entire irrigation is used, rather than any portion of the advance curve, e.g., in an attempt to discern the infiltration parameters as the irrigation progresses. 13.7.4.2 Optimization to match outflow hydrographs. Infiltration parameters suitable for the entire duration of an irrigation should yield simulated runoff in agreement with measured values. Accordingly, Ley (1978), incorporated a match of total runoff volumes into his extension of Christiansen’s et al. (1966) match of advance curves. Walker (2005) developed a systematic procedure for selecting appropriate values of K, a, B, and Manning n in a simulation that would match measured outflow hydrographs, when measured inflow, furrow cross-sections, and bottom slope were given. The search procedure differs from classical optimization of several parameters simultaneously in that the parameters are sought sequentially, albeit iteratively. The concept of sequential determination of the parameters is based on the general observation in a wide range of simulations that the four sought after parameters affect the results in different ways. Specifically, the K value appears the most influential in determining advance time (the time outflow begins), a is the most important factor affecting the shape of the outflow hydrograph, and B is the most important in determining the absolute values of runoff rate. The roughness n was seen as most influential in determining the time recession ends (the time outflow ends). The necessity of iteration stems from the imprecision of these observations and the interactive effects of parameter values. In the multiple simulations comprising the method, the measured furrow geometry and inflow hydrograph are augmented by successive approximations to the four parameters. Starting guesses follow from Equation 13.13 for B, the two-point method (Equations 13.72–13.74) for K and a, and values in the literature for n. In the sequence of nested optimizations, K is adjusted in the innermost nest of simulations to make computed and measured advance times equal. The secant method provides the successive approximations as changes in tL are nearly proportional to changes in K. In the next outer nest, a Fibonacci search (Kahaner et al., 1989) for B minimizes rmsRO the root mean square of the differences between computed and measured runoff rates, rms RO =

1 m⎛ ∑ ⎜ QRO i M − QRO i S m i =1 ⎝

⎞ ⎟ Kˆ , aˆ , Bˆ , nˆ ⎠

2

(13.81)

Design and Operation of Farm Irrigation Systems

491

where m is the number of points in the hydrograph, the subscript M refers to measured values, S refers to simulated values, and the hat above a variable, e.g., Kˆ , refers to the current value that satisfies the search. At each value of B found in the process, the inner nest of searches adjusts K to match the advance time. The next ring of the search is for a, which also minimizes the root-mean square of the computed and measured values of runoff. At each a value, Bˆ and Kˆ are re-computed. The final, outer ring determines nˆ to match computed and measured recession times. A key advantage of the method is that all data collection during the irrigation is confined to the inlet and outlet of the furrow, with no need to enter the wet field. 13.7.4.3 Matching measured depths. An estimation technique based on a few selected depth measurements at various locations or times during advance (not enough to define stream volume, but only to compare with computed depths) was presented by Katopodes et al. (1990). Several similar gradient procedures were applied by Playán and Garcia-Navarro (1997) in a search for the n-k parameter pair and also the a-k pair, with the third component of the parameter vector held fixed. A quite different procedure for estimating whole-border infiltration and roughness for the advance phase was developed by Valiantzas (1994) utilizing a combination of measurements and theoretical (zero-inertia) simulations (Section 13.6.7 et seq.). The field parameters sought are the Kostiakov k and a and Manning n (actually its equivalent, normal depth, for the given inflow and bottom slope). Both the advance trajectory xA(ti), i = 1,..., N, and a depth hydrograph yR(tj), j = 1,..., M, in a reference point xR at or near the inlet end of the border, are measured.

REFERENCES Abramowitz, M., and I. A. Stegun, eds. 1964. Handbook of Mathematical Functions. Appl. Math. Series 55, June. Gaithersburg, Md.: Nat. Bur. Standards. Akanbi, A. A., and N. D. Katopodes. 1988. Model for flood propagation on initially dry land. J. Hydr. Eng. 114(7): 689-706. Alazba, A. A. 1999. Explicit volume balance model solution. J. Irrig. Drain. Eng. 125(5): 273-279. Alazba, A. A., and D. D. Fangmeier. 1995. Hydrograph shape and border irrigation efficiency. J. Irrig. Drain. Eng. 121(6): 452-457. Alazba, A. A., and T. S. Strelkoff. 1994. Correct form of Hall technique for border irrigation advance. J. Irrig. Drain. Eng. 120(2): 292-307. Alvarez, J. A. R. 2003 Estimation of advance and infiltration equations in furrow irrigation for untested discharges. Agric. Water Mgmt. 60(3): 227-239. Azevedo, C. A. V. 1992. Real-time simulation of the inverse furrow advance problem. Ph.D. diss. Logan, Utah: Utah State Univ. Bautista, E., and W. W. Wallender. 1993. Identification of furrow intake parameters from advance times and rates. J. Irrig. Drain. Eng. 119(2): 295-311. Bellos, C. V., J. V. Soulis, and J. G. Sakkas. 1991. Computation of two-dimensional dam-break induced flows. Adv. in Water Resour. 14(1): 31-41. Bjorneberg, D. L., T. J. Trout, R. E. Sojka, and K. K. Aase. 1999. Evaluating WEPP predicted infiltration, runoff, and soil erosion for furrow irrigation. Trans. ASAE 42(6): 1733-1741. Blair, A. W., and E. T. Smerdon. 1985. Effect of wetted perimeter on infiltration in furrows. In Proc. Spec. Conf. Irrig. and Drain. Div., 162-169. Reston, Va.: Ameri-

492

Chapter 13 Hydraulics of Surface Systems

can Soc. Civil Engineers. Blair, A. W., E. T. Smerdon, and J. Rutledge. 1984. An infiltration model for surge flow irrigation. In Proc. Spec. Conf. Irrig. and Drain. Div., 691-700. Reston, Va.: American Soc. Civil Engineers. Boas, M. L. 1966. Mathematical Methods in the Physical Sciences. New York, N.Y.: John Wiley and Sons. Bondurant, J. 1971. Proc. of the Annual Report to Regional Research Committee W-65 on the Hydraulics of Surface Irrigation. Washington, D.C.: USDA-ARS. Bouwer, H. 1957. Infiltration patterns for surface irrigation. Agric. Eng, 38(9): 662664, 676. Bradford, S., and N. D. Katopodes. 1998 Nonhydrostatic model for surface irrigation. J. Irrig. Drain. Eng. 124(4): 200-212. Burt, C. M., A. J. Clemmens, T. S. Strelkoff, K. H. Solomon, R. D. Bliesner, L. A. Hardy, T. A. Howell, and D. E. Eisenhauer. 1997. Irrigation performance measures: Efficiency and uniformity. J. Irrig. Drain. Eng. 123(6): 423-442. Carslaw, H. S., and J. C. Jaeger. 1959. Conduction of Heat in Solids. New York, N.Y.: John Wiley & Sons. Christiansen, J. E., A. A. Bishop, F. W. Kiefer, Jr., and Y. S. Fok. 1966. Evaluation of intake rate constants as related to advance of water in surface irrigation. Trans. ASAE 9(5): 671-674. Clemmens, A. J. 1981. Evaluation of infiltration measurements for border irrigation. Agric. Water Mgmt. 3: 251-267. Clemmens, A. J. 1982. Evaluating infiltration for border irrigation models. Agric. Water Mgmt. 5: 159-170. Clemmens, A. J. 1991. Direct solution to surface irrigation advance inverse problem. J. Irrig. Drain. Eng. 117(4): 578-594. Clemmens, A. J. 1998. Level basin design based on cutoff criteria. Irrig. Drain. Systems 12: 85-113. Clemmens, A. J. and Dedrick, A.R. 1982. Limits for practical level basin design. J. Irrig. Drain. Div., ASCE 108(IR2): 127-141 Clemmens, A. J., and A. R. Dedrick. 1984. Multi-purpose head detection and monitoring unit. Trans. ASAE 27(6): 1825-1828. Clemmens, A. J., A. R. Dedrick, and R. J. Strand. 1995a. BASIN: A Computer Program for the Design of Level-Basin Irrigation Systems. Version 2.0. WCL Report #19. Phoenix, Ariz.: USDA-ARS, U.S. Water Conservation Laboratory. Clemmens, A. J., D. E. Eisenhauer, and B.L. Maheshwari. 2001a. Infiltration and roughness equations for surface irrigation: How form influences estimation. ASAE Paper No. 01-2255. St. Joseph, Mich.: ASAE. Clemmens, A. J., and K. H. Solomon. 1995. Procedures for combining distribution uniformity components. In Proc., First Int’l. Conf. Sponsored by the Water Resources Eng. Div., 1531-1535. Reston, Va.: American Soc. Civil Engineers. Clemmens, A. J., T. S. Strelkoff, and C. M. Burt. 1995b. Defining efficiency and uniformity: Problems and perspectives. In Proc. First Int’l. Conf., Water Resources Eng. Div., 1521-1525. Reston, Va.: American Soc. Civil Engineers. Clemmens, A. J., T. S. Strelkoff, and E. Playan. 2003. Field verification of twodimensional surface irrigation model. J. Irrig. Drain. Eng. 129(6): 402-411. Clemmens, A. J., T. L. Wahl, M. G. Bos, and J. A. Replogle. 2001b. Water

Design and Operation of Farm Irrigation Systems

493

Measurement in Flumes and Weirs. Publication #58. Wageningen, The Netherlands: Int’l. Inst. for Land Reclamation and Improvement (ILRI). Collis-George, N. 1974. A laboratory study of infiltration advance. Soil Sci. 117(5): 282-287. Cunge, J. A., F. M. Holly, and A. Verwey. 1980. Practical Aspects of Computational River Hydraulics. Boston, Mass.: Pitman Advanced Publishing Program. Dedrick, A. R. 1991. Surface-drained level basins for rice production. In Proc. Spec. Conf. Irrig. and Drain. Div. Reston, Va.: American Soc. Civil Engineers. Dudley, S. J., S. R. Abt, C. D. Bonham, and C. Fischenich. 1996. Efficiency of estimating vegetative density using point frame. In Proc. North Am. Water and Environ. Congr. Reston, Va.: American Soc. Civil Engineers. Dunn, C., F. Lopez, and M. Garcia. 1996. Vegetation-induced drag: An experimental study. In Proc. North Am. Water and Environ. Congr. Reston, Va.: American Soc. Civil Engineers. Elliot, R. L., and W. R. Walker. 1982. Field evaluation of furrow infiltration and advance functions. Trans. ASAE 25: 396-400. Enciso-Medina, J., D. Martin, and D. Eisenhauer. 1998. Infiltration model for furrow irrigation. J. Irrig. Drain. Eng. 124(2): 73-80. Esfandiari, M., and B. L. Maheshwari. 1997. Application of the optimization method for estimating infiltration characteristics in furrow irrigation and its comparison with other methods. Agric. Water Mgmt. 34: 169-185. Fangmeier, D. D., A. J. Clemmens, M. El-Ansary, T. S. Strelkoff, and H. E. Osman. 1999. Influence of land leveling precision on level-basin advance and performance. Trans. ASAE 42(4): 1019-1025. Finkel, H. J., and D. Nir. 1960. Determining infiltration rates in irrigation borders. J. Geophys. Res. 65: 2125-2131. Garabedian, P. 1964. Partial Differential Equations. New York, N.Y.: John Wiley & Sons. Green, W. H., and G. Ampt. 1911. Studies of soil physics. Part I: The flow of air and water through soils, J. Agric. Sci. 4: 1-24. Hall, W. A. 1956. Estimating irrigation border flow. Agric. Eng. 37: 263-265. Hart, W. E., D. L. Bassett, and T. S. Strelkoff. 1968. Surface irrigation hydraulics: Kinematics. J. Irrig. Drain. Div., ASCE 94(IR4): 419-440. Hartley, D. M. 1992. Interpretation of Kostiakov infiltration parameters for borders. J. Irrig. Drain. Eng. 118(1): 156-165. Heermann, D. F., and N. A. Evans. 1969. Estimate of resistance from roughness measurements. Unpublished contribution from the Northern Plains Branch, Soil and Water Conservation Res. Div., ARS, USDA and the Colo. Agric. Expt. Stn., Fort Collins, Colo. to W. Reg. Res. Proj. W-65, USDA-CSRS. Hromadka II, T. V., and J. J. DeVries. 1988. Kinematic wave routing and computational error. J. Hydr. Eng. 114(2): 207-217. Hutson, S. S., N. L.Barber, J. F. Kenny, K. S. Linsey, D. S. Lumia, and M. A Maupin. 2004. Estimated Use of Water in the United States in 2000. USGS Circular 1268. Available at: water.usgs.gov/pubs/circ/2004/circ1268/. ISED. 1989. SIRMOD, Surface Irrigation Simulation Software. User’s Guide. Logan, Utah: Irrig. Software Eng. Div., Dept. Agric. and Irrig. Eng., Utah State Univ. Israelsen, O. W. 1932 Irrigation Principles and Practices. New York, N.Y.: John

494

Chapter 13 Hydraulics of Surface Systems

Wiley & Sons. Izuno, F. T., and T. H. Podmore. 1985. Kinematic wave model for surge irrigation research in furrows. Trans. ASAE 28(4): 1145-1150. Kahaner, D., C. Moler, and S. Nash. 1989. Numerical Methods and Software. Englewood Cliffs, N.J.: Prentice Hall. Katopodes, N. D. 1990. Observability of surface irrigation advance. J. Irrig. Drain. Eng. 116(5): 656-675. Katopodes, N. D., and T. S. Strelkoff. 1977a. Hydrodynamics of border irrigation: Complete model. J. Irrig. Drain. Div., ASCE 103(IR3): 309-324. Katopodes, N. D., and T. S. Strelkoff. 1977b. Dimensionless solutions of border irrigation advance. J. Irrig. Drain. Div., ASCE 103(IR4): 401-417. Katopodes, N. D., J. H. Tang, and A. J. Clemmens. 1990. Estimation of surface irrigation parameters. J. Irrig. Drain. Eng. 116(5): 676-695. Kemper W. D., W. H. Heinemann, D. C. Kincaid, and R. V. Worstell. 1981. Cablegation: I. Cable controlled plugs in perforated supply pipes for automatic furrow irrigation. Trans. ASAE 24(6): 1526-1532. Kincaid, D. S., and J. L. Stevens. 1992. CABLE: A Design Program for the Cablegation Automatic Surface Irrigation System. Kimberly, Idaho: USDA-ARS. Kostiakov, A. N. 1932. On the dynamics of the coefficient of water percolation in soils and on the necessity for studying it from a dynamic point of view for purposes of amelioration. In Trans., 6th Comm., Int’l. Soc. of Soil Sci. (In Russian) Part A: 17-21. Moscow: ISSS, Russian Federation. Kouwen, N., R-M. Li, and D. S. Simons. 1981. Flow resistance in vegetated waterways, Trans. ASAE 24(3): 684-690, 698. Kruse, E. G., C. W. Huntley, and A. R. Robinson. 1965. Flow resistance in simulated irrigation borders and furrows. Conserv. Res. Report 3: 1-56. Washington, D.C.: USDA-ARS. Lenau, C. W. 1969. Discussion of Hart et al. (1968), Surface irrigation hydraulics: Kinematics. J. Irrig. Drain. Div., ASCE 95(IR4): 624-627. Lewis, M. R. 1937. The rate of infiltration of water in irrigation practice. Trans. AGU 18: 361-368. Lewis, M. R., and W. E. Milne. 1938. Analysis of border irrigation. Agric. Eng. 19: 267-272. Ley, T. W. 1978. Sensitivity of furrow irrigation performance to field and operation variables. M.Sc. thesis. Fort Collins, Colo.: Dept. of Ag. and Chem Eng., Colorado State Univ. Lighthill, M. J., and G. B. Whitham. 1955. On kinematic waves. I: Flood movements in long rivers. Proc. Royal Soc. of London, Series A 229: 281-316. Maheshwari, B. L. 1992. Suitability of different flow equations and hydraulic resistance parameters for flows in surface irrigation: A review. Water Resour. Res. 28(8): 2059-2066. Maheshwari, B. L., A. K. Turner, T. A. McMahon, and B. J. Campbell. 1988. An optimization technique for estimating infiltration characteristics in border irrigation. Agric. Water Mgmt. 13(1): 13-24. Merriam, J. L. 1971. Adjusting cylinder infiltrometer data for field use. ASAE paper No. 71-13. St. Joseph, Mich.: ASAE. Merriam, J. L., and A. J. Clemmens. 1985. Time rated infiltrated depth families. In

Design and Operation of Farm Irrigation Systems

495

Development and Management Aspects of Irrigation and Drainage, Spec. Conf. Proc., Irrig. and Drain. Div., 67-74. Reston, Va.: American Soc. Civil Engineers. Monserrat, J. 1994 Solucion al Problema inverso del riego por tablares mediante un modelo hidrologico mixto, Tesis Doctoral. Universidad de Leida, Escuela Tecnica Superior de Ingenieria Agraria. Monserrat, J., and J. Barragan. 1998. Estimation of the surface volume in hydrological models for border irrigation. J. Irrig. Drain. Eng. 124(6): 238-247. NRCS. 1984. Section 15: Irrigation, Chapter 5: Furrow Irrigation. In National Engineering Handbook. Washington, D.C.: USDA NRCS. Palomo, M. J., N. A. Oyonarte, L. Mateos, and J. Roldan. 1996. Infiltracion en riego por surcos mediante pulsaciones intermitentes. In Proc. XIV Congreso Nacional de Riegos, 235-243. Aguadulce (Almeria). (in Spanish). Perea, H. 2005. Development, verification, and evaluation of a solute transport model in surface irrigation. Ph.D. diss. Tucson, Ariz.: Univ. Arizona. Perea, H., T. S. Strelkoff, J. Šimunek, E. Bautista, and A. J. Clemmens. 2003. Unsteady furrow infiltration in the light of the Richards equation. In Proc. 2nd Int’l Conf. on Irrig. and Drain., 625-636. U.S. Com. on Irrig. and Drain. Petryk, S., and G. Bosmajian. 1975. Analysis of flow through vegetation. J. Hydr. Eng. 101(7): 871-884. Philip. J. R. 1969. Theory of infiltration. In Advances in Hydroscience. Ven Te Chow, ed. 5: 215-295. New York, N.Y.: Academic Press. Philip, J. R., and D. A. Farrell. 1964. General solution of the infiltration advance problem in irrigation hydraulics. J. Geophys. Res. 69: 621-631. Playan, E., J. M. Faci, and A. Serreta. 1996. Modeling microtopography in basin irrigation. J. Irrig. Drain. Eng. 122(6): 339-347. Playán, E., and P. García-Navarro. 1997. Radial flow modeling for estimating levelbasin irrigation parameters. J. Irrig. Drain. Eng. 123(4): 229-237. Playan, E., W. R. Walker, and G. P. Merkley. 1994. Two-dimensional simulation of basin irrigation. I: Theory. J. Irrig. Drain. Eng. 120: 837-856. Press, W. H., S. A. Teukolsky, W. T. Vetterling, and B. P. Flannery. 2001. Numerical Recipes in FORTRAN 77: The Art of Scientific Computing. 2nd ed. Cambridge, England: Cambridge Univ. Press. Purkey, D. R., and W. W. Wallender. 1989. Surge infiltration variability. Trans. ASAE 32(3): 894-900. Ree, W. O., and V. J. Palmer. 1949. Flow of Water in Channels Protected by Vegetated Linings, 1-115. Bulletin No. 967. Washington, D. C.: U.S. Soil Conserv. Roth, R. L., D. W. Fonken, D. D. Fangmeier, and K. T. Atchison. 1974. Data for border irrigation models. Trans. ASAE 17(1): 157-161. Saint Venant, B. D. 1871. Théorie du mouvement non-permanente des eaux avec application aux crues des riviéres et à l’introduction des marées dans leur lit. Compte Rendus, Acad. Sci. 73: 148-154, 237-240. Sakkas, J. G., and T. S. Strelkoff. 1974. Hydrodynamics of surface irrigation: Advance phase. J. Irrig. Drain. Div., ASCE 100(IR1): 31-48. Sakkas, J. G., C. V. Bellos, and M. N. Klonaraki. 1994. Numerical computation of surface irrigation. Irrig. Sci. 15: 83-99. Salt River Project. 1960. Major Facts in Brief. Phoenix, Ariz.: Salt River Project. Sayre, W. W., and M. L. Albertson. 1961. Roughness spacing in rigid open channels.

496

Chapter 13 Hydraulics of Surface Systems

J. Hydr. Div. 87(3): 121-149. Schmitz, G. H., and G. J Seus. 1987. Analytical solution of simplified surge flow. J. Irrig. Drain. Eng. 113(4): 603-610. Schmitz, G. H. 1993a. Transient infiltration from cavities. I: Theory. J. Irrig. Drain. Eng. 119(3): 443-457. Schmitz, G. H. 1993b. Transient infiltration from cavities. II: Analysis and application. J. Irrig. Drain. Eng. 119(3): 458-470. Schmitz, G. H., and G. J. Seus. 1990. Mathematical zero-inertia modeling of surface irrigation: Advance in borders. J. Irrig. Drain. Eng. 116(5): 603-615. Schmitz, G. H., and G. J. Seus. 1992. Mathematical zero-inertia modeling of surface irrigation: Advance in furrows. J. Irrig. Drain. Eng. 118(1): 1-18. Shatanawi, M.R. and Strelkoff, T.S. 1984. Management contours for border irrigation, J. Irrig. Drain. Eng. 110(4): 393-399. Shayya, W. H., V. F. Bralts, and L. J. Segerlind. 1991. Zero-inertia surface irrigation design: A finite element approach. ASAE Paper No. 91-2099. St. Joseph, Mich.: ASAE. Shepard, J. S., W. W. Wallender, and J. W. Hopmans. 1993. One-point method for estimating furrow infiltration. Trans. ASAE 36(2): 395-404. Sherman, B., and V. J. Singh. 1982. A kinematic model for surface irrigation: An extension. Water Resour. Res. 18(3): 659-667. Shockley, H. J., G. Woodard, and J. T. Phelan. 1964. A quasi-rational method of border irrigation design. Trans. ASAE 7(4): 420-423, 426. Šimunek, J., M. Šejna, and M. T. van Genuchten. 1999. The HYDRUS-2D software package for simulating two-dimensional movement of water, heat, and multiple solutes in variably saturated media. Version 2.0, IGWMC-TPS-53. Golden, Colo.: Int’l. Grd.Water Modeling Ctr., Colorado School of Mines. Singh, V. J., and R. S. Ram. 1983. A kinematic model for surface irrigation: Verification by experimental data. Water Resour. Res. 19(6): 1599-1612. Singh, V. P., and F. X. Yu. 1989a. An analytical closed border irrigation model I: Theory. Agric. Water Mgmt. 15: 223-241. Singh, V. P., and F. X. Yu. 1989b. An analytical closed border irrigation model II: Experimental verification. Agric. Water Mgmt. 15: 243-252. Skonard, C. 2002. A field-scale furrow irrigation model. Ph.D. diss. Lincoln, Nebr.: Univ. of Nebraska. Smith, R. E. 1972. Border irrigation advance and ephemeral flood waves. J. Irrig. Drain. Div., ASCE 98(IR2): 289-307. Strelkoff, T. 1969. One-dimensional equations of open channel flow. J. Hydr. Eng. 95(HY3): 861-876. Strelkoff, T. 1977. Algebraic computation of flow in border irrigation. J. Irrig. Drain. Div., ASCE 103(IR3): 357-377. Strelkoff, T. 1985. BRDRFLW: A Mathematical Model of Border Irrigation, 1-100. Publ. ARS-29. Washington, D.C.: USDA-ARS. Strelkoff, T. 1990. SRFR: A computer program for simulating flow in surface irrigation: Furrows–Basins–Borders. WCL Report #17. Phoenix, Ariz.: U.S. Water Conserv. Lab. Strelkoff, T. 1992. EQSWP: Extended unsteady flow double-sweep equation solver. J. Hydr. Eng. 118: 735-742.

Design and Operation of Farm Irrigation Systems

497

Strelkoff, T. S., E. Bautista, and A. J. Clemmens. 2003b. Errors inherent in simplified infiltration parameter estimation. In Proc. 2nd Int’l. Conf. on Irrig. and Drain., 735-745. Phoenix, Ariz: U.S. Com. on Irrig. and Drain. Strelkoff, T. S., and D. L. Bjorneberg. 2001. Hydraulic modeling of irrigation-induced furrow erosion. Selected papers from the 10th Int’l Soil Conserv. Organiz. Conf., ISCO ‘99. CD-ROM: 699-705. Strelkoff, T. S., and A. J. Clemmens. 1994. Dimensional analysis in surface irrigation. Irrig. Sci. 15: 57-82. Strelkoff, T. S., and A. J. Clemmens. 1996a. Managing border irrigation for near-zero discharge. In Proc., North Am. Water and Environ. Congr. ‘96, CD-ROM. Reston, Va.: American Soc. Civil Engineers. Strelkoff, T. S., and A. J. Clemmens. 1996b. Overviews of border irrigation performance for optimization. ASAE Paper No. 962140. St. Joseph, Mich.: ASAE. Strelkoff, T. S., and A. J. Clemmens. 2001. Data-driven organization of field methods for estimation of soil and crop hydraulic properties. ASAE Paper No. 01-2256. St. Joseph, Mich.: ASAE. Strelkoff, T. S., A. J. Clemmens, M. El-Ansary, and M. Awad, M. 1999. Surfaceirrigation evaluation models: Application to level basins in Egypt. Trans. ASAE 42(4): 1027-1036. Strelkoff, T.S., Clemmens, A. J., and H. Perea-Estrada. 2006. Calculation of nonreactive chemical distribution in surface fertigation. Agric. Water Mgmt., 86:93-101. Strelkoff, T. S., A. J. Clemmens, and B. V. Schmidt. 1998. SRFR, Version 3.31: A Model for Simulating Surface Irrigation in Borders, Basins and Furrows. Phoenix, Ariz.: USWCL, USDA-ARS. Strelkoff, T. S., A. J. Clemmens, B. V. Schmidt, and E. J. Slosky. 1996. Border:A design and management aid for sloping border irrigation systems. In Model and Computer Program Software Documentation. WCL Report #21. Strelkoff, T. S., and H. T. Falvey. 1993. Numerical methods used to model unsteady canal flow. J. Irrig Drain. Eng. 119(4): 637-655. Strelkoff, T. S., and N. D. Katopodes. 1977a. End depth under zero-inertia conditions. J. Hydr. Eng. 103(HY7): 699-711. Strelkoff, T. S., and N. D. Katopodes. 1977b. Border-irrigation hydraulics with zero inertia. J. Irrig. Drain. Div., ASCE 103(IR3): 325-343. Strelkoff, T.S. and Shatanawi, M.R. 1984. Normalized graphs of border-irrigation performance, J. Irrig. Drain. Eng. 110(4): 359-374 Strelkoff, T. S., A. H. Tamimi, and A. J. Clemmens. 2003a. Two-dimensional basin flow with irregular bottom configuration. J. Irrig. Drain. Eng. 129(6): 391-401. Tabuada, M. A., Z. J. C. Rego, G. Vachaud, and L. S. Pereira. 1995. Modelling of furrow irrigation. Advance with two-dimensional infiltration. Agric. Water Mgmt. 28: 201-221. Tinney, E. R., and D. L. Bassett. 1961. Terminal shape of a shallow liquid front. J. Hydr. Div. 87(HY5): 117-133. Trout, T. J. 1992. Flow velocity and wetted perimeter effects on furrow infiltration. Trans. ASAE 35(3): 855-862. Trout, T. J., and G. S. Johnson. 1989. Earthworms and furrow irrigation infiltration. Trans. ASAE 32(5): 1594-1598. Trout, T. J., R. E. Sojka, and R. D. Lentz. 1995. Polyacrylamide effect on furrow

498

Chapter 13 Hydraulics of Surface Systems

erosion and infiltration. Trans. ASAE 38: 761-765. USDA. 1974. Chapt. 4, Section 15: Border Irrigation. In National Engineering Handbook. Washington, D.C.: USDA, Soil Conservation Service. USDA. 1985. Chapt. 5, Section 15: Furrow Irrigation. In National Engineering Handbook. Washington, D.C.: USDA, Soil Conservation Service. USDA. 1999. SRFR, Version 4.06. Maricopa, Ariz.: U.S. ALARC, USDA-ARS. USDA-ARS. 2006. WinSRFR: Surface Irrigation Analysis, Design, and Simulation. V. 1.1. U.S. Arid Land Agricultural Research Center. Maricopa, AZ 85239. Available at: http://www.ars.usda.gov/services/software/download.htm?softwareid=146. Valiantzas, J. D. 1994. Simple method for identification of border infiltration and roughness characteristics, J. Irrig. Drain. Eng. 120(2): 233-249. Valiantzas, J. D. 1997. Surface irrigation advance equation: Variation of subsurface shape factor, J. Irrig. Drain. Eng. 123(4): 300-306. Valiantzas J. D., S. Aggelides, and A. Sassalou. 2001, Furrow infiltration estimation from time to a single advance point. Agric. Water Mgmt. 52: 17-32. Walker, W. R. 2003. SIRMOD III: Surface Irrigation Simulation, Evaluation, and Design. Guide and technical documentation. Logan, Utah: Dept. of Bio and Irrig. Eng., Utah State Univ. Walker, W.R. 2005. Multilevel calibration of furrow infiltration and roughness, J. Irrig. Drain. Eng., 131(2): 129-136 Walker, W.R. and Busman, J.D. 1990. Real-time estimation of furrow infiltration, J. Irrig. Drain. Eng., 116(3): 299-318 Walker, W. R., and A. S. Humpherys. 1983. Kinematic-wave furrow irrigation model. J. Irrig. Drain. Eng. 109(4): 377-392. Walker, W. R., and B. Kasilingam. 2004. Another look at wetted perimeter along irrigated furrows: Modeling implications. In Proc. World Water and Environ. Resour. Congr. CD-ROM. Reston, Va.: EWRI/ASCE. Walker, W. R., and G. V. Skogerboe. 1987. Surface Irrigation Theory and Practice. Englewood Cliffs, N.J.: Prentice Hall. Wallender, W. W., and M. Rayej. 1985. Zero-inertia surge model with wet-dry advance. Trans. ASAE 28(5): 1530-1534. Weir, G. J. 1983. A mathematical model for border strip irrigation. Water Resour. Res. 19(4): 1011-1018. Weir, G. J. 1985. Surface irrigation advance motion after gate closure. Water Resour. Res. 21(9): 1409-1414. Whitham, G. B. 1955. The effects of hydraulic resistance in the dam-break problem. Proc., Royal Society of London, Series A. 227: 399-407. London, England. Wu, I. 1972. Recession flow in surface irrigation. J. Irrig. Drain. Div., ASCE 98(IR1): 77-90. Yost, S. A., and N. D. Katopodes. 1998. Global identification of surface irrigation parameters. J. Irrig. Drain. Eng. 124(3): 131-139. Zerihun, D., C. A. Sanchez, A. Furman, and A. W. Warrick. 2005. A coupled surfacesubsurface flow model for improved basin irrigation management. J. Irrig. Drain. Eng. 131(2): 111-128. Zerihun, D., C. A. Sanchez, and K. L. Farrell-Poe. 2004. Modified two-point method for closed-end level-bed furrows. In Proc. 2004 World Water and Environ. Resour. Congr. CD-ROM. Reston, Va.: EWRI/ASCE.

CHAPTER 14 DESIGN OF SURFACE SYSTEMS Albert J. Clemmens (USDA-ARS, Maricopa, Arizona) Wynn R. Walker (Utah State University, Logan, Utah) Delmar D. Fangmeier (University of Arizona, Tucson, Arizona) Leland A. Hardy (H & R Engineering, Inc., Salem, Oregon) Abstract. Surface irrigation design and operation are a challenge because soil infiltration and soil and crop resistance influence the movement of water over the field and thus the water distribution. In the past, design for each surface irrigation method was treated differently because of differences in the simplicity with which different phases of the irrigation could be described. This has tended to make surface irrigation analysis and design appear disjointed. In this chapter, we apply the same basic procedures for the design of all surface systems, deviating where needed to make the procedures both straightforward and sufficiently accurate. The basis for these designs is the ability to predict advance, recession, the distribution of infiltrated water, and the performance for a given set of conditions. Conservation of mass is the main concept, with empirical approximations used where needed. Examples are provided for each method. Keywords. Basin, Border, Design, Furrow, Irrigation, Surface irrigation.

14.1 INTRODUCTION Surface irrigation methods are so named because water is distributed across the field by flowing over the field surface. Thus, soil infiltration and soil and crop flow resistance have major influences on the distribution of water. Because these conditions change with both time and space, the design and operation of surface irrigation systems is often more difficult than pressurized irrigation systems. However, surface irrigation systems still comprise more than 90% of the world’s irrigated land (FAO, 2005) and almost half of the irrigated land in the U.S. (Hutson, 2004). While some land is continually being converted to pressurized irrigation, a significant portion of land is likely to remain surface irrigated for the foreseeable future. Surface irrigation can be an effective irrigation method and in some cases can equal the efficiencies of pressurized irrigation methods (Kennedy, 1994: 166). Surface irrigation is still the most economical method for many situations.

500

Chapter 14 Design of Surface Systems

Method Sloping furrow

Border strip

Level basins/ Level furrows

Table 14.1. Categories of surface irrigation. Inflow Control of Slope Control Lateral Flow furrows steep or to individual low gradient; furrows either can have cross slope flat planted or steep or distributed corrugations low gradient across cross slope on upper end milder slopes flat planted, zero in all can be point furrowed or directions inflow bedded furrows zero in direction to individual of run, can have furrows cross slope

End Conditions open, blocked, or group of furrows blocked open, blocked, or partially blocked blocked if furrowed, all interconnected blocked, or group of furrows blocked

Surface irrigation methods can be categorized according to how they function hydraulically. Distinctions can be made by advance and recession curves, which describe the time when the advancing stream reaches a particular location and the time when standing water no longer occurs on the surface. This hydraulic comparison assumes that water enters the field or irrigation set along one end and flows to the other end uniformly across the set width. We distinguish three categories of surface irrigation: sloping furrows, border strips, and level basins and furrows. The main physical differences are given in Table 14.1. These physical differences result in hydraulic differences, primarily the magnitude of the inflow rate (and pattern) and the general shape of the recession curves and runoff hydrographs. The purpose of this chapter is to present methods for the design of modern surface irrigation systems, where modern implies a reasonable control over the water supply. The chapter focuses on the three major categories of surface irrigation; design of rice paddies and methods such as contour levee, contour ditch, wild flooding, etc. are not discussed.

14.2 DESIGN CONSIDERATIONS AND APPROACHES Many surface irrigation systems are ineffective and inefficient. This can be caused by physical constraints (e.g., steep land slopes, shallow soils, poor water supplies, etc.), by poor design and layout, or by improper operation and management. A thorough discussion of the constraints and limitations of surface irrigation systems is beyond the scope of this chapter. For more details, see Walker and Skogerboe (1987), Clemmens and Dedrick (1994), or Burt et al. (1999). One advantage of surface irrigation over pressurized irrigation methods is that they often do not require a good, reliable water supply. They can be adapted to different rates of flow, flows that vary randomly, and flows with poor water quality (sediment, debris, etc.). Efforts in surface irrigation research and extension have focused on methods for providing better water control—control over flow rate or control over volume applied. These generally focus on how the system is operated. Of equal importance is field design and layout. Good operation cannot make up for a poor field design. However, when surface irrigation systems are properly designed and more modern operating procedures are used, irrigation efficiencies and uniformities can be high.

Design and Operation of Farm Irrigation Systems

501

14.2.1 Design Objectives The amount of water to be supplied during an irrigation event, referred to as the target or required depth of application, is a major design consideration. Typically, surface irrigation systems have a narrow range of target depths for which they are reasonably efficient and uniform. The irrigator must adjust irrigation practices (typically flow rate and application time) to account for changing field conditions (infiltration and roughness). Irrigators typically develop rules of operation that they use to make adjustments. A poor field design will make these judgments difficult for the operator. A good design will include guidance on system operation. Surface irrigation systems are most applicable on mild to level slopes. On steeper slopes, erosion can become excessive and the range of operating conditions can be narrow (e.g., narrow range of target depths). The design objectives are typically stated in terms of achieving some desired application efficiency, Ea. This efficiency is called potential application efficiency, PEa, here to distinguish it from a field-measured Ea. Design is typically based on supplying the target depth of water everywhere in the field. The PEmin is the application efficiency when the minimum depth infiltrated just equals the target depth: PE min =

d min d req = da da

(14.1)

where dreq is the required depth, dmin is the minimum infiltrated depth, and da is the depth of applied irrigation water, in this case, the depth applied that results in the minimum depth just equal to the required depth. In practice, some under-irrigation is usually allowed and operation is based on satisfying the low-quarter depth. Because design does not take into account all of the variability which exists within the field, we use a design based on satisfying the minimum depth, with the expectation that when properly operated the low quarter depth would be satisfied. 14.2.2 Choosing Design Conditions and Parameters Surface irrigation design requires the estimation of parameters that define the infiltration of water into the soil and the resistance to water movement caused by the soil surface and vegetation. These are key factors in the design and are often very difficult to determine accurately. These factors can change over the course of the growing season, further complicating design. Published values of the Manning roughness coefficient, n, are usually sufficient for design purposes. However, infiltration is much more difficult to estimate. Whenever possible, estimates of these parameters should be obtained from field measurements, preferably from irrigation events. Soil infiltration conditions can change significantly when land is newly converted to surface irrigation. Once a reasonable estimate of infiltration parameters is obtained and a design selected, a sensitivity analysis should be performed to determine how the design might change under different conditions of infiltration, flow resistance, and target infiltration depth; conditions that will likely be faced by the irrigator. 14.2.3 Equations for Infiltration in Borders and Basins Infiltration is one of the most important factors affecting the design and performance of surface irrigation systems. Unfortunately, estimating infiltration conditions is one of the most difficult things to do in the field. Several equations have been used to define infiltration. Philip (1957) developed equations for ideal soil conditions (e.g., uniform physical properties) for infiltrated depth versus time for short times:

502

Chapter 14 Design of Surface Systems

d( τ ) = kτ

1/ 2

+ bτ

i( τ ) = 1 2 kτ −1/ 2 + b

(14.2)

and for long times: d( τ ) = c2 + bτ

i( τ ) = b

(14.3)

where τ = infiltration time k = sorptivity b is related to the hydraulic conductivity c2 = a derived constant. Unfortunately, soil surface conditions are not ideal and many soils do not follow Equation 14.2 very well at short times. Many do follow Equation 14.3 at long times, but with different values for c2 than those derived theoretically. Irrigation engineers have tended to use the Kostiakov or Kostiakov-Lewis equation, or a variation thereof, defined by d( τ ) = kτ a

i( τ ) = akτ a −1

(14.4)

or in modified form, d( τ ) = c + kτ a + bτ

i( τ ) = akτ a −1 + b

(14.5)

where b, c and k are all empirically derived. In Equation 14.4, the exponent a is often below 1/2, indicating the poor fit to Equation 14.2. However, with Equation 14.4, the infiltration rate approaches zero as time approaches infinity. Soils generally reach a constant final infiltration rate. For course-textured soils, where the required infiltration time is short, this occurs well after typical irrigation opportunity times. For finetextured soils, the final infiltration rate is reached during the irrigation and should be included. In keeping with the theoretical development, a Kostiakov branch function has been used, where the first branch uses Equation 14.4 (or 14.5 with b = 0) for short times, and switches to Equation 14.3 when the infiltration rate (derivative of equation with respect to time) equals b. For soils that reach a nearly constant, final infiltration rate during the irrigation, design and evaluation can be greatly simplified with use of this equation. The Kostiakov branch function equations are: d( τ ) = c + k τ a i( τ ) = ak τ a-1 = c 2 + bτ =b

for τ ≤ τ B for τ ≥ τ B

(14.6)

where τB is the time at which the infiltration depth and rate for the two branches match. For cracking-clay soils, the depth infiltrated over the first few minutes of wetting can be large relative to the total depth infiltrated during the irrigation. With Equation 14.4, this could be fit with a very small value of the exponent a. However, this makes the problem unnecessarily difficult, since whether the depth infiltrates immediately or, say, during the first 10 min has very little effect on the hydraulics of the irrigation or on the distribution of infiltrated depths. For these soils, it makes more sense to fit infiltration with an equation such as Equation 14.3 and ignore the initial part of the infiltration curve (e.g., use second half of Equation 14.6). The U.S. Soil Conservation Service (SCS, currently Natural Resources Conservation Service) developed a series of infiltration families to assist field personnel in determining infiltration relationships. These families are defined according to Equation 14.5 with b = 0 and c = 7 mm. The infiltration relationships for many soils do not fol-

Design and Operation of Farm Irrigation Systems

503

low the SCS families very well. The value of exponent a in this equation is around 0.7 for all the families. Non-cracking soils typically have exponents that range from 0.3 to 0.8. Merriam and Clemmens (1985) developed a relationship between the exponent a and the time to infiltrate 100 mm from a wide range of field data. From this they developed time-rated intake families, based on the time to infiltrate 100 mm, a typical required depth for surface irrigation. These time-rated intake families are given in Table 14.2. These infiltration relationships are entirely empirical. Emphasis should be placed on the general characteristics of the infiltration relationship and not on the equation chosen or the values of the various parameters. As long as the equation and constants chosen fit the observed infiltration relationship and are useful, this is sufficient justification for their use. In particular, it is most important to have the proper range of infiltrated depths over the range of infiltration opportunity times. Table 14.3 shows the parameters for five infiltration functions, representing the various equations fit to infiltration described by the 6-hr time-rated family (i.e., 6 h to infiltrate 100 mm with a Kostiakov exponent of 0.51). This corresponds to an SCS intake family of about 0.45. These infiltration functions are plotted in Figure 14.1. There is very little difference between the infiltrated depths predicted by the three different Kostiakov equations. A linear equation even fits well between 4 and 8 h.

100

Depth (mm)

[a]

Table 14.2. Time-rated intake families (from Merriam and Clemmens, 1985). Time to Infiltrate Kostiakov Kostiakov Constant (k) Exponent (a)[a] (in/hr[a]) (mm/hr[a]) 100 mm (hours) 0.5 0.739 167.0 6.57 1 0.675 100.0 3.94 2 0.611 65.5 2.58 4 0.547 46.8 1.84 8 0.483 36.6 1.44 16 0.419 31.3 1.23 32 0.355 29.2 1.15 Exponent a = 0.675 – 0.2125 log10(τ100), where τ100 is the time (in hours) to infiltrate 100 mm.

50

SCS Kostiakov Modified Linear Branch

0 0

2

4

6

Time (hours) Figure 14.1. Comparison of five infiltration functions.

8

504

Chapter 14 Design of Surface Systems

Table 14.3. Infiltration functions matching 6-hr time-rated family. SCS Kostiakov Modified Kostiakov Linear Parameters Equation Equation Kostiakov Branch Equation k (mm/hr) 24.26 40.1 37.39 40.1 0 a 0.75 0.51 0.35 0.51 0 b (mm/hr) 0 0 5.0 8.5 8.9 c (mm) 7 0 0 0 49.0

14.2.4 Equations for Infiltration in Furrows Infiltration in furrows can be significantly more complicated than in flat borders or basins due to the two-dimensional nature of the furrow cross-section. Infiltration in borders and basins is generally considered to be one dimensional: downward. Infiltration in furrows can be influenced by the wetted width of the stream and lateral flow into the furrow bed. If gravitational forces dominate the infiltration process (e.g., a very sandy soil), then infiltration may be directly proportional to the wetted width. If furrows are closely spaced and the soil is very heavy such that capillary forces dominate infiltration, the lateral infiltration from adjacent furrows will meet such that infiltration essentially occurs over the entire set width; i.e., for each furrow, infiltration becomes a function of the furrow spacing. The main problem is that for many situations, infiltration conditions are in between these two extremes, and may differ significantly over time or along the length of run. For example, infiltration at the head end of the field, where the flow rate and water depth are high, may result in infiltration after, say, 12 hours essentially governed by furrow spacing; while at the tail end where the flow rate and wetted width are significantly reduced and the infiltration opportunity time is short, infiltration may be strongly influenced by the wetted width. The SCS design procedures for sloping furrows (USDA, 1984) use an infiltration function based on the wetted perimeter at normal depth, plus an empirical constant of 213 mm to account for the lateral flow. This is reported as an average value from numerous tests. For irrigation of every furrow, this width should not exceed the furrow spacing. This width is then assumed constant over time and distance. The assumption with this procedure is that the SCS intake family for a given soil used for border-strip irrigation design can also be used for sloping-furrow design. In practice, this approach has not been very successful. For general use, infiltration really needs to be judged from field performance of the irrigation system. If a constant width is to be chosen for infiltration, it is simpler to express infiltration as a function of furrow spacing; recognizing that significant changes in wetted perimeter (e.g., from furrow shape or flow-rate changes) or furrow spacing (or whether all furrows are irrigated) may change infiltration constants based on furrow spacing. The design procedures presented here do not consider the influence of wetted width on furrow infiltration. Thus, these design procedures should be used with some caution as they may overestimate irrigation uniformity. Where no furrow infiltration data are available, it may be necessary to estimate furrow infiltration from ring infiltration data or from border irrigation trials. For mild slopes, high flow rates, closely spaced furrows and heavy soils, the infiltration can be assumed to be somewhat similar to that for borders (i.e., one-dimensional and based on furrow spacing), except there is a tendency for the infiltration rates to start higher and drop more quickly (i.e., a smaller value of the exponent a in the Kostiakov-Lewis equation). For steeper slopes, low flow rates and coarse-textured soils, the border infil-

Design and Operation of Farm Irrigation Systems

505

tration rates can be applied to the wetted width, with perhaps an increase in k to account for some lateral movement and a decrease in the exponent a. Such adjustments must be considered just a guess since the effects of differences in tillage and compaction from tractor traffic between borders and furrows can be substantial. Furrow irrigation has the additional complexity in that equipment wheels travel through some furrows and not others. Wheel traffic generally causes compaction and smoothing of the soil surface, a reduction in infiltration, and a reduction in flow resistance. In some cases with sloping-furrow systems, only wheel or only non-wheel rows are irrigated. Where both are irrigated, the inflow rates to wheel and non-wheel rows are typically adjusted by the irrigator so that advance in all rows is approximately the same. These adjustments are not needed in level furrow or furrowed level-basin irrigation systems. In other cases, non-wheel furrows are intentionally compacted by dragging heavy objects (sometimes called torpedoes) through them. When considering also changes in properties over the season, design procedures often have to be applied to a wide range of conditions. Because of this, some irrigation evaluations only attempt to determine field-average conditions (e.g., an infiltration function determined for an entire set), which assumes that irrigation and tillage practices will be similar after design. 14.2.5 Equations for Flow Resistance Resistance to flow is usually described by the Manning roughness coefficient. This equation relates the flow rate, Q, to the flow area, A, the hydraulic radius, R (area over wetted perimeter), the friction slope, Sf, and the Manning n: AR 2 / 3 S 1f / 2 Cu

(14.7) n The units coefficient, Cu, allows the same value of the Manning n in both footpound and metric units. For units of m and m3/s, Cu has a value of 1.0. Expressed in terms of flow rate per unit width, q, for border strips and basins, this becomes Q=

y5/3 S1/2 f Cu

(14.8) n where y is the flow depth and q = Q/W, where W is the basin or border width. All units in Equations 14.7 and 14.8 are in meters and seconds. In theory, the Manning n should change with flow depth when vegetation is present. However, the changes in the Manning n over the season and with changes in planting density are as large as changes caused by flow depth. For design purposes, a constant Manning n seems to be satisfactory. Published values of the Manning n for various crops are shown in Table 14.4. Manning n values have been observed as high q=

Source USDA (1974, 1984) USDA (1974, 1984) USDA (1974) Clemmens (1991) USDA (1974)

Table 14.4. Recommended Manning n values. n Conditions for Use 0.04 smooth, bare soil surfaces and furrows 0.10 drilled, small-grain crops if the drill rows run in the direction of water flow and corrugations 0.15 alfalfa, broadcast small grains 0.20 dense alfalfa or alfalfa on long fields with no secondary ditches 0.25 dense sod crops and small grains drilled perpendicular to the flow

506

Chapter 14 Design of Surface Systems

as 0.4 on closely planted wheat, a value significantly higher than that given in the table. Also, furrow-irrigated crops with vegetation that hangs into the water can cause much greater resistance than the Manning n of 0.04 given in Table 14.4. Here again, field observations of advance and recession can be used to determine whether design estimates of Manning n are reasonable. 14.2.6 Consideration of Operating Practices and Guidelines We know that the conditions assumed in design will never be exactly encountered in practice. The irrigator of a surface irrigated field will adapt operations to account for differences from the assumed conditions. In addition to variations in infiltration and roughness, the irrigator must adapt to changes in flow rate and target depth of application. Frequently, the flow rate is only approximately known and changes in application time to account for this are not easily made. The operator frequently uses information about the advancing stream to make these adjustments, either in inflow rate or in application time. Usually, these adjustments are based solely on experience. However, such experience takes time to accumulate and often experience on one type of soil does not translate to another type of soil. To aid in this process, design should consider the operating criteria to be used; either as recommendations after design or used explicitly in the design procedure to help determine the field layout. In any event, a sensitivity analysis can determine what will happen if the irrigator operates the system with the defined operating criteria over the range of possible conditions. 14.2.7 General Design Approaches The approach taken for design assumes that repeated trials are needed based on existing conditions, local farming and irrigation practices, and designer experience. The procedures provided in this chapter allow the calculation of irrigation results and performance for each individual trial. It is up to the designer to explore various options in arriving at a final design recommendation. The methods presented here allow one to compute advance and recession curves, the distribution of infiltrated water, the amount of deep percolation, and runoff, if applicable. The approach assumes that the designer knows the infiltration function of the soil. If infiltration is not well known or changes over the season, the designer is responsible for evaluating the impact of variations in infiltration on design and operations. This is beyond the scope of this chapter. The design procedures assume that the depth to be infiltrated during a particular irrigation event is known by the designer. This required depth of infiltration may be different for different crops and can change over the season. It is the designer’s responsibility to take this into account during the design process. Design is typically based on assuming that one end of the field or the other will receive the least infiltrated depth. Then, the inflow and application time are adjusted such that the required depth is infiltrated at that location. The time to infiltrate the required depth, τreq, becomes an important design parameter. Typically, it has more influence on the design than the infiltration constants themselves. The infiltration opportunity time at any location, x, along the length of run, τopp(x), is defined as the time between advance, tA(x), and recession, tR(x) or

τ opp ( x) = t R ( x) − t A ( x )

(14.9)

If the minimum depth is at the head end of the field (x = 0), the opportunity time at that point is equal to the recession time, or

Design and Operation of Farm Irrigation Systems

τ opp (0) = t R (0) = t co + tlag

507

(14.10)

where tco is the time of cutoff or application time and tlag is the recession lag time, or the time required for the water depth at the upstream end to drop to zero after cutoff. Design based on meeting the requirement at the upstream end does not require advance and recession curves to be computed. However, an estimate of PEmin is required, along with a method for estimating recession lag time (USDA, 1974). For well designed systems, the minimum depth is typically at the downstream end of the field, furthest from the water source (x = L). Then advance and recession curves must be computed, and estimation of inflow rate and time required to achieve τreq is more difficult. Specifically, (14.11) t R ( L) = t A ( L) +τ req Precise solutions of advance and recession are possible with solution of the Saint Venant equations of continuity and momentum. However, this solution requires numerical analysis and is done on a case-by-case basis. It does not directly produce general equations for advance and recession. The time of cutoff is t co = t A ( L) +τ req − [t R ( L) − t R (0) − tlag ]

(14.12)

where the term in brackets is the time between cutoff and recession at the downstream end, which is sometimes assumed to be zero (e.g., sloping furrows with runoff). In this chapter, we provide methods for design based on satisfying the minimum depth at the downstream end. 14.2.8 Assumed Surface-Volume Method All design methods must use procedures that satisfy a volume balance, regardless of how sophisticated they are. In this chapter, we make assumptions regarding the surface volume in order to use a volume balance to determine an advance curve. This has advantages over strictly empirical equations since the assumptions regarding the surface volume can be verified with field observations or from computer simulation. During advance, the cumulative infiltrated volume is equal to the difference between the accumulated inflow volume and the surface storage volume at that time. This volume balance relationship may be expressed as: V in (t ) = V y (t ) + V z (t )

(14.13)

where Vin(t) is the inflow volume at time t, Vy(t) is the volume in surface storage at time t, and Vz(t) is the infiltrated volume at time t. Using this relationship to determine advance time, tx, to distance x requires that the surface and subsurface volumes at any time during advance are known. Typically, Equation 14.13 is put in the following form: Qin t x = σ y A0 (t x ) x + σ z Wz (t x ) x

(14.14)

where Ao(t) is the cross-sectional flow area at the inlet at time t, σy is the surface storage shape factor, W is the width, z is the infiltrated depth, and σz is the subsurface shape factor (all units are in meters and seconds). The surface shape factor is the ratio between the average cross-sectional flow area and that at the head of the field. For furrows, it is typically between 0.70 and 0.80. The subsurface shape factor is the ratio between the average infiltrated crosssectional area (infiltrated depth times width), and the infiltrated cross-sectional area

508

Chapter 14 Design of Surface Systems

(depth times width) at the head of the field. When infiltration is defined by Equation 14.5, the subsurface shape factor is difficult to determine. It is easier to rewrite the subsurface volume in the following form:

h ⎛ ⎞ V z = W ⎜ c + σ z1 k t xa + bt x ⎟ x 1+ h ⎝ ⎠

(14.15)

where h is the exponent on the advance equation, t x = s xh

(14.16)

where s is a constant. Then σz1 can be found from (ASAE, 1991)

σ z1 = =

h + a(h − 1) + 1 (1 + a)(1 + h)

(14.17)

Determining an advance curve requires knowledge of Qin, A0, σy, the infiltration constants (c, k, a, and b), and the width. Defining this curve is still an iterative process since the advance exponent h and advance times are interrelated and must be solved for simultaneously. The main problem with determining advance from Equations 14.14 through 14.17 is that the A0 and σy are not known in general. For steep slopes, this method is reasonably good since after a short time, the surface volume is typically a small amount of the total inflow volume. Also, A0 can be assumed equal to the flow area at normal depth and σy can be estimated with sufficient accuracy. For milder slopes, the surface volume is often a large portion of the inflow volume, even at the time of cutoff. Also, A0 changes continuously during the irrigation event and in some cases never reaches normal depth (see Chapter 13). Similarly, σy can vary (see Chapter 13). On a unit width basis, A0 = y0 W. 14.2.9 Simulation Approach One method for more accurately determining advance is to simulate it with the unsteady flow equations, e.g., with the simulation programs SIRMOD (Walker, 1989) or SRFR (Strelkoff, 1990). Recession can be found either from continued solution of the equations, or from assuming that recession occurs at the same time everywhere. The latter approach was used in the level-basin design program, BASCAD (Boonstra and Juriens, 1988). BASCAD will search through various input parameters to arrive at a useful solution. It does not allow PEmin as a design input, but displays it as a design output. The above unsteady-flow simulation programs can also be used through trial and error by the user to arrive at a design solution. 14.2.10 Dimensional-Analysis Approach Because of the large number of variables involved in surface irrigation, it is virtually impossible to present the results of simulations for the range of possible field conditions. Dimensional analysis methods have been used to reduce the number of parameters so that meaningful results can be displayed on a limited number of graphs. Solutions have been developed for flat-planted level basins and open-ended sloping border strips, and are available in the form of computer programs BASIN (Clemmens et al., 1995) and BORDER (Strelkoff et al., 1996). No nondimensional solutions are currently available for furrows. This methodology is presented in some detail in Chapter 13.

Design and Operation of Farm Irrigation Systems

509

14.3 SLOPING-FURROW IRRIGATION With sloping-furrow irrigation, water advance must be sufficiently fast so that the downstream end will receive adequate water while the upstream end is not excessively over irrigated. However, advance that is too rapid can result in a large percentage of the applied water running off the field, unless inflow after completion of advance is reduced, for example with a cutback, surge, or cablegation system, or the runoff is collected for reuse. The Soil Conservation Service (USDA, 1984) provided the following guidance for limiting irrigation-induced erosion from furrows. Furrow slopes in areas of high rainfall should be great enough to allow adequate drainage (> 0.03%), yet not so great as to cause significant erosion. For erodible soils (e.g., silty soils), the maximum furrow slope should be limited to 60/(P30)1.3, where P30 is the 30-min rainfall in mm for a 2year frequency. This limit can be exceeded by about 25% for less erodible soils (e.g., sandy and clayey soils). Further, erosion can be limited by placing a limit on the irrigation stream size. The maximum flow velocity should be limited to 8 and 13 m/min for erosive and non-erosive soils, respectively. The relationship between velocity and flow rate can be obtained from Equation 14.7, since velocity is flow rate, Q, divided by flow area, A. 14.3.1 Design of Sloping Furrows with No Cutback or Tailwater Reuse Design of sloping furrows with runoff is very straightforward under the following assumptions. First, advance can be computed from Equation 14.13 or 14.14, which give an accurate estimate of advance times when a good estimate of the surfacevolume shape factor is made. Second, we can assume that the flow depth at the upstream end is at normal depth for the entire period of inflow and the friction slope equals the bottom slope, Sf = S0. These assumptions are particularly appropriate for slopes greater than 0.05%. Next, recession along the entire furrow length is assumed to occur immediately after cutoff for steeply sloped furrows, particularly above 0.5% slope. For flatter slopes, the recession at the downstream end will take some time, resulting in a greater infiltration depth there and underprediction of PEa. Adjustment to the application time and recession curve are made based on the surface volume at the completion of advance. Finally, this design approach assumes that the wetted width does not vary over the length of the furrow. That is, the infiltrated volume over a unit length of furrow is dependent only on the infiltration opportunity time and not on the depth of flow or wetted width. For heavy soils with significant lateral sorption, the wetted width from adjacent furrows overlap, making the wetted width per furrow essentially equal to the furrow spacing. For coarse-textured soils, wetted width can cause a significant reduction in infiltration along the furrow due to a reduction in flow depth as the flow rate gets smaller with distance from the head end. For this latter case, field design should not attempt to minimize runoff since this would drastically reduce furrow flow rates toward the downstream end. Other measures (e.g., return flow) are needed to improve performance in these cases. Design for a specific set of field conditions (i.e., infiltration, roughness, and required infiltration depth) is essentially by trial and error for furrow length, slope, and inflow rate. The furrow cross-sectional area and wetted perimeter must be specified as a function of flow depth (i.e., normal depth). Any function can be useful provided that it properly describes the cross-section. Trapezoidal and power-function shapes are

510

Chapter 14 Design of Surface Systems

commonly used. Then the normal depth for a given discharge can be found by trial and error from Equation 14.7. The flow area at the upstream end, A0, is then computed from the cross-section definition. Since this does not vary during inflow, the surface volume can be computed as a function of distance only. The assumed surface-volume method provides a relationship between advance distance x and advance time tx from Equation 14.14, where W is the furrow spacing, under the assumption that Qin, W, A0, σz, and σy are fixed. Fortunately, for most situations of interest σy does not vary significantly, generally falling between 0.7 and 0.8. However, σz can vary considerably and should be adjusted for the particular case of interest. For a power infiltration function, it can be estimated from Equation 14.17, which required a power advance function (Equation 14.16). Estimation of the exponent h in Equation 14.16 requires that at least two points on the advance curve be estimated. The design procedure starts with estimates for the time for the advancing stream to reach half the field length and the entire field length. From these two, the advance exponent can be computed from log(t L / 2 /t L ) (14.18) h= log(1/2) With Equations 14.15 and 14.17, we can compute the subsurface volume. A new guess for the advance times can then be estimated from the volume balance (e.g., Equation 14.14) and new values for h calculated. The procedure converges fairly rapidly. An alternate procedure which is computationally faster for spreadsheets is to estimate h, allow the spreadsheet to iterate on the time to achieve a volume balance at both distances independently, and manually iterate on h until it converges. To provide the required depth at the downstream end, the cutoff time is computed from Equation 14.12, with the last term (in brackets) assumed equal to zero. The application efficiency is computed as the required volume (required depth × furrow spacing × furrow length) divided by the applied depth (inflow rate × cutoff time). The volume of runoff can be estimated by computing a distribution of infiltrated depths, the associated infiltrated volume, and subtracting this volume from the inflow volume. The deep percolation volume is the infiltrated volume minus the required volume, since it is assumed that all points receive at least the required depth. The infiltrated volume can be computed from the infiltrated depths at the head, middle, and tail of the field, since the infiltration opportunity times are known there (i.e., recession is assumed to occur at cutoff). Better estimates can be made by closer numerical integration, for example by computing additional points along the advance curve. Example 1. Consider a field 400 m long with a slope of 0.002, infiltration (based on a width equal to the 1 m furrow spacing) described by k = 40 mm/hra and a = 0.35, and Manning roughness of 0.05. (Carefully smoothed and compacted furrows can have a roughness as low as 0.03.) The furrow spacing is 1 m and the furrow shape is defined as a trapezoid with a bottom width of 100 mm with 2:1 side slopes, horizontal to vertical. Initial design will be made for a required depth of 80 mm. The resulting required opportunity time is 435 min. This design example starts by assuming an inflow rate of 1.0 L/s. Iterative solution of Equation 14.7 gives; a normal depth of 53 mm, a wetted perimeter of 338 mm, a

Design and Operation of Farm Irrigation Systems

511

velocity of 5.5 m/min (well within the acceptable range), and an upstream crosssectional flow area of 0.011 m2. For sloping furrows, we assume that the upstream depth reaches normal depth quickly and remains there. Then the surface volume during advance is not a function of time, and Vy(t) becomes Vy(x). If we assume that σy = 0.75, then from the expression for Vy in Equation 14.14, we get surface volume of 1.65 and 3.29 m3 for advance to 200 and 400 m, respectively (e.g., 0.75 × 0.011 × 200 = 1.65). The design procedure starts with a guess for the advance time to half the field length and to the field end. For this example, we arbitrarily guess 100 and 250 min, respectively. For a flow rate of 1 L/s, this represents inflow volumes of 6.0 and 15.0 m3. From Equation 14.18, we get h = 1.32. With h = 1.32 and a = 0.35, Equation 14.17 gives σz = 0.777. (Note that for the Kostiakov equation σz and σz1 are the same.) The infiltrated depths at the head of the field at 100 and 250 min are 47.8 and 65.9 mm, respectively. Substituting these into the expression for Vz in Equation 14.14, we get subsurface volumes of 7.43 and 20.48 m3 at these two times, respectively. At 100 min, the computed surface and subsurface volumes total 1.65 + 7.43 = 9.08, which is much larger than the inflow volume of 6.0. Thus the advance time to 200 m must be much longer. Similarly, at 250 min Vy + Vz = 23.77 m3 > Vin = 15.0 m3. If we adjust the advance times to remove the volume error, our next guess of advance times would be 100 × (9.08/6.0) = 151 min and 250 × (23.77/15.0) = 396 min, respectively. If we repeat the calculations, we get h = 1.39, σz = 0.783, Vin = 8.65 and 24.25, Vy + Vz = 10.30 and 27.55, for the two advance locations, respectively. Clearly, this linear adjustment is too small. This is due to the fact that the infiltrated depth also changes with the change in time. From this point, if we take 11/2 times the correction we get advance times of 181 and 490 min, h = 1.44, σz = 0.787, Vin = 9.27 and 26.27, Vy + Vz = 10.92 and 29.56, for the two advance locations, respectively. Again the correction is insufficient, but much closer. The solution for advance finally converges as shown in Table 14.5. The cutoff time is then 494 + 435 = 929 min (15.5 h). The required volume is 32 m3 (400 m × 1 m × 0.080 m), while the applied volume is 55.7 m3 (0.0010 m3/s × 929 min × 60 sec/min), giving an application efficiency of 57.4% (32/55.7). This assumed surface-volume method can be used to compute the advance time to intermediate locations. The advance times computed from the power-function advance relationship derived from the two-point method usually do not exactly satisfy a volume balance, but can be used as a good first guess of the advance time. If we use the value of the advance exponent h from the two-point method, the advance times calculated at intermediate points are shown in Table 14.6. If we assume that recession occurs everywhere at the time of cutoff, then the depths infiltrated at the head, middle, and tail end of the field are 104, 97, and 80 mm, respect-

Q (L/s) 1.0 1.0

Table 14.5. Advance calculations for assumed surface-volume sloping-furrow design example. Vin(t) Vy(t) σz A0 h x tx (min) (m3) (mm) (m3) (m) 200 182.3 10.94 0.011 1.65 1.44 0.787 400 494.0 29.64 0.011 3.29 1.44 0.787

Vz(t) (m3) 9.29 26.35

512

Chapter 14 Design of Surface Systems

Table 14.6. Advance and recession for sloping-furrow design example (Qin = 1 L/s), with recession times adjusted. Distance Advance Time Recession Time Opportunity Time Infiltrated Depth (m) (min) (min) (min) (mm) 0 0 874 874 102 50 25 881 856 101 100 67 888 820 100 150 121 894 774 98 200 182 901 719 95 250 251 908 657 92 300 327 915 580 89 350 408 922 514 85 400 494 929 435 80

tively. Numerical integration with the trapezoidal rule (i.e. numerically averaging depths over the two intervals) gives an average infiltrated depth of 94.5 mm and infiltrated volume of 37.8 m3. The runoff volume is the difference between the applied and infiltrated volumes, or 55.7 – 37.8 = 17.9 m3, or 32.1%. The deep percolation volume is 37.8 – 32 = 5.8 m3, or 10.4%. Better estimates for these volumes can be found by dividing the field into several increments and computing advance either from the volume balance or from Equation 14.16 (the difference is insignificant). In this example, dividing the field into eight increments (as shown for advance in Table 14.6) for advance and recession gave runoff and deep percolation volumes of 31.7% and 10.9% (rather than 32.1% and 10.4%). Our assumption with the method is that the volume of water on the surface at the time of cutoff is insignificant. In this example, the slope is relatively shallow and the volume on the surface at cutoff from Table 14.5 is 3.3 m3, or 5.9%, not an insignificant volume. If we compare these results with those from simulation, we find that recession at the downstream end takes more than an hour, implying that we could cut off 1200

Time (min)

1000 800 Vol-Bal Advance

600

Vol-Bal Recesssion SRFR Advance SRFR Recession

400

SRFR-adjusted cutoff Vol-Bal adjusted cutoff

200 0 0

100

200

Distance (m)

300

400

Figure 14.2. Advance and recession curves for sloping-furrow design example.

Design and Operation of Farm Irrigation Systems

513

the stream much earlier. Figure 14.2 shows the advance and recession curves from the volume balance and from simulation with the SRFR program. Note the excellent agreement in the two advance curves, associated with the assumed surface-volume and the unsteady-flow simulation. The computed runoff and deep percolation volumes were 29.4% and 13.3%, respectively. For this relatively mild slope, the effect of the recession curve is significant, as reflected by the shift in runoff and deep percolation volumes. An adjustment to the recession curve can be made by subtracting the volume of water on the surface at the time of cutoff from the total applied volume. Then the application time is found from ⎛ V y ( L) ⎞ (14.19) ⎟ t co = t A ( L) +τ req − ⎜ ⎜ Q ⎟ in ⎠ ⎝ where Vy(L) is the steady state surface volume for the inflow rate after advance is complete. For convenience, we use the surface volume computed when advance is just complete. The recession time at the upstream end is tco and the infiltration opportunity time at the downstream end is assumed equal to τreq. This results in a cutoff time of 856 min. (14.3 h), an applied volume of 55.7 – 3.3 = 52.4 m3, and an application efficiency of 61.0%. Here the recession curve is not horizontal, but can be assumed linear between the upstream and downstream ends (Table 14.6). This assumption, with advance and recession computed over eight intervals, gives runoff and deep percolation of 28.3% and 10.7%, respectively. From simulation, this design is closer to matching the required depth at the downstream end, as shown in Figure 14.2, than assuming instantaneous recession. Further, simulation resulted in runoff and deep percolation volumes of 26.9% and 12.2%, respectively. It appears that the inflow volume could be reduced even more than the volume on the surface. However, this simple procedure provides reasonable results and is generally on the conservative side. For furrows with a steeper slopes, this adjustment will be small and is not necessary. With the runoff percentage significantly higher than deep percolation percentage, one would expect a higher potential application efficiency with a smaller flow rate. Repeating these calculations at 0.7 L/s results in an application efficiency of 64.4%, tL = 808 min (13.5 h), tco = 1182 min (19.7 h), and runoff and deep percolation of 18.1% and 17.5%, respectively. A field is typically broken up into an integer number of irrigation sets, with the total inflow divided among the furrows. Thus furrow flow rate is not a continuous variable, but typically can only take on discrete values. A field is also broken into an integer number of lengths, making furrow length also not a continuous variable for design. 14.3.2 Design of Sloping Furrows with Cutback The efficiency of furrow irrigation systems can often be improved by reducing the inflow rate after water has advanced to the end of the field. A high initial flow rate can provide rapid advance, and thus more uniform opportunity time, while cutting back the stream will reduce the amount of runoff. If the cutback stream is too small to keep up with infiltration, recession can occur at the downstream end and move back up the field. Rather than cutting off at the completion of advance, cutoff is typically delayed until infiltration is somewhat reduced. This will also help assure that the downstream end receives sufficient flow depth and wetted perimeter.

514

Chapter 14 Design of Surface Systems

A common practice is to cut back to 50% of the inflow. Dividing the cutback inflow rate by the wetted field area gives the average infiltration rate that matches the cutback inflow Q iCB ≤ CB (14.20) WL The average infiltration rate at any time after completion of advance can be computed from numerical integration over the distance (e.g., the eight intervals used above). A direct analytical solution for average infiltration rate is not possible, but a numerical approximation can be found by integrating infiltration rate over distance. Infiltration at any point and any time is a function of the infiltration time, or the current time minus advance time. The advance time is a function of xh, from Equation 14.16. If this term is replaced with a truncated series expansion, with higher order terms removed (e.g., xh = 1 + h(x – 1)), an analytical expression for the infiltration rate, averaged over the field length, can be found, namely iCB

aktCB ( a −1) ≈b+ ah

a a ⎡⎛ ⎧ ⎞ ⎛ ⎞ ⎤ ⎢⎜ ⎨ tCB ⎫⎬ − 1 + h ⎟ − ⎜ ⎧⎨ tCB ⎫⎬ − 1⎟ ⎥ ⎟ ⎜ t ⎟ ⎢⎜⎝ ⎩ t L ⎭ ⎠ ⎝ ⎩ L ⎭ ⎠ ⎥⎦ ⎣

(14.21)

where tCB is the time of cutback. The cutback time to achieve the necessary average infiltration rate can be determined from Equation 14.21 by trial and error. This is a very conservative estimate of cutback time since the reduction in flow results in less surface storage on the field, and this change in surface storage can contribute to infiltration. A conservative estimate of the correction in cutoff time can be found by dividing the change in surface volume by the cutback flow rate, which is related to the distance averaged infiltration rate through Equation 14.20. The adjusted cutback time is simply V y (Qin ) − V y (QCB ) (14.22) tCB adj = tCB − QCB where the surface volume after completion of advance is now a function only of flow rate. The cutoff time is computed according to Equation 14.19, but with Vy(Qin) replaced with Vy(QCB). Because less volume is on the surface during cutback, the cutoff time is actually slightly longer. For some soils, the reduction in wetted perimeter caused by cutback might require an increase in total application time. Example 2. For a 50% cutback flow of 0.5 L/s, Equation 14.20 indicates that the average infiltration rate would need to be below 4.5 mm/hr: 3600sec 1m 3 1000mm 0.5L/s × × × = 4.5mm/hr 1hr 1000L 400m × 1m m From Equation 14.22, the time of cutback is 582 min. (Trial and error solution with numerical integration of the average infiltration rate over eight intervals gives tCB = 609 min.) The surface volume for the cutback flow is 1.98 m3. Equation 14.22 gives an adjusted cutback time of 538 min. The adjusted cutoff time, from Equation 14.19

Design and Operation of Farm Irrigation Systems

515

with Vy(QCB) replacing Vy(L), is 863 min. The total inflow volume is now 42.0 m3 rather than 55.7 m3. This gives an application efficiency of 76.1%. Runoff is reduced to 4.53 m3, or 10.8%, while deep percolation is 13.1%. SRFR simulation gave 9.4% and 14.4% for runoff and deep percolation, respectively. Advance, recession and the distribution of infiltrated water are shown in Table 14.7. Table 14.7. Advance and recession for sloping furrows with cutback design example (Qin = 1 L/s). Distance Advance Time Recession Time Opportunity Time Infiltrated Depth (m) (min) (min) (min) (mm) 0 0 863 863 102 50 25 871 845 101 100 67 879 810 100 150 121 888 767 98 200 182 896 713 95 250 251 904 654 92 300 327 912 586 89 350 408 921 513 85 400 494 929 435 80

14.3.3 Design of Sloping Furrows with Tailwater Reuse Furrow irrigation systems on sloping fields usually need some runoff in order to provide a water depth at the downstream end to get sufficient wetted perimeter and lateral water movement. The design procedures used here ignore this level of detail and can sometimes give misleading results, particularly when cutback flows are used. Tailwater reuse provides another mechanism for providing a sufficient flow rate and water depth at the downstream end, while still improving furrow application efficiency. Tailwater can be reused in fields which are downstream from the field where the tailwater is generated (e.g. cascaded from field to field). Alternately, the tailwater from a field can be collected in a storage pond, pumped back, and reused on the same field. Improvements in irrigation efficiency resulting from tailwater reuse depend upon the type of system, the amount of the tailwater that is reused, and the number of times it is reused. Solomon and Davidoff (1997) developed equations for determining the combined efficiencies associated with the reuse of water through a series of areas. Applied to tailwater reuse systems, the application efficiency of the combined system with reuse, Ea_RU, can be estimated from the application efficiency from a single irrigation set from E a _ RU =

Ea ⎛ n −1 ⎞ 1 − RU ⎜ ⎟ ⎝ n ⎠

(14.23)

where RU is the fraction of the applied water that is reused and n is the number of subunits from which water is reused. For a cascaded reuse system, n is the number of systems (e.g., individual fields, farms, projects, etc.) irrigated in sequence. For a recycling reuse system, n is infinity and the term (n – 1)/n becomes unity. With a recycling runoff reuse systems, a very high runoff fraction can result in excessive pumping. The volume of water pumped or recycled, VRC, relative to the inflow

516

Chapter 14 Design of Surface Systems

volume can be estimated from the reused runoff fraction by VRC = Vin

RU 1 − RU

(14.24)

There are two main types of recycling runoff reuse systems: those that recycle runoff at a rate nearly the same as the runoff rate, and those that recycle at a constant rate. The advantage of the runoff-rate type is that storage volume is minimized. Two methods used for these systems are to cycle the pump on and off at short intervals, the exact interval related to the rate of runoff. Maximum recommended cycle rates for these systems are 15 times per hour. Sump surface area depends upon the maximum runoff rate and the allowable depth fluctuations between cycles. The other method for runoffrate recycling systems is to use variable speed pumps (Trout and Kincaid, 1994). For these systems, runoff should be applied to another field or a different irrigation set, or pumped to a larger, upstream reservoir. It is not recommended to recycle water back into the supply stream for the field that is creating the runoff, unless additional furrows are started as return flow increases (ASAE, 1997). Using a variable returnflow rate to irrigate additional furrows is difficult to manage and increases labor, and therefore is not recommended. Sump sizes for these systems are on the order of tens of cubic meters. The intent of a constant flow recycling system is to provide a constant flow rate to all furrows. This can be accomplished by pumping water from a reuse storage reservoir and adding it to the normal irrigation stream from the beginning of the irrigation. The return-flow rate should be such that the return-flow volume for one set is approximately equal to the runoff volume for one set. Then the reservoir will be at approximately the same level at the end of the irrigation as it was at the beginning. This essentially requires a full reservoir at the start of the irrigation. Also, there must be a sufficient volume of water in storage to provide the return flow until the runoff rate from the first set matches or exceeds the pumping rate. This operating scheme might be practical where several fields share the same return-flow system and reservoir. However, for many systems this is not practical. Rather than starting with the reservoir full, the same general scheme can be used with the reservoir initially empty. This is accomplished by splitting the width of the first set into a portion irrigated with the supply inflow only and a portion irrigated with the reuse portion only. The second part of the first set is actually irrigated after the main supply flow is turned off and irrigated with return flow only (i.e., it is moved to the end of the irrigation and the end of the field). This scheme was first proposed by Stringham and Hamad (1975). For a given field (length, width, slope, infiltration) and a given inflow rate Qin, the design of these systems is based on balancing three parameters: the desired infiltrated depth (e.g., average, di or low quarter, dlq); the individual furrow flow rate, q; and the number of irrigation sets across the field width, N (an integer number). These cannot be selected independently, but must assure that the set runoff and return-flow volumes approximately match. Once these three are determined, the other design parameters can be found. A number of alternative combinations can be examined to find a workable combination (e.g., optimum). The number of sets, the individual furrow flow rate, Qf, and relationship between the supply inflow rate, QS, and the return-flow or pumpflow rate, QP, should satisfy:

Design and Operation of Farm Irrigation Systems

517

⎛Q Qp ⎞ ⎟ = N ns + n p F = N⎜ s + ⎜ Qf Qf ⎟ ⎝ ⎠

(

)

(14.25)

where F is the total number of furrows in the field, N is the number of sets, and ns and nf are the number of furrows in one set irrigated with the supply and pumpback flows, respectively. The selection of return-flow rate must assure that there is sufficient runoff volume which is satisfied when RO ≥

Qp QS + Q P

⎛ RO ⎞ or QP ≤ ⎜ ⎟QS ⎝ 1 − RO ⎠

(14.26)

where RO is the runoff fraction (i.e., fraction of inflow that runs off for an average furrow). If this equation isn’t satisfied, there may not be enough water in the reservoir to irrigate the last set. When this inequality is satisfied as an equality, the pumping rate is the maximum allowable pumping rate, QP-max. If the pumping rate is less than the maximum allowable, a portion of the runoff volume will not be returned to the same field. Here, we assume that this portion of the runoff volume is unused or a loss. The reservoir volume, VR, needed for operation of such systems is the amount of water required in the reservoir to irrigate the last set with only return flow, VR ≥ QP (1 − RO)t co ≤ Qs ROt co

(14.27)

where tco is the application time for each set. When Equation 14.26 is an equality, this volume equals the volume of runoff from the first set, irrigated without return flow, and the right inequality of Equation 14.27 is an equality. Note that the volume required in the reservoir at the start of this last set is less than the volume needed to irrigate the last set. For that set, its own runoff is recirculated to provide the necessary volume. The runoff volume lost is equal to the total volume applied times the difference between the actual runoff fraction and the reused runoff fraction Vin(RO –RU). The method presented above for determining sloping-furrow advance and recession curves can be used to determine the runoff fraction for a given furrow flow rate and application time. This information can be used to determine whether this satisfies the desired depth of infiltration and Equations 14.25 and 14.26. The application efficiency can be adjusted according to Equation 14.23 with n = ∞ and RU = RO(QP/QP-max), or some fraction thereof to account for return-flow system losses. This assumes that only the water pumped during this irrigation event is reused. Example 3. Consider the example given in the previous sections. In this case, there are a total of 800 furrows (800 m at 1 m/furrow) and the available flow rate is 80 L/s. An initial trial at 1 L/s/furrow gives the advance and recession curves of Table 14.6 and nS = 80 furrows. From Equation 14.26, the maximum return-flow pumping rate is 31.6 L/s. Choosing QP = 31.6 L/s results in nP = 31.6 and N = 7.2 (800/[80 +31.6]), which is not an integer and so is not feasible. If the solution were feasible, we would get RO = 28.3%, PEa = 85.1% (from Equation 14.23), and a reservoir volume of 1188 m3. One way to make this solution feasible is to choose a smaller pumping rate. At QP = 20 L/s, nf = 20 and N = 8, and a smaller reservoir is needed. However, then only 20/31.6 or 63% of the runoff is reused. If only the reused volume were considered

518

Chapter 14 Design of Surface Systems

useful, then PEa would drop to 74.3% with a corresponding runoff volume lost would be 3480 m3 (Table 14.8). This emphasizes the need to satisfy the inequality in Equation 14.26 as close as possible to an equality. A number of adjustments can be made in the other variables to produce an integer number of sets, with a better fit to Equation 14.26. Table 14.8 provides the results of changing various parameters on PEa, VR, and VRO-lost. If the inflow could be increased to 82 L/s, a feasible solution would result (Table 14.8, run 3). The target application depth could be changed, either increased slightly or decreased slightly (runs 4 and 5). The lower efficiency for run 5 reflects the lower uniformity associated with the reduced target depth. Or, the individual furrow flow rate could be adjusted to arrive at feasible solutions, as in run 6. In run 7, the flow rate decrease did not provide a very useful solution without also a change in the required depth; run 8. Selection of the best alternative from among these and other potential solutions should be based on economic considerations; labor, reservoir construction, pumping plant facilities, etc. Effective operation of these systems requires a matching of runoff to return-flow pumping rate. Monitoring of runoff volume can be used to adjust set duration and/or pumping rate to assure efficient use of the runoff (i.e., a balancing of Equation 14.26).

Run 1 2 3 4 5 6 7 8

Table 14.8. Examples of solutions for tailwater recovery system design. QS dreq Qf tco RO QP_max QP nS nP N RU PEa T VR (L/s) (mm) (L/s) (min) (%) (L/s) (L/s) (%) (%) (hrs) (m3) 80 80 1 874 28.3 31.6 31.6 80 32 7.2 28.3 85.1 104 1188 80 80 1 874 28.3 31.6 20 80 20 8 17.9 74.3 117 752 82 80 1 874 28.3 32.4 32 80 32 7 28.0 84.7 102 1203 80 83 1 922 30.4 35.0 34.0 80 34 7 29.6 85.2 108 1309 80 68 1 713 20.1 20.1 20 80 20 8 20.0 79.5 95 684 80 80 1.2 720 34.3 41.8 40 67 33 8 32.8 85.9 103 1135 80 80 0.97 890 27.5 30.2 17.6 82 18 8 15.9 73.1 119 681 80 85 0.97 973 31.0 35.9 31.4 82 32 7 27.1 81.9 114 1265

VRO-lost (m3) 0 3480 96 256 21 510 3952 1268

14.4 BORDER-STRIP IRRIGATION With border-strip irrigation, application efficiency typically is reduced when advance is either too fast or too slow. Thus design needs to provide a layout such that inflow rate and time can be adjusted within reason to provide satisfactory performance. The Soil Conservation Service (USDA, 1974) provided the following recommendations. The maximum recommended inflow rate to limit erosion on non-sodforming crops, such as alfalfa and small grain, is found from qin

max

= 0.00018 S 0−0.75

(14.28)

where S0 is in m/m and qin is in m2/s. For sod-forming crops, twice this value can be used. A minimum inflow rate has also been suggested so that the water depth will be sufficient to spread laterally: qin

min

=

0.000006 L S10/ 2 n

(14.29)

Design and Operation of Farm Irrigation Systems

519

Border-strip irrigation is typically practiced on slopes less than 0.05 m/m (5%). On fine-textured soils, slopes are typically less than 0.01 m/m. The maximum slope based on the criteria for minimum flow depth (and discharge) can be found by solving Equation 14.29 for S0. This does not consider erosion potential. 14.4.1 Design of Open-Ended Border Strips For short border strips and steep slopes, the minimum depth infiltrated can be at the head end of the strip. However, numerous runs with the BORDER design program indicated that for most situations when near the peak potential efficiency, the minimum depth was at the downstream end. Even when the minimum was at the upstream end, the range of low-quarter depths was split between the upper and lower ends. Thus design based on the downstream end should give more consistent results over the range of typical design conditions. Since, in addition, estimates for potential efficiency are needed, recession lag-time design (or design based on satisfying the requirement at the upstream end, as in USDA [1974]) is no longer recommended. Simple volume-balance procedures can be used for border strip design, as was done for furrow design, under the assumption of a surface shape factor. For sloping borders, we can assume that the flow depth at the upstream end approaches normal depth. The normal depth can be computed from Equation 14.8 with q as the unit inflow rate and Sf set to the bottom slope, S0. Solving for y gives y0 =

3/5 (n / Cu )3/5 qin

S 03/10

(14.30)

As with furrows, Vy = σy y0 W, where σy ≈ 0.7. The remaining problem in solving the volume balance is computing the recession curve, where the recession time varies with distance. The recession time at the downstream end is set so that the required depth is just satisfied there, as in Equation 14.12. An empirical relationship, adapted from Walker and Skogerboe (1987), is used to determine the cutoff time for this required downstream recession time. An approximate upstream recession time is found from

Δt R = t R ( L) − t R (0) =

0.666 n 0.4757 S 0y .2074 L0.6829 I 0.5244 S 00.2378

(14.31)

where all units are in meters and seconds, and Sy is 1 S y= L

⎡ (qin − IL)n ⎤ ⎢ ⎥ ⎢⎣ S 0.5 ⎥⎦ 0

0.6

(14.32)

and I is the infiltration rate (m/s) at tR(0)S, averaged over the length. For the branch infiltration function after the branch point, it is simply b. For the other infiltration functions, the infiltration rate can be numerically integrated or approximated by averaging the values at the upstream and downstream ends, τ(0) = tR(0)S and τ(L) = tR(L) – tA(L). Equation 14.31 is essentially an empirical fit to a series of computer runs over a wide variety of conditions. It is based on an estimate of the upstream recession lag time from (Strelkoff, 1977), which results in a cutoff time of

520

Chapter 14 Design of Surface Systems

tco = t R (0) S −

y0 L

(14.33)

2 q in

Solution of Equations 14.31 and 14.32 for tR(0)S is essentially a trial-and-error process if the average infiltration rate is not fixed (i.e., with the Kostiakov branch function, the trial and error is not needed). While the upstream recession time from Equation 14.31 is needed for the procedure used to compute cutoff time, it does not give a very realistic estimate for the actual upstream recession time at small slopes. The following equation can be used to estimate the upstream recession time (adapted from Hart et al., 1980): t R (0) = tco +

1.2 q0.2 in n

⎡ ⎛ 0.345 n q0.175 ⎞⎤ in ⎟⎥ ⎢S 0 + ⎜ 1/2 ⎟⎥ ⎜ τ 0.88 ⎢ S req 0 ⎝ ⎠⎦ ⎣

1.6

(14.34)

where units are in meters and seconds. The recession lag times for steeper slopes are generally very small and either equation gives reasonable results. For smaller slopes (e.g., τB. The resulting equation is ⎧ bht A ( L) ⎫ V z = LW ⎨d req + ⎬ (h +1) ⎭ ⎩

(14.41)

The cutoff time is found by dividing this volume by the inflow rate, Qin. For very small values of AR, this equation is not appropriate, however in such cases the cutoff time can be based solely on required volume (i.e., PEmin = 100%). This procedure assumes that advance is complete prior to cutoff. For large level basins as used in the U.S., this is frequently not the case, particularly when flow resistance is high (e.g., alfalfa or grass). One of the biggest errors associated with this procedure is that the surface volume during advance is large. This is particularly true when cutoff precedes completion of advance. Example 6. The assumed surface-volume method given above was applied to the design example for border strips, except that the slope is zero. (Such light application depths are more difficult to apply with level basins unless the length is significantly reduced. Variations in field surface elevations also play a role in potential efficiencies if attempts are made to apply too little water.) The length is reduced to 200 m and the flow rate to 2.0 L/(s m). The surface shape factor is 0.80. The advance curve computed with the assumed surface-volume method (Equations 14.38 and 14.39) is given in Table 14.12. From Equation 14.40, the total volume applied is 20.23 m3, resulting in a cutoff time of 169 min and PAEmin = 79.1%. If we assume that recession occurs at the time required to infiltrated the desired depth at the far end (195.4 + 232.4 = 428 min), opportunity times at various points along the basin can be computed by subtracting the advance time from this recession time, as shown in Table 14.13. Using these computed points to estimate the volume infiltrated gives (with eight-point numerical integration) PAEmin = 80.5%, suggesting that Equation 14.40 is a reasonable estimate for the required volume. (Walker and Skogerboe’s method gives PAEmin = 60.1%.)

526

Chapter 14 Design of Surface Systems

Q (L/(s m)) 2.0 2.0

Table 14.12. Results for assumed surface-volume level-basin design. x tx Vin(t) y0 Vy(t) Vz(t) σz (m) (min) (m3/m) (mm) (m3/m) h (m3/m) 100 72.2 8.66 68.5 5.47 1.44 0.723 3.18 200 195.4 23.45 80.4 12.86 1.44 0.723 10.58

Table 14.13. Advance and recession for level-basin design example. Distance Advance Time Recession Time Opportunity Time Infiltrated Depth (m) (min) (min) (min) (mm) 0 0 428 428 109 25 8 428 420 108 50 21 428 407 106 75 38 428 390 104 100 58 428 370 101 125 80 428 348 98 150 104 428 324 95 175 131 428 297 91 200 195 428 232 80

500

Time (min)

400 Vol-Bal Advance

300

Vol-Bal Recession SRFR Advance

200

SRFR Recession

100 0 0

50

100

Distance (m)

150

200

Figure 14.4. Advance and recession curves for level-basin design example.

For this example, the advance time exceeds the application time. This assumed surface-volume method likely overestimates PAEmin. The BASIN design program for this set of conditions gives PAEmin = 78.4%, tA(L) = 211 min, tco = 170 min, and cutoff at advance to 88% of the basin length (R = 0.88). This is close to the limit line presented by Clemmens and Dedrick (1982), which occurs at approximately R = 0.85, and beyond the limit in BASCAD for design. While the differences are small in this case (2% change in PAEmin), errors increase rapidly as cutoff occurs at shorter advance distances. A comparison of advance and recession curves for the assumed surfacevolume method and unsteady-flow simulation are given in Figure 14.4.

Design and Operation of Farm Irrigation Systems

527

14.5.2 Design of Level Furrows and Furrowed Level Basins The same assumed surface-volume method can be used for level furrows or furrowed level basins as was used for flat-planted level basins, except Equation 14.38 is replaced with (from Equations 14.7 and 14.37) 2 A02 R04 / 3 y0 = qin ( n / Cu )2 x

(14.42)

Since A0 and R0 are functions of y0 as defined by the furrow cross-section shape, determination of A0 for the volume balance calculation is iterative rather than direct (i.e., it adds one more internal iteration loop to the procedure). While the design procedures for the two systems are essentially the same (i.e., based on advance and recession of a single furrow), the cultural and irrigation operations for level furrows and furrowed basins are quite different. Example 7. The infiltration and roughness conditions for the sloping-furrow irrigation example are used here: k = 40 mm/hra, a = 0.35, and n = 0.05. The furrow spacing is 1 m, and as before, the furrow bottom width is 100 mm with 2:1 side slopes (horizontal:vertical). In this case, we start the design assuming 2 L/s per furrow and a length of 200 m. The design starts with a solution for the depth at the upstream end as a function of distance x, from Equation 14.42. This relationship does not depend on the advance time. At 100 m, we guess a depth of y0 = 0.1 m. For a trapezoidal shape, this results in an area A0 of 0.0485 m2, a wetted perimeter of 0.6935 m, and a hydraulic radius R0 of 0.0699 m. Solving Equation 14.42 for y0 gives 0.0148 m. This new estimate of y0 does not provide a good value for the next guess (i.e., the solution diverges). Instead, we use 0.8 times the original guess plus 0.2 times the value computed from Equation 14.42; 0.8(0.1) + 0.2(0.0148) = 0.1091 m. After three iterations, the solution converges to 0.0891 m. At 200 m, solution with this procedure gives y0 = 0.1012 m. The surface volume for advance to these two distances is found from Equation 14.14, with σy = 0.80, as 1.983 and 4.897 m3, respectively. The subsurface volumes are a function of time and the advance exponent. Initial guesses of 60 and 250 min give h = 1.32, and from Equation 14.15, subsurface volumes of 3.107 and 4.897 m3, respectively. The inflow volumes at these times are 7.2 and 18.0 m3, respectively, or volume errors of more than 20%. From Equation 14.39, new estimates of advance time are 42 and 122 min, respectively. The solution eventually converges to 39 and 104 min, respectively, h = 1.415 and an advance ratio AR = 0.239. From Equation 14.40 with a target depth of 80 mm, the ultimate infiltrated volume is 16.89 m3, which occurs at a cutoff time of 140.7 min, or approximately 37 min after the completion of advance. The potential application efficiency for this case is 94.7%. Numerical integration with eight points on the advance and recession curve, assuming recession occurs at the time needed to infiltrate 80 mm at the downstream end, gives an infiltrated volume of 16.75 m3, an application time of 139.5 min and a potential application efficiency of 95.6%. Clearly in this case, either a longer basin or a smaller application depth can be applied. Further application of this procedure at different lengths, furrow flow rates, and application depths should be explored to arrive at an economic design.

528

Chapter 14 Design of Surface Systems

14.6 SURFACE IRRIGATION SYSTEM HEADWORKS AND CONTROL OF INFLOW The headworks for a surface irrigation system are used to convey and distribute water from the supply point to individual irrigation sets, furrows, etc. They provide a transition from the supply to the field, and must be compatible with both the type of supply available and the type of surface irrigation system in use. There are two main types of supply: from a farmer-operated groundwater well and from an irrigation district. The irrigation district supply may be from a pressure pipeline or from an open canal or pipeline. Flow from a groundwater well is typically constant and nonadjustable. Flow from an irrigation district with open canals or pipelines is typically variable and not under control of the irrigator. Flow from an irrigation district pressure pipeline may be adjustable by the irrigator (e.g., by applying backpressure on the outlet). Farm field conveyance and distribution works may be a pipeline or an open ditch. Water from groundwater can be pumped into a field pipe distribution system or an open canal. Where sufficient head is available, water from canals and open pipelines can also supply field pipe distribution systems (e.g., gated pipe). However, debris must be kept out of the pipeline since it could clog outlets. Turbulent fountain screens are an effective way to achieve this (Bondurant and Kemper, 1985). Where earthen farm ditches are used, water can be distributed to individual furrows with siphon tubes or spiles; or to individual borders or basins through siphon tubes, pipe outlets (such as large spiles that are closable) or cuts in the canal wall. For concrete ditches, options for water distribution include siphon tubes (for all methods), pipe outlets (for borders and basins), or open gated outlets (e.g., with jack gates). For farm pipe conveyance and distribution systems, alfalfa valves are common for borders and basins, while gated pipe systems are most common for furrows. The selection from among these various alternatives should be based on compatibility with the supply and water control to the irrigated field. This latter must consider the available head of the water above the field surface, which may change substantially over the farm, so that water will flow out of the distribution system into the desired field area and not elsewhere. Systems should typically have 0.15 to 0.3 m of head available at all points. Outlets should also be non-erosive. This can be a severe problem with steep slopes and erodible soils, or where very high flow rates are used (e.g., > 300 L/s). The headworks must often supply fields that may irrigate both row and field crops, such that, for example, both furrow and border-strip irrigation are practiced on the same field. The headworks should not limit this possibility. Selection of headworks is often dictated by common practices in the area and economics. However, modernization or improvement in surface irrigation practices usually requires better control of inflow and may suggest other alternatives. Control of inflow rate is necessary to help balance advance and recession curves and to minimize runoff that is lost. Is some cases, some form of automation will help improve water control. Cutback furrow irrigation has not been widely practiced because of the labor involved in making effective cutback. Several methods have been developed to accomplish cutback automatically or semi-automatically. Surge flow was developed in attempts to apply cutback flow to gated pipe. With these systems, if the pipe gates are

Design and Operation of Farm Irrigation Systems

529

set to give the desired distribution at full flow (i.e., gate openings vary due to slope, wheel vs. non-wheel row, etc.), reducing the pipe flow in half will not give each furrow half of its previous flow. By cycling the full flow at a relatively short interval (e.g., 10 min), a more effective cutback is achieved. Automatic surge valves that cycle water have made surge flow, and the associated cutback, a practical method for improving sloping-furrow irrigation performance. Cablegation is another method of achieving an effective cutback flow. Under these systems, gated pipe is placed on a uniform slope. A plug moves at a constant speed down the gated pipeline. Immediately upstream from the plug, the outlets have essentially full pressure. Further upstream, outlets have a reduced pressure. By allowing the plug to move downstream, the flow gradually decreases, essentially producing a gradual cutback. Different flow distributions can be produced with different supply inflows and plug speeds (Kincaid, 1984). Automation only requires control over the plug speed. These systems are not in common use. Cutback flows from field ditches was first proposed by Garton (1966), where spiles for subsequent irrigation sets are located at a lower elevation. When the next set is irrigated, the current set has a lower head on it, producing the cutback flow. These have not proven to be practical in the field, partly because of physical limitations (i.e., they need to have the correct cross-slope), their relative nonadjustability, and plugging of the spiles with debris. Canal side weirs or ditch notches have been used effectively in some areas. Here small weirs, one for each furrow, are made when the concrete ditch is poured. Flow into a set of furrows is controlled by the water surface elevation in the ditch. Blocking the canal gives full flow. Cutback flow can be achieved and controlled by unblocking the canal to allow most of the flow to continue downstream, but by maintaining the water level (e.g., with check boards). This system is also less susceptible to weed plugging. These systems have not seen widespread use due to the cost of construction. (Eftekharzadeh et al., 1987). Improved management of borders and basins usually requires selection of the right flow rate and duration. With level-basin irrigation, the tendency has been to increase the flow rate to speed advance and reduce labor costs. Level basins as large as 15 ha with flows above 1 m3/s have been used. Outlet erosion control becomes a major issue for such systems. With fixed-width borders and basins, flow rate adjustments are not always feasible, making control of duration the only operational control over performance. A number of methods have been proposed for the automatic control of duration to such systems, including automatic control of jack-gates, pipe outlets, and drop-open or drop-closed gates within a concrete-lined supply canal. Most of these use simple timers to move water from one set to another. Field sensors have also been developed to sense the arrival of water and signal the change in irrigation set (e.g., based on advance distance). These sensors are also not in common use. Drain-back level basins were developed to reduce many of the limitations of the more traditional, modern level basins used in southwestern Arizona. With these systems, an earthen ditch is placed below the field grade. When the ditch is blocked, the water level rises and simply flows into the field. When the ditch is opened, the supply water is conveyed past the field, below its surface elevation, to the next field. In addition, some of the applied water drains back off the field just irrigated, thereby reducing the applied depth. The conveyance ditch is used as the turn row. This system significantly reduces the cost of level-basin construction (i.e., no concrete ditch to construct),

530

Chapter 14 Design of Surface Systems

makes lighter water applications more feasible, and places less restriction on machinery operations (i.e., one can drive though ditches from one basin to the next) (Dedrick, 1984). More recently, similar systems have been utilized for level-basin irrigation in high rainfall areas, including paddy rice production. A common feature is below-grade ditches on one to three sides. These replace the contour levee system, in that rather than water flowing from levee to levee through the field, the water flows in the side ditches from levee to levee. This actually promotes both more efficient irrigation and fastest surface drainage (Clemmens, 2000). The headwork facilities are an extremely important aspect in system design since they often represent a major part of the development cost for a surface irrigation system. The surface irrigation design procedures presented herein should be used in conjunction with an economic analysis of all of the factors that influence the cost of development and operations. Consideration should also be given to the future pressure to reduce water applied and environmental protection. Designs that potentially allow better water control offer more flexibility to meet these changing demands. More details on these the considerations in irrigation system selection can be found in Burt et al. (1999).

REFERENCES ASAE. 1991. EP419: Evaluation of furrow irrigation systems. St. Joseph, Mich.: ASAE. ASAE. 1997. EP408.1: Design and installation of surface irrigation runoff reuse systems. St. Joseph, Mich.: ASAE. Bondurant, J. A., and W. D. Kemper. 1985. Self-cleaning, non-powered trash screens for small irrigation flows. Trans. ASAE 28(1): 113-117. Boonstra, J., and M. Jurriens. 1988. BASCAD A Mathematical Model for Level Basin Irrigation. ILRI Publication 43. Wageningen, The Netherlands: Int’l. Inst. for Land Reclamation and Improvement. Burt, C. M., A. J. Clemmens, R. D. Bliesner, J. L. Merriam, and L. A. Hardy. 1999. Selection of irrigation methods for agriculture. ASCE On-Farm Irrigation Committee Report. Reston, Va.: American Soc. Civil Engineers. Clemmens, A. J. 1991. Feedback control of a basin irrigation system. J. Irrig. Drain. Eng. 118(3): 480-496. Clemmens, A. J. 2000. Level basin irrigation systems: adoption, practices, and the resulting performance. In Proc. 4th Decennial Nat’l Irrigation Symp., 273-282. St. Joseph, Mich.: ASAE. Clemmens, A. J., and A. R. Dedrick. 1981. Estimating distribution uniformity in level basins. Trans. ASAE 24(5): 177-1180, 1187. Clemmens, A. J., and A. R. Dedrick. 1982. Limits for practical level-basin design. J. Irrig. Drain. Div., ASCE 108(IR2): 127-141. Clemmens, A. J., and A. R. Dedrick. 1994. Chapt. 7: Irrigation techniques and evaluations. In Management of Water Use in Agriculture, 64-103. K. K. Tanji and B. Yaron, eds. Adv. Series in Agricultural Sciences, Vol. 22. Berlin, Germany: Springer-Verlag. Clemmens, A. J., A. R. Dedrick, and R. J. Strand. 1995. BASIN: A Computer Program for the Design of Level-Basin Irrigation Systems. WCL Report #19. Phoenix, Ariz.: U.S. Water Conservation Lab., USDA-ARS. Dedrick, A. R. 1984. Water delivery and distribution to level basins. In Water Today

Design and Operation of Farm Irrigation Systems

531

and Tomorrow, Proc. Am. Soc. Civ. Eng. conf., 1-8. Reston, Va.: American Soc. Civil Engineers. Eftekharzadeh, S., A. J. Clemmens, and D. D. Fangmeier. 1987. Furrow irrigation using canal side weirs. J. Irrig. Drain. Eng. 113(2): 251-265. FAO. 2005. AQUASTAT online database, FAO’s Information System on Water and Agriculture. Food and Agriculture Organization, United Nations, Rome, Italy. Available at: www.fao.org/ag/agl/aglw/aquastat/dbase/index2.jsp. Garton, J. E. 1966. Designing an automatic cut-back furrow irrigation system. Okla. Agr. Bull. B-651. Hart, W. E., H. G. Collins, G. Woodward, and A. S. Humpherys. 1980. Chapt. 13: Design and operation of gravity or surface irrigation systems. In Design and Operation of Farm Irrigation Systems. M. E. Jensen, ed. St. Joseph, Mich.: ASAE. Hutson, S. S., N. L. Barber, J. F. Kenny, K. S. Linsey, D. S. Lumia, and M. A. Maupin. 2004. Estimated Use of Water in the United States in 2000. USGS Circular 1268. Washington, D.C.: USGS. Kennedy, D. N. 1994. California Water Plan Update. Vol. 1. Sacramento, Calif.: State of California Dept. of Water Resources. Kincaid, D. C. 1984. Cablegation: V. Dimensionless design relationships. Trans. ASAE 27(3): 769-722. Merriam, J. L., and A. J. Clemmens. 1985. Time rated infiltrated depth families. In Development and Management Aspects of Irrigation and Drainage, Spec. Conf. Proc., 67-74. Reston, Va.: American Soc. Civil Engineers, Irrig. and Drain. Div. Philip, J. R. 1957. The theory of infiltration: 4. Sorptivity and algebraic infiltration equations. Soil Science 84: 257-264. Solomon, K. H., and B. Davidoff. 1997. On the relationship between unit and subunit irrigation performance. Draft copy. Strelkoff, T. 1977. Algebraic computation of flow in border irrigation. J. Irrig. Drain. Div., ASCE 103(1R3): 357-377. Strelkoff, T. 1990. SRFR: A Computer Program for Simulating Flow in Surface Irrigation Furrows-Basins-Borders. WCL Report #17. Phoenix, Ariz.: U.S. Water Conservation Laboratory, USDA/ARS. Strelkoff, T. S., A. J. Clemmens, B. V. Schmidt, and E. J. Slosky. 1996. Border: A Design and Management Aid for Sloping Border Irrigation Systems. Version 1.0. WCL Report #21. Phoenix, Ariz.: U.S. Water Conservation Laboratory, USDA-ARS. Stringham, G. E., and S. N. Hamad. 1975. Design of irrigation runoff recovery systems. J. Irrig. Drain. Div., ASCE 101(IR3): 209-219. Trout, T. J., and D. C. Kincaid. 1994. Float-activated variable-speed irrigation tailwater pump. ASAE Paper No. 942124. St. Joseph, Mich.: ASAE. USDA. 1974. Chapt. 4, Sect. 15: Border Irrigation. In National Engineering Handbook. Washington, D.C.: Soil Conserv. Serv., USDA. USDA. 1984. Chapt. 5, Sect. 15: Furrow Irrigation. In National Engineering Handbook. Washington, D.C.: Soil Conserv. Serv., USDA. Walker, W. R. 1989. Guidelines for designing and evaluating surface irrigation systems. FAO Irrigation and Drainage Paper 45. Rome, Italy: Food and Agriculture Organization of the United Nations. Walker, W. R., and G. V. Skogerboe. 1987. Surface Irrigation: Theory and Practice. Englewood Cliffs, N.J.: Prentice-Hall, Inc.

CHAPTER 15

HYDRAULICS OF SPRINKLER AND MICROIRRIGATION SYSTEMS Derrel L. Martin (University of Nebraska, Lincoln, Nebraska) Dale F. Heermann (USDA-ARS, Fort Collins, Colorado) Mark Madison (CH2M Hill, Portland, Oregon) Abstract. This chapter presents and discusses the hydraulic characteristics of closed-pipe irrigation systems needed for the design of sprinkler and microirrigation systems. Several of the common equations for calculating friction head loss in closedpipe systems are presented. The accuracy and limitations of the methods are stressed. A discussion of irrigation system valves and their proper use is provided for inclusion in the design of closed-pipe systems that are frequently drained and refilled during operation. Water distribution in the soil to enhance uniformity is also mentioned. Keywords. Air and vacuum relief valves, Check valves, Friction resistance, Head loss, Irrigation control valves, Pipe friction loss, Lateral design criteria, Surge valves.

15.1 INTRODUCTION The fluid dynamics of sprinkler and microirrigation systems are complex. Water moves dynamically from the water source through the pump into the pipe network. Water often goes through a series of screens and filters depending on the source and type of irrigation system. From the pipe network, water is supplied under pressure to sprinkler systems and through a sprinkler nozzle into the air at a high velocity where it breaks up into droplets and falls to the soil or crop surface and is redistributed. Sprinklers are mounted at various heights with many types of nozzles that experience diverse pressure and wind speeds. The diversity of operating conditions produces a large range of drop sizes. When the network provides water to a microirrigation system the

Design and Operation of Farm Irrigation Systems

533

water is discharged through emitters that are located on the surface or buried in the soil. In this chapter, we discuss the flow within the closed-pipe network to either the sprinkler or microirrigation emitter. The importance of the redistribution on the surface and in the soil is discussed. We emphasize general design considerations with limited focus on operation. Chapters 16 and 17 present detailed procedures for designing and operating sprinkler and microirrigation systems.

15.2 HYDRAULICS OF PIPE SYSTEMS Pressurized irrigation systems consist of main lines which convey the entire flow from the water source to lateral pipelines positioned across the field. Some of the flow in the main line branches into laterals which direct the discharge to emitters which wet the immediate soil area or sprinkler heads, and nozzles, which direct the jet stream into the air. The main line energy losses due to surface resistance can be determined with uniform flow methods, whereas laterals and outlets with branching flow require nonuniform flow analysis due to surface resistance and form losses. 15.2.1 Energy Losses in Pipes and Couplers The Darcy-Weisbach formula, as discussed in Chapter 11, is recommended for calculating energy losses in pipes. The Darcy-Weisbach equation for circular pipes flowing full is given by L V 2 ⎛⎜ 8 f ⎞⎟ LQ 2 (15.1) = hf = f D 2 g ⎜⎝ π 2 g ⎟⎠ D5 where hf = the head loss, m f = the Darcy-Weisbach resistance coefficient L = the pipe length, m D = the inside diameter of the pipe, m V = the velocity, m/s Q = the flow in the pipe, m3/s g = the gravitational acceleration, 9.81 m/s2. The resistance coefficient (f), also called the friction factor, is a function of the relative roughness (e/D) and Reynolds number (Re). The absolute roughness height (e) is a constant for a given pipe material. The resistance coefficient is dimensionally consistent and can be represented for most commercial materials by the semi-empirical equation as ⎛e 1 9.35 ⎞⎟ = 1.14 − 2 log⎜ + (15.2) ⎜D R f ⎟ f e ⎝ ⎠ where e = the absolute roughness of the inside of the pipe, mm Re = the Reynolds number (VD/v) v = the kinematic viscosity, m2/s. The average flow velocity is equal to the discharge divided by the cross-sectional area so the Reynolds number for circular pipes flowing full can be computed from Re = 1.273×106 Q/D

(15.3)

534

Chapter 15 Hydraulics of Sprinkler and Microirrigation Systems

where Q is the discharge in L/s, D is the inside diameter of the pipe in mm, and the kinematic viscosity is 1.0×10-6 m2/s, which represents a water temperature of approximately 20°C. The Reynolds number is generally between 300,000 and 700,000 at the inlet for center pivot or lateral move systems designed with representative system capacities (Figure 15.1). The Reynolds number decreases along a lateral as water exits through nozzles or emitters. The flow in a center pivot lateral at a point 90% of the way along the lateral should be about 20% of the inflow to the system. Thus, the Reynolds number exceeds 50,000 for most of the lateral for pivots with typical capacities. The Reynolds number for many microirrigation and other types of sprinkler systems are usually less than for center pivots. Large diameter pipes used for main lines may have Reynolds numbers larger than shown in Figure 15.1. The resistance coefficient is dimensionless and consistent with customary (footpound) or metric units; however, Equation 15.2 must be solved iteratively. Table 15.1 presents equivalent sand-grain size roughness values for determining the friction factor for various pipe materials. General resistance diagrams for determining the friction factor are available in most fluid mechanics textbooks. Values for the resistance coefficient typical of irrigation systems are presented in Figure 15.2. The resistance coefficient varies along sprinkler or microirrigation laterals. The relative roughness depends on the smoothness and inside diameter of the pipe; therefore, the relative roughness is constant for a given pipe. However, the flow in a lateral decreases along the pipeline. For example, consider a center pivot with a galvanizedsteel lateral that has an inside diameter of 163 mm. The roughness (e) is about 0.11 mm from Table 15.1 which gives a relative roughness (e/D) of 0.00068. If the pivot was designed for an inflow of 51 L/s the Reynolds number at the inlet would be about 400,000. The resistance coefficient at the lateral inlet would be about 0.019 (Figure 15.2). At a point 90% of the way along the lateral the flow should be about 20% of the inflow to the lateral giving a Reynolds number of approximately 80,000 which produces a resistance coefficient of 0.022. This represents a change of approximately Table 15.1. Values of roughness e (equivalent sand grain size) for pipe materials for determining the Darcy-Weisbach friction factor and roughness coefficient C for the Hazen-Williams equation (adapted from Albertson et al., 1960, and Meadows and Walski, 1998). Material e (mm) C Aluminum with a coupler spacing of 9.2 m Asbestos cement Commercial steel or wrought iron Concrete Corrugated metal pipe Galvanized iron Galvanized steel for center pivot and lateral move systems New wrought iron or steel schedule 40 pipe Plastic, polyethylene and PVC Riveted steel

0.10 - 0.30 0.0015 – 0.0025 0.03 - 0.09 0.20 - 3.00 30.0 - 60.0 0.10 - 0.25 0.10 - 0.12 0.04 – 0.050 0.0015 - 0.0025 0.90 - 9.00

115 - 125 135 - 145 135 - 145 100 - 140 55 - 60 120 - 130 130 - 140 140 - 150 140 - 150 100 - 110

Design and Operation of Farm Irrigation Systems

535

700,000 168

650,000

203

254

600,000

REYNOLDS NUMBER

550,000

PIPE DIAMETER, mm

152

500,000 450,000 400,000 350,000 300,000 250,000 200,000 150,000 100,000 50,000 0 0

10

20

30

40

50

60

70

80

90

100 110 120

130 140

DISCHARGE, L/s

Figure 15.1. Reynolds number for pipes typical of center pivot and lateral move systems.

Figure 15.2. Resistance coefficient for flows used in irrigation, based on Equation 15.2 using 20°C water.

536

Chapter 15 Hydraulics of Sprinkler and Microirrigation Systems

16% in the resistance coefficient along the lateral. Figure 15.2 also illustrates that the resistance coefficient varies with the diameter of pipe. If 51 L/s flowed in a 254-mm pivot lateral as in the above example the relative roughness would be about 0.00043 and the resistance coefficient would drop to 0.018 for a Reynolds number of 260,000. The irrigation industry has generally accepted empirical formulae to determine friction losses. Commonly used empirical equations include the Scobey method and the Hazen-Williams equation. The Hazen-Williams and Scobey equations are given by hf = K

LQ d1 Dd2

(15.4)

where hf = the friction loss in a pipe that conveys the flow throughout the length, m L = the length of pipe, m Q = the flow in L/s D = the inside diameter of the pipe, mm. The value of K varies for the two equations. For the Hazen-Williams equation K = 1.21 × 1010 C-1.852 where C is the pipe roughness coefficient. The exponents are given by d1 = 1.852 and d2 = 4.87 for the Hazen-Williams equation. Typical values of C are listed in Table 15.1 for various types of pipe. For the Scobey method K = 4.10 × 106 Ks where Ks is Scobey’s coefficient of retardation and the exponent values are d1 = 1.9 and d2 = 4.9. The empirical equations (Equation 15.4) are similar in form to the Darcy-Weisbach equation when expressed in terms of flow. The exponents for the flow and pipe diameter for the Hazen-Williams and Scobey equations were determined from experimental evaluation and are smaller than for the Darcy-Weisbach method. However, the primary difference is that the roughness coefficient for the Hazen-Williams equation and the Scobey retardation coefficient are often taken as constants for a given type of pipe. In reality, the pipe size, type of piping material, coupler design, alignment of pipe, and distance between couplers affect flow resistance. Normally, as pipe size increases, resistance decreases (i.e., C should increase for the Hazen-Williams equation). As the number of couplers decreases, C should increase. Piping materials with smoother inside walls will have less resistance and higher C values for the Hazen-Williams equation. The resistance coefficient for the Darcy-Weisbach equation depends on the pipe roughness, pipe diameter, flow rate and the temperature of the water. Thus, the DarcyWeisbach equation has broader applications than the Hazen-Williams or Scobey equations. Other friction loss equations have been proposed that have a similar form. Keller and Bliesner (1990) combined the Reynolds number, Darcy-Weisbach, and Blausis equations to develop functions similar to the Hazen-Williams but with exponents for flow of 1.75 for plastic pipe with diameters smaller than 125 mm and 1.83 for pipes larger than 125 mm. The diameter exponent was 4.75 and 4.83 for pipes smaller than 125 mm and larger than 125 mm, respectively. The Hazen-Williams and other empirical equations should be used with caution. Liou (1998) analyzed several datasets to establish limitations and recommended uses of the Hazen-Williams equation. He demonstrated significant variations of C with Reynolds number, relative roughness, and, separately, diameter or absolute roughness of the pipe. The error resulting from applying the Hazen-Williams equation outside its

Design and Operation of Farm Irrigation Systems

537

calibration range can be as large as 20%. This error is similar to the variation in the resistance coefficient along pivot laterals. Variations in pipe diameter do not introduce errors for the Darcy-Weisbach equation if the correct roughness height is used. Tables and nomographs of energy loss with the Darcy-Weisbach equation are needed to encourage greater industry adoption. The energy loss for pipe with couplers can be determined by increasing the friction factor by 11% for aluminum irrigation pipe with couplers spaced 9 m apart. Decreasing the spacing to 6 m would increase the energy loss by 17%, while increasing the coupler spacing to 12 m should decrease the energy loss by 8%. Lyle and Wimberly (1962) found these losses were the upper limits in their study of various styles of couplers. Either manufacturers’ data or individual tests should be obtained for a given coupler to determine energy losses for pipe and couplers. Kincaid and Heermann (1970) used numerical techniques to calculate the pressure distribution on a center pivot system with the Darcy-Weisbach equation. Their results showed that the head loss in the sprinkler riser can be approximated by hr =

⎛ 9.2q s V2 exp⎜⎜ 2g ⎝ Q

⎞ ⎟⎟ ⎠

(15.5)

where hr = the head loss in the riser, m qs = the sprinkler discharge, L/s Q = the upstream lateral discharge, L/s. The branching loss due to the change in direction and flow into the riser was generally less than 0.6 m (6 kPa) and is often neglected in the design of high pressure systems (>400 kPa). Pressures along pipelines are computed in an iterative fashion. The outlet pressure head Hi for section i along the pipeline is given by Hi = Hi-1 – hfi + (Ei-1 – Ei)

(15.6)

where Hi-1 = the pressure head at the inlet to the section hfi = the head loss due to friction in the section Ei = the elevation at the distal end of the section Ei-1 = the elevation at the inlet to the section. 15.2.2 Lateral Pressure Distribution Many systems are designed for nearly constant sprinkler discharge along the lateral at a specified average nozzle pressure. The piping is then specified in terms of maximum lateral length and multiple sizes to maintain the pressure variation along the lateral within specified limits. The laterals are usually designed first, then the main line is designed, and the inlet pressure is set high enough to maintain the minimum desired sprinkler pressure. In designing portable laterals, normally only one or two tubing sizes are used per lateral. However, for permanent laterals several sizes of pipe along the lateral may be the most economical design. In sizing main or supply and lateral lines, the designer should not only consider uniformity of application, but also pumping costs and elevation differences between the water source and the fields. When water is removed at intervals along the lateral, the friction loss will be less than if the flow was constant for the entire length of the pipeline. To accurately compute friction loss in the lateral, start at the last outlet on the line and work back to the

538

Chapter 15 Hydraulics of Sprinkler and Microirrigation Systems

supply line, computing the friction loss between outlets. This tedious process has been simplified by a procedure developed by Christiansen (1942). He developed an adjustment factor (F) which is the ratio of the friction loss in a lateral with multiple outlets having equal spacings and discharges, to that calculated from the general formula that assumes all of the water is carried to the distal end of the pipeline. The F factor assumes that the first sprinkler is located one full sprinkler spacing from the main line. Jensen and Fratini (1957) developed a modified factor F′ when the first sprinkler is located one-half the sprinkler spacing from the supply line. Equations have been developed to predict the values for F and F′. The equation for F is based on Christiansen’s formulation while the equation for F′ is based on the approximation from Keller and Bliesner (1990) to the function developed by Jensen and Fratini: F=

1 1 m −1 + + m + 1 2N 6N 2

and F ′ =

2N 2N − 1

⎛ 1 m − 1 ⎞⎟ ⎜ + ⎜ m + 1 6N 2 ⎟ ⎝ ⎠

(15.7)

where N is the number of sprinkler outlets on the lateral and m is the exponent in the friction loss equation. The exponent m = 1.852 for the Hazen-Williams equation, 2.0 for Darcy-Weisbach, 1.9 for the Scobey function, and either 1.75 or 1.83 for the functions developed by Keller and Bliesner (1990). Values of F and F′ are listed in Table 15.2. Table 15.2. Factors to use in computing head loss in a lateral with equal sprinkler discharge and spacing for equations commonly used to compute friction loss (based on Equation 15.7). m coefficient Number of Sprinklers

Keller and Bliesner

Hazen-Williams

1.75

Scobey

1.852

Darcy-Weisbach

1.9

2

F

F′

F

F′

F

F′

F

F′

2

0.65

0.53

0.64

0.52

0.63

0.51

0.63

0.50

3

0.55

0.46

0.53

0.44

0.53

0.43

0.52

0.42

4

0.50

0.43

0.49

0.41

0.48

0.41

0.47

0.39

5

0.47

0.41

0.46

0.40

0.45

0.39

0.44

0.38

6

0.45

0.40

0.44

0.39

0.43

0.38

0.42

0.37

7

0.44

0.39

0.43

0.38

0.42

0.37

0.41

0.36

8

0.43

0.39

0.42

0.38

0.41

0.37

0.40

0.36

9

0.42

0.39

0.41

0.37

0.40

0.37

0.39

0.36

10

0.42

0.38

0.40

0.37

0.40

0.36

0.39

0.35

12

0.41

0.38

0.39

0.37

0.39

0.36

0.38

0.35

14

0.40

0.38

0.39

0.36

0.38

0.36

0.37

0.35

16

0.40

0.38

0.38

0.36

0.38

0.36

0.37

0.34

18

0.39

0.37

0.38

0.36

0.37

0.36

0.36

0.34

20

0.39

0.37

0.38

0.36

0.37

0.35

0.36

0.34

25

0.38

0.37

0.37

0.36

0.37

0.35

0.35

0.34

30

0.38

0.37

0.37

0.36

0.36

0.35

0.35

0.34

35

0.38

0.37

0.37

0.36

0.36

0.35

0.35

0.34

40

0.38

0.37

0.36

0.36

0.36

0.35

0.35

0.34

50

0.37

0.37

0.36

0.35

0.35

0.35

0.34

0.34

Design and Operation of Farm Irrigation Systems

539

The relationship between pressure and discharge for sprinklers shows that the discharge varies with the square root of the pressure. Since pressure varies along the lateral due to friction and elevation differences, the sprinkler discharge also varies. The pressure ratio on a lateral, P/Pd, is the ratio of the pressure at any point on the lateral to the pressure at the end sprinkler (with constant elevation). The discharge ratio is equal to the square root of the pressure ratio: qs P (15.8) = qd Pd where qs is the discharge of any sprinkler whose pressure is P and qd is the discharge of the last sprinkler on the lateral with pressure of Pd. Thus, with 20% variation in pressure along a lateral, the discharge varies by about 10%. To obtain high application efficiencies and stay within economical pipe sizes, the variation in lateral pressure must be held to a practical minimum. The variation should be determined by the system designer with the approval of the purchaser. Generally the variation in pressure should not exceed 20% of the average lateral pressure. Ultimately the allowable pressure loss involves balancing capital cost of the pipe against pumping costs due to friction. Keller and Bliesner (1990) provide methods to determine the most economical pipe size. The variation of pressure along a lateral can be approximated by assuming a uniform discharge per unit length along the lateral. The total flow in the lateral (Qx) at any point x along the lateral is given by Qx = QL (1 – x/L) = N qs (1 – x/L)

(15.9)

where QL is the total inflow to the lateral. If it is assumed that the slope of the lateral is uniform, the head loss per unit length at a point along the lateral for the HazenWilliams equation is given by dH = − KQ1x.852 D − 4.87 − S f (15.10) dx where Sf is the uniform slope of the lateral from the inlet. A positive slope means the lateral is running uphill. Combining these equations and integrating from the distal end of the lateral to a point along the lateral gives the pressure head H(x) along a uniformly sloping lateral: ⎡ K L H ( x) = H d + ⎢ ( Nq s )1.852 4 . 2 852 ⎢⎣ D .87

x⎞ ⎛ ⎜1 − ⎟ L ⎝ ⎠

2.852

⎤ + S f ( L − x)⎥ ⎥⎦

(15.11)

When this equation is applied to the inlet of the lateral (i.e., x = 0) the pressure loss in the lateral is 35% (1/2.852) of the loss in an enclosed pipeline where all of the flow is delivered to the distal end of the pipeline. This corresponds to an F value of 0.35 which would equate to an infinite number of sprinklers on a lateral. The values for the appropriate F or F′ value can be substituted for 1/2.852 in Equation 15.11. Chu and Moe (1972) derived an analytical approximation for calculating head loss in a center pivot lateral. They assumed an infinite number of tiny sprinklers evenly distributed along the lateral with the discharge proportional to the radius for a uniform

540

Chapter 15 Hydraulics of Sprinkler and Microirrigation Systems

irrigation. The total head loss in the lateral is 54% of the head loss for the lateral operating as a supply line with the inflow discharge. This supply line head loss can be determined from an energy-loss table or one of the previously discussed equations. The distribution of the pressure loss was approximated by 3 5 ⎫ ⎧ ⎪ ⎛ 15 ⎞ ⎡ x 2 ⎛ x ⎞ 1 ⎛ x ⎞ ⎤ ⎪ H x = H L + (H o − H L ) ⎨1 − ⎜ ⎟ ⎢ − ⎜ ⎟ + ⎜ ⎟ ⎥ ⎬ ⎪⎩ ⎝ 8 ⎠ ⎢⎣ L 3 ⎝ L ⎠ 5 ⎝ L ⎠ ⎥⎦ ⎪⎭

(15.12)

where Hx = the pressure head at a radial distance of x from the pivot, m Ho = the pressure head at the pivot, m HL = the pressure head at the distal end of the lateral at radius L, m x = the radial distance from pivot, m. 15.2.3 Average Pressure The average pressure head along the lateral can be computed by integrating the pressure distribution for the lateral and dividing by the length. This results in Ha = Hd +

Fh f 3.852

+

Sf L 2

(15.13)

where hf is the friction loss for an equal sized pipe with no outlets and the adjustment factor can be equal to F or F′ depending on the spacing of the first sprinkler from the main line. The average pressure depends on the sprinkler discharge, the number of sprinklers on the lateral, the diameter and type of pipe, and the slope along the lateral. The layout of laterals determines the length and number of sprinklers on the lateral. The discharge per sprinkler is determined by crop water requirements, soil characteristics, sprinkler spacings, and management constraints. Once this information is known the average pressure along the lateral can be determined. The average pressure and the required discharge are used to determine the sizes of nozzles for the sprinkler. The selected nozzle and pressure combination must provide adequate coverage to maintain uniformity. The average pressure may need to be raised to satisfy both conditions. 15.2.4 Pressure Variation The pressure head at the inlet (Ho) and distal (Hd) ends of the lateral can be computed using the pressure distribution equation: Ho = Ha + 0.74 F hf + 0.5 Sf L and Hd = Ha – 0.26 F hf – 0.5Sf L

(15.14)

The pressure variation for an example system is shown in Figure 15.3. The average pressure generally occurs in the second quarter of the pipeline. The average pressure point moves down the pipeline as the slope increases. The guideline for maintaining the pressure variation to less than 20% of the average pressure requires that the maximum and minimum pressure be identified. The maximum pressure occurs at the inlet to the lateral when the elevation increases along the lateral (i.e., lateral runs uphill) or when the downward slope is less than the friction loss along the lateral (i.e., –Sf L < F hf). Conversely, the maximum occurs at the distal end of the lateral when the lateral slopes downhill at an angle that exceeds the friction

Design and Operation of Farm Irrigation Systems

541

loss along the lateral (i.e., –Sf L > F hf). The maximum pressure can be computed using either Equation 15.14 or 15.15. The minimum pressure can occur anywhere along the lateral depending on the friction loss and slope (Figure 15.3). The minimum pressure and its location can be determined from Equation 15.11. The uniformity criteria require that the difference between the maximum and minimum head be less than 20% of the average pressure head. The minimum acceptable average pressure is then five times the pressure variation along the lateral. Using this criterion and the pressure distribution, the minimum acceptable average pressure is given by 1. For Sf ≥ 0: Hmax = Ho and Hmin = Hd, so Pmin = 5(F hf + Sf L) 9.8

2. Sf < 0: xmin

1 ⎤ ⎡ ⎢ ⎛ S f L ⎞ 1.852 ⎥ ⎟ = L ⎢1 − ⎜ − ⎥ so ⎥ ⎢ ⎜⎝ h f ⎟⎠ ⎥⎦ ⎢⎣ 1

⎡⎛ S f L ⎞⎤ ⎛ S L ⎞ 1.852 ⎟ ⎥ + S f L⎜ − f ⎟ H min = H d + Fh f ⎢⎜ − ⎟ ⎜ hf ⎟ ⎜ ⎢⎣⎝ h f ⎠⎥⎦ ⎠ ⎝

(15.15)

For (– F hf /L) ≤ Sf ≤ 0: Hmax = Ho and xmax = 0 ⎧ ⎪⎪ Pmin = 5⎨ Fh f ⎪ ⎩⎪

2.852 ⎤ 1 ⎤⎫ ⎡ ⎡ ⎢ ⎛ S f L ⎞ 1.852 ⎥ ⎪⎪ ⎢ ⎛ S f L ⎞ 1.852 ⎥ ⎟ ⎟ ⎥ ⎬9.8 ⎢1 − ⎜⎜ − ⎥ + S f L ⎢1 − ⎜⎜ − ⎢ ⎝ h f ⎟⎠ ⎥⎪ ⎢ ⎝ h f ⎟⎠ ⎥ ⎣⎢ ⎦⎥ ⎭⎪ ⎣⎢ ⎦⎥

and for Sf ≤ (– F hf /L): Hmax = Hd and xmax = L 2.852 1 ⎤⎫ ⎧ ⎡ ⎪⎪ ⎛ S f L ⎞ 1.852 ⎢ ⎛ S f L ⎞1.852 ⎥ ⎪⎪ ⎟ ⎟ − S f L ⎢1 − ⎜ − Pmin = 5⎨− Fh f ⎜ − ⎥ ⎬9.8 ⎜ hf ⎟ ⎜ hf ⎟ ⎢ ⎥⎪ ⎪ ⎠ ⎠ ⎝ ⎝ ⎣⎢ ⎦⎥ ⎭⎪ ⎩⎪

where the Pmin (in kPa) is the minimum value that can be used for the average pressure in selecting nozzle sizes. The above equations can be used to design constant-sized laterals. Guidelines can be developed for specific conditions such as shown in Figure 15.4. For this example the number of sprinklers on the lateral and the discharge per sprinkler are known from the system layout and preliminary calculations. The chart can be used to quickly estimate the average pressure to use in selecting nozzles for the sprinkler, or the maximum number of sprinklers for the lateral. A series of charts can be made for varying spacings, pipe sizes, and slopes. Slide rules and other tools have been developed and are widely used for quick estimates of friction loss.

542

Chapter 15 Hydraulics of Sprinkler and Microirrigation Systems 900 LATERAL LENGTH LATERAL INFLOW ALUMINUM PIPE, C PIPE I.D END PRESSURE

UPSLOPE FROM INLET, m/m

800 0.06

400 m 16 L/s 120 99 mm 350 kPA

PRESSURE, kPa

700 0.03 600 0.0

AVERAGE PRESSURE, kPa

524

500 465 -0.03

400

406 347

-0.06

300

288

200 0

50

100

150

200

250

300

350

400

DISTANCE ALONG LATERAL, m

Figure 15.3. Example pressure distribution along a sprinkler lateral versus slope and the location along the lateral where the average pressure occurs. 700 650

SPRINKLER DISCHARGE, L/s

MINIMUM AVERAGE PRESSURE, kPa

600

1.2 1.0

550

0.8

0.6

0.5

0.4

500 450 400 350 300 250 200 150

Sprinkler Spacing Pipe Diameter, C value

100 50

9.1 m 101.6 mm 120

0 0

10

20

30

40

50

NUMBER OF SPRINKLERS

Figure 15.4. Design chart based on the minimum average pressure for a 101.6 mm lateral with sprinklers spaced at 9.1 m.

60

Design and Operation of Farm Irrigation Systems

543

When multiple pipe sizes are used along the lateral, or topographic variations are complex, an iterative section-by-section calculation scheme is required. With the widespread availability of personal computers and spreadsheet software an iterative technique is proficient. Friction losses in valves and fittings such as sprinkler risers and lateral valves should be included in pressure calculations. Losses in pipe couplers are commonly included in the factor for friction losses (see Chapter 11 for details on computing losses in fittings and valves). Valve flow coefficients (flow in L/s for a loss of 1 kPa) are available from manufacturers. The actual head loss through a valve is usually proportional to the square of the flow rate. 15.2.5 System-Head Curve The design of a lateral only produces one flow rate and pressure head at the inlet. For pump specification a head-discharge curve should be developed for the sprinkler system. The head-discharge relationship can be developed from the pressure distribution function and the sprinkler discharge equation: 1 ⎛⎜ q s Ho = 9.8 ⎜⎝ C d Dna1

1 / b1

⎞ ⎟ ⎟ ⎠

+ 0.74 FK ( Nq s )1.862

L D

4.87

+

Sf L 2

(15.16)

where Cd is the discharge coefficient for the sprinkler used on the lateral, Dn is the diameter of the nozzle, and parameters a1 and b1 are the coefficients for the discharge equation: q s = C d Dna1 P

b1

(15.17)

Coefficients a1 and b1 frequently have the value of 2.0 and 0.5, respectively. A system-head curve can also be developed by treating the lateral as a single circular nozzle. Using this relationship the inlet head will increase approximately in proportion to the square of the flow rate (Ho = k QL2). The value of k can be computed from the design values of Ho and QL and can be used for other flows. If the pump will operate at various flow rates, for example with variable numbers of laterals functioning, various head-flow combinations may be specified. The pump pressure can be controlled if necessary, although it is desirable to operate the pump close to full capacity. Multiple pumps operating in parallel would be a good choice for systems in which the flow rate varies greatly but pressure head is nearly constant. A head-booster pump can be used where a portion of the system is at a high elevation relative to the pump. It is desirable to apply water as uniformly as possible, whether the entire system or only a segment of the system is operating. This requires that pressure loss due to friction in the main line be minimized or that flow control devices be used. Elevation differences in a field must also be included in the calculations. Normally, flow velocities should not exceed 3 m/s. For permanent systems in which polyvinyl chloride (PVC) plastic pipe and asbestos cement (AC) pipe are used for supply lines, velocities should not exceed 2.25 m/s, and most manufacturers caution against using velocities in excess of 1.6 m/s. When more than one lateral on a system is operated at one time (especially adjacent to each other), the supply line becomes multi-outlet. This means more water is flowing through the portion nearer the pump and friction loss must be computed in each segment.

544

Chapter 15 Hydraulics of Sprinkler and Microirrigation Systems

15.2.6 Pressure Regulation/Flow Control The operating pressure of a sprinkler may vary from the design value. Variations in pressure may result from changes in elevation as the sprinkler lateral is moved across the field. Variations can also occur when the numbers of laterals or sprinklers per lateral vary during the operation. When a lateral is shutoff for repositioning, the pressure increases for the laterals that remain in operation. If the variation is too severe, pressure regulators or flow control nozzles may be required to provide the desired uniformity. Pressure regulators are devices that maintain a nearly constant downstream pressure for varying inlet pressures as illustrated in Figure 15.5. The advantage of pressure regulators is that the discharge from the sprinkler is nearly constant over a wide range of inlet pressures. An indicator that regulators are needed is based on the concept that the pressure at a sprinkler should not vary more than 20%. Guidelines such as shown in Figure 15.6 can be developed from these criteria. Pressure regulators are not without some limitations. First, there is a loss of pressure across regulators. The pressure loss occurs within the regulator even when the inlet pressure is below the design pressure. Commonly designers use a loss of 20 to 40 kPa across the regulator. For proper design the pressure in the lateral should be adequate to satisfy the required pressure for the sprinkler, pressure losses in fittings and riser used to connect the sprinkler to the lateral and losses through the regulator. Pressure regulators increase the cost of the irrigation systems. The analysis by von Bernuth (1983) compared the economic benefit of the increase in uniformity for using pressure regulators to the amortized investment. This type of comparison is very site specific and results cannot be generalized; however, the analysis provides a procedure for localized evaluation.

OUTLET PRESSURE, kPa

300

250

FLOW, L/s

200

0.05 0.6 1.0

150

FLOW, L/s

100

0.05 0.6 1.0

50

0 0

100

200

300

400

500

600

700

800

INLET PRESSURE, kPa

Figure 15.5. Regulator control of outlet pressure with regulators set to maintain 70 and 210 kPa.

Design and Operation of Farm Irrigation Systems

545

8.0 7.0

Pressure Regulators Recommended Elevation Change, m

6.0 5.0 4.0 3.0 2.0

Pressure Regulators Not Recommended 1.0 0.0 50

100

150

200

250

300

350

Design Pressure, kPa

Figure 15.6. Elevation change equivalent to 20% variation in the design pressure.

Proper operation of pressure regulators is difficult to monitor. As von Bernuth and Baird (1990) showed, some regulators display a hystersis when the inlet pressure gradually changes. Thus, the regulator does not restrict flow to the extent desired for high pressure situations and excessively restricts flow when the pressure drops. Pressure regulators operate by restricting the opening for water flow when the pressure increases. This requires relatively small openings that may plug if foreign material is contained in the irrigation water. This makes utilization of the devices inconsistent for some sources of water. If the regulators are plugged, have hystersis problems or simply stick in one position it is difficult to observe the inappropriate operation. The sprinkler may appear to be working yet the pressure could be inappropriate. Regulators are occasionally used when they are not needed. Regulators are habitually used on many center pivots that use low pressure sprinkler devices. Some designers use regulators as insurance so that the system will operate at the appropriate pressure even though the terrain would not merit pressure regulation. Obviously this drives up the investment and operating costs and should be avoided. If pressure regulators are used the minimum pressure in the system should be determined. This usually occurs at a point near the end of the system where the elevation is the greatest. The minimum pressure may vary during the growing seasons if the supply capacity of wells or other sources of water declines. The system should be designed for this pressure and the appropriate regulators should be used for this condition. The uniformity of the system can be reduced substantially if the inlet pressure drops below the design pressure for the system. Special nozzle designs are available that compensate for pressure variation in the field. These nozzles, called flow control nozzles, are constructed so that the diameter

546

Chapter 15 Hydraulics of Sprinkler and Microirrigation Systems 0.6

NORMAL 3.97 mm NOZZLE

DISCHARGE, L/s

0.5

0.4 FLOW CONTROL NOZZLE 0.3

0.2

0.1

0.0 0

200

400

600

800

1000

NOZZLE PRESSURE, kPa

Figure 15.7. Illustration of flow control nozzle compared to a regular nozzle.

of the nozzle varies with nozzle pressure. When the pressure rises the nozzle diameter decreases and a relatively uniform flow rate is maintained as illustrated in Figure 15.7. Flow control nozzles can be effective in providing uniform discharge and are cheaper to purchase and operate than pressure regulators. Flow control nozzles, however, will not compensate for as large of variation in pressure as regulators. As with pressure regulators, it is difficult to determine if the nozzles are operating properly and flow measurements are generally required to monitor operation. 15.2.7 Energy Losses in Regulators and Emitters The industry recommends that the inlet pressure exceed the nominal regulated pressure by 30 to 35 kPa. This in general will allow for the pressure loss through the regulator at most design flow rates. As illustrated in Fig. 15.6, the regulated outlet pressure is dependent on the flow rate. The head loss through the regulator increases as the flow increases. Emitters also have head losses that are variable depending on pressure and thus flow rate. It is recommended to search the literature and obtain manufacturers’ recommendations on head loss for specific regulators and emitters. 15.2.8 Energy Losses in Filtration Systems Filtration systems included screens and media type filters. The inlets to pumping plants often use a rotating screen filter that has minimal head loss with the relatively large screen area sized to the design flow. Filters are also used prior to the inflow into the lateral systems particularly for microirrigation systems. In some cases filters are used to clean the water to prevent the plugging of application devices for sprinkler systems. The manufacturers can supply the required head for proper operation. Many of the filter systems are designed to provide self cleaning when the pressure drop across the filter exceeds approximately 50 kPa. The head loss depends on the flow rate and manufacturers’ curves illustrate losses of 1 to 100 kPa for a range of discharges.

Design and Operation of Farm Irrigation Systems

547

15.3 IRRIGATION SYSTEM VALVES The following sections provide considerations for special valves when installing pressurized pipelines. Specific technical data and installation recommendations should be obtained from the equipment manufacturers. Often these data can be obtained from internet sites. 15.3.1 Surge Control and Check Valves Irrigation systems require valves for many functions. Some common valves and their functions and applications are described below. Undesirable high and low pressures can occur in a pipeline following power failure. Carefully timed and controlled closure settings on the valves at each pump discharge will help reduce the high pressure spikes that occur when the valves fully close. Low pressures are problematic if the water begins to vaporize. Bubbles form in the liquid if the pressure in the pipeline drops below the vapor pressure of the water. If the bubbles coalesce they may form a pocket of vapor in the pipeline. This is called column separation and is a condition that must be guarded against during flow. Column separation that occurs at high flow rates and high head conditions cannot be corrected by adjusting valve closure times. Vacuum relief valves along portions of the pipeline will be required to minimize column separation. 15.3.1.1 Surge chambers. A surge chamber (Figure 15.8) is the most reliable water hammer control device. Factors used to determine the required surge chamber size— which can be large—are pipeline length, maximum flow capacity, acoustic wave speed in the line, pipe profile, and flow distribution. The required size can only be determined with computer analysis, but an approximate size may be estimated by

VT =

QL 100c

(15.18)

where VT = the total volume, m3 Q = the discharge, L/s L = the length of the pipeline, m c = the velocity of the pressure wave in the pipeline, m/s. The velocity of the pressure wave as a function of pipe material and the diameter ratio of the pipe are shown in Figure 15.9 for materials commonly used for irrigation. Equations to compute the velocity of the pressure wave for other pipe materials or temperatures are provided in fluid mechanics textbooks. The temperature of the water has little effect on the wave velocity. The diameter ratio used to determine the velocity of the wave is the ratio of the inside diameter to the wall thickness of the pipe. Surge chambers do not work properly with in-line control valves or surge anticipator valves. To prevent forcing flow back through the pumps, it is appropriate to use a check valve in conjunction with surge chambers on the discharge side of pumps. Different outlet and inlet line sizes may significantly reduce the required surge chamber size. Surge chambers act as energy sources following power failure and as shock absorbers during upsurge events such as a pump start or reverse flow following a pump shutdown. Characteristically, surge chambers remove sharp pressure spikes and create smooth, controlled pressure oscillations until friction damps out transient pressure waves.

548

Chapter 15 Hydraulics of Sprinkler and Microirrigation Systems

Figure 15.8. Picture of surge tank assembly with sketches of an air vent/vacuum relief valve and a pressure release valve.

Figure 15.9. Velocity of pressure wave based on the diameter ratio of the pipe (for 20°C water and pipes with expansion joints).

Design and Operation of Farm Irrigation Systems

549

15.3.1.2 Combination air valves. Air admission/air release valves (see Figure 15.8) allow air into the line when the pressure drops below atmospheric pressure and discharge air (if any) when the pressure becomes positive. Negative pressures in a line during a surge may be better understood by visualizing the negative pressures as a drooping string hanging below the pipe profile. The string is pulled up to the pipe at control points, such as reservoirs or air valves. The string still droops between these control points. The only way to prevent column separation is to make the valves large enough to pull the pressures up to near atmospheric pressure at the valves and place the valves close enough together to keep the drooping pressure between valves at a safe level. Pressures should normally be above 50 kPa, which yields a safety factor of about two in avoiding column separation. Generally, the admission of large quantities of air into pipelines is not a good practice. Air cavities are created at low pressure; this limits the cushioning characteristics of the cavities when water columns rejoin. High pressures can result when water columns collide. In addition, air must be expelled to ensure a smooth, efficient pipeline operation. 15.3.1.3 Surge anticipator valves. Surge anticipator valves open when the pressure drops below a specified set point. They remain open for a pre-set period of time, and then close in a manner that prevents the high pressure caused by rapid valve closure. Properly installed, surge anticipator valves prevent high pressures at pump stations but have little ability to alleviate column separation, especially in long pipelines. Furthermore, valve movements must be set with care; otherwise, they may cause more severe water hammer problems than they correct. Anticipator valves can also be set to “open” when line pressures exceed a pre-set value, thus operating as a surge relief valve. 15.3.1.4 Surge relief valves. Pressure relief valves consist of a movable seal plate that is held in position by a spring and jam nut as illustrated in Figure 15.8. When pressure in the pipeline exerts more force on the seal plate than the spring can resist the seal plate rises and allows high pressure water to exit from the pipeline. Surge or pressure relief valves can reduce maximum surge pressures if they are adjusted properly. They open at a pre-set pressure level and relieve excessive pressure at a critical point in a pipeline, often at the discharge of pumps. Like surge anticipator valves, surge relief valves cannot prevent column separation. 15.3.1.5 Check valves. Check valves prevent or limit reverse flow and should be designed to prevent slamming. Excessive noise and pressure spikes upon final closure are two negative results of slamming valves. Check valves are available in numerous types, styles, and shapes. Below we summarize the desirable and undesirable features of the five basic types of check valves: the swinging check valve (both single disc and double disc), the slanting-disc check, the silent check (poppet, disc, ball), the control disc check, and the foot check. The various types of check valves may look different, but all check valves perform the same basic function: they facilitate flow in one direction and control for a no-flow (check) return. Check valves are commonly installed on the discharge side of the pump. The most important role of a check valve is to act as the automatic shutoff valve when the pump stops. This prevents the draining of the system that the pump is filling. However, each check valve has different shutoff characteristics.

550

Chapter 15 Hydraulics of Sprinkler and Microirrigation Systems

Pressure surge (water hammer) can be greatly reduced by selecting the correct check valve. For example, silent-check valves shut off very quickly, in 1/10 to 1/20 of a second; this makes them excellent surge protection devices when pumping short distances (less than 800 meters). Conversely, the control-disc check valve has extremely slow and variable speed shutoff; it is an excellent surge protection device when pumping long distances (more than 800 meters). 15.3.2 Check Valve Comparisons 15.3.2.1 Swinging check (single disc). Swinging check valves (Figure 15.10) depend on gravity and reverse flow to close the valve and shut off reverse flow. Many models include a spring to help close the valve. The pivot point of the swinging check is outside of the periphery of the disc and the larger the head, the greater the possibility that the fluid will flow back through the valve before the disc can shut. To affect complete shutoff, the disc of a swinging check valve must travel through a 90° arc to the valve seat. Without resistance to slow the disc’s downward thrust, the shutoff results in slamming and potentially damaging water hammer. Caution should be used when applying this type of check valve when velocities exceed 1 m/s reverse flow and pressures are higher than 200 kPa.

Figure 15.10. Sketch and photograph of single-disc swinging check valve.

15.3.2.2 Double-disc swinging check (split-swing disc). The double-disc swinging check valve (Figure 15.11) is similar to the conventional swinging check valve except that the single disc is split for the double-disc valve. Splitting the disc reduces the mass of the disc and the travel distance from the open to the shutoff position. Further improvement results from torsion springs that force the double discs to shut with minimal flow reversal. In comparison to the conventional swinging check, this valve’s shutoff characteristic minimizes the slam potential. With double-disc checks, the hinge pin is stationary and each disc swings freely when opening or closing. Multiple springs are used on larger sizes to compensate for heavier discs and improve shutoff speed.

Design and Operation of Farm Irrigation Systems

551

Figure 15.11. Diagram of the components of a double-disc swinging check valve.

15.3.2.3 Slanting-disc check (with off-center pivot). Slanting-disc check valves have an enlarged flow area through the disc and seat sections. The disc pivots offcenter, with approximately 30% of the disc area above the pivot point offering resistance against the 70% disc area below the pivot point. Hence, this type of check valve has built-in non-slam shutoff characteristics. The shutoff seating angle is 55°. The disc swings open through this 55° seat angle, travels a short distance, and stops at 15° off of the horizontal. Hence, the disc travels only a short distance, allowing minimal reverse flow and non-slam shutoff. 15.3.2.4 Silent check. Silent check valves have a center-guided poppet that is spring loaded and is normally closed. The feature that results in the silent shutoff is the relation of the poppet to the seat when in the open position. This distance is approximately 1/4 of the valve size; in other words, a 100-mm valve has a 25-mm distance to shut off. It is the short poppet travel distance coupled with the spring force that results in a silent shutoff. The approximate time of shutoff is between 1/20 and 1/10 of a second. Silent check valves are furnished with springs, eliminating any possibility of reverse flow and minimizing the water hammer or pressure surge. 15.3.2.5 Control-disc check. The control-disc check valve is a unique adaptation of the slanting-disc check valve. All previously described check valves open and close when flow from the pump starts or stops. The opening and closing of the disc, poppet, or flapper is uncontrolled. The control-disc check valve is recommended where pump flow control is essential to prevent pressure surges when flow starts or stops in the pipeline. The valve disc can be electrically controlled to permit remote operation of automatic pump stations. The valve functions as a shutoff valve, slowly opening after flow starts and closing slowly when stopping the flow to close the control-disc check valve. The valve shuts off automatically if electrical power is interrupted. This type of valve is available with hydraulic controls and disc position indicators. Similar control functions can be achieved with hydraulic cylinder-actuated or motor-actuated butterfly valves.

552

Chapter 15 Hydraulics of Sprinkler and Microirrigation Systems

15.3.2.6 Foot valve. A foot valve is a form of check valve installed at the bottom of the pump suction line and below water. Foot valves are inexpensive ways to prime centrifugal pumps. The foot valve is installed in the vertical position with the direction of flow upward. In this position, the foot valve is normally closed. It is necessary to fill the suction line with water prior to starting the centrifugal pump. Once the suction line is filled, the foot valve takes over and opens while the centrifugal pump is running and closes when the pump stops. This maintains a flooded suction and primed pump. Since foot valves are continually submerged and are not readily accessible for inspection or repair, it is important to select high-quality, long-wearing valves. 15.3.3 Theory and Use of Air Valves 15.3.3.1 Air release valves. Air entrained in conveyed water will often separate from the liquid and collect at high points within the pipeline. If provisions are not made to remove air from high points, pockets of air will collect and grow in size. Air pocket growth will gradually reduce the effective liquid flow area, creating a throttling effect similar to a partially closed valve. Often, the velocity of liquid flow will be sufficient to carry a portion of the enlarging air pocket downstream and the air will lodge at another high point. The ability of flow velocity to trim the size of air pockets may minimize the effect of reduced flow area in the pipe, but the throttling effect caused by the presence of air will always reduce the flow rate below design. Air pocket problems are difficult to detect, but if permitted to continue may cause a constant power drain. Air release valves installed at high points of the system shut off when filled with liquid and are then subjected to pressure. During system operation, small particles of air will separate from the liquid and enter the valve. Each particle of air will displace an equal amount of the liquid within the valve and lower the liquid level relative to the float. When the liquid level lowers to a point where the float is no longer buoyant, the float will drop. This action opens the valve orifice and allows the air that has accumulated in the upper portion of the valve to be released into the atmosphere. As air is released, the liquid level within the valve once again rises, lifting the float and closing the valve orifice. This cycle repeats itself as often as air accumulates in the valve. 15.3.3.2 Air and vacuum valves. An air and vacuum valve (AVV) is floatoperated with a large discharge orifice equal in size to the valve’s inlet. This valve allows great volumes of air to exhaust from or enter into a system as it fills or drains. The following conditions prevail with AVVs used in pipelines: ƒ Prior to filling, a pipeline may be thought of as empty, but this is not true. In reality it is filled with air. The air must exhaust from the pipeline in a smooth, uniform manner to prevent pressure surges and other destructive phenomena. ƒ To prevent potentially destructive vacuums from forming, air must re-enter the pipeline in response to a negative pressure. Even in those instances where vacuum protection is not a primary concern, air re-entry is still essential to efficiently drain the pipeline and prevent water column separation, which can be as damaging as pressure surges. To perform the functions outlined, an AVV must be installed at each high point or change in grade, at the ends of pipelines, and at isolation valves. As a pipeline fills air exhausts into the atmosphere through an AVV. As air is exhausted, water enters the valve and lifts the float to close the valve orifice. The rate at which air is exhausted is a function of the pressure differential that develops across the valve discharge orifice. This pressure differential develops as water filling the pipeline

Design and Operation of Farm Irrigation Systems

553

compresses the air sufficiently to provide an escape velocity equal to that of the incoming fluid. Since the size of the valve controls the pressure differential at which the air is exhausted, valve size selection is very important. Should the internal pressure of the pipeline approach a negative value at any time during system operation due to column separation, draining of the pipeline, power outage, or pipeline break, the float will immediately drop away from the orifice and allow air to re-enter the pipeline. Air re-entry will prevent a vacuum from occurring and protect the pipeline against collapse or water column separation. The size of the AVV will dictate the degree to which a vacuum is prevented; therefore, selecting the correct valve size is very important. The AVV, having opened to admit air into the pipeline in response to a negative pressure, is now once again ready to exhaust air. This cycle will repeat as often as necessary. During system operation and while under pressure, small amounts of air will enter the AVV from the pipeline and displace the fluid. Eventually, the entire AVV may fill with air but it will not open because the system pressure will continue to hold the float closed against the valve orifice. To reiterate, an AVV is designed to exhaust air during pipeline fill and to admit air during pipeline drain. It will not open and vent air as it accumulates during system operation—air release valves are used for this purpose. 15.3.3.3 Combination air valves. Combination air valves consist of a float that moves vertically inside the valve to control the discharge and entry of air into the pipeline (Figure 15.12). As the name implies, combination air valves (CAVs) have operating features of both air and vacuum valves and air-release valves. CAVs are also called double-orifice air valves. Combination air valves are installed at high points of a system where it has been determined that dual-purpose air-and-vacuum and airrelease valves are needed to vent and protect a pipeline. Generally, it is sound engineering practice to use CAVs instead of single-purpose air and vacuum valves at high points. Combination air valves prevent accumulations of air at high points within a system by exhausting large volumes of air as the system is filled and by releasing accumulated pockets of air while the system is operational and under pressure. Combination

Figure 15.12. Picture of a combination air control valve for when the float is in the down position and when it is raised to operate in a continuous position.

554

Chapter 15 Hydraulics of Sprinkler and Microirrigation Systems

air valves also prevent potentially destructive vacuums from forming by admitting air into the system during power outages, water column separations, or a sudden drainage of the pipeline.

15.4 WATER DISTRIBUTION TO THE SOIL The objective of microirrigation is to apply water for individual plants at rates slow enough to prevent water translocation on the surface. Tree crops are often irrigated with the emitter applying the water under the canopy, whereas with row crops the water is applied near the plant or in the row to provide the required water for production but limit the evaporation from a wet soil surface. Buried microirrigation minimizes the soil surface evaporation while providing sufficient water for plant growth. In contrast, the objective of sprinkler irrigation is to uniformly apply water over the area for crop use. Sprinkler irrigation systems should be designed to apply water at a rate less than the intake rate to prevent surface runoff. The water application uniformity of a sprinkler system depends less on the dynamics of the individual sprinklers and more on the spacing and operating pressure. The spacing of sprinklers and, on moving systems, the travel speed and path affect irrigation uniformity. The effects of wind and evaporation on uniformity are discussed in more detail in Chapter 16.

15.5 REDISTRIBUTION IN THE SOIL The uniformity of sprinkler irrigation systems normally is evaluated on the basis of distribution at the soil surface. The crop, however, is dependent on available water in the root zone. Water will move within the soil profile due to potential gradients, and plant roots have a horizontal distribution. Hart (1972) analyzed the subsurface redistribution, using finite-difference solutions of the two-dimensional diffusion equation. He concluded that there is substantial redistribution within 1 m and final soil water distribution is not significantly affected by the nonuniformity on the soil surface within that distance (as illustrated in Figure 6.6). An estimated 10% reduction in total water requirements resulted due to the redistribution of soil water for one test field. Where the nonuniformity exists over larger than 1- to 3-m distances, the subsurface redistribution will have little effect on the final soil-water uniformity. Seginer (1978) estimated a net loss of $180/ha for a 0.1 decrease in the uniformity coefficient in cotton production due to nonuniform water application. The effective nonuniformity for various crops is a function of the extent of horizontal root development of individual plants. Seginer (1979) analyzed the increase in effective uniformity of water distribution in orchards with various tree and sprinkler spacing and calculated an increase in effective uniformity of 0.1. The “actual” uniformity is appropriate for “point” plants that have a limited horizontal root distribution. He also found that when rows are parallel to the shorter sprinkler spacing of solid set systems, a better effective uniformity can be obtained. Microirrigation systems generally apply the water directly to the root zone. However, in some cases the spacing of the emitters does rely on horizontal movement to become available for the root system. This is particularly true for row crops where emitters are not located in each row but in between or every other row spacing.

15.6 SUMMARY Design and operation of sprinkler and microirrigation systems depends on proper selection and sizing of components, which rely on the water pressure distribution

Design and Operation of Farm Irrigation Systems

555

throughout the system. This chapter discusses the hydraulics of mainline and lateral pipelines, performance of pressure regulation devices, and characteristics of valves and pipeline protection devices commonly used in pressurized irrigation systems. Figures and charts are provided for use in design and operation. The ability of soils to redistribute soil water and thereby affect the uniformity of water availability to plants is briefly reviewed.

LIST OF SYMBOLS a1,b1 C c Cd D Dn d1,d2 e Ei Ei -1 f F F′ g Ha Hd hf hfi Hi-1 HL Hmax Hmin Ho hr Hx Ks L L m N P Pd Pmin Q qs qd Qx

coefficients for the sprinkler discharge equation pipe roughness coefficient for the Hazen-Williams equation velocity of a pressure wave in the pipeline, m/s discharge coefficient for sprinkler discharge inside diameter of the pipe, m diameter of the nozzle, mm coefficients for the Hazen-Williams and Scobey equations absolute roughness of the inside of the pipe, mm elevation at the inlet section i of a pipeline, m elevation at the outlet of section i of a pipeline, m Darcy-Weisbach resistance coefficient adjustment factor for lateral pressure loss when first sprinkler is one sprinkler spacing from the inlet adjustment factor for lateral pressure loss when first sprinkler is one-half the sprinkler spacing from the inlet gravitational acceleration, 9.81 m/s2 the average pressure head along a sprinkler lateral, m the pressure head at the distal end of a sprinkler lateral, m head loss, m head loss due to friction in the section pressure head at the inlet to the section pressure head at radial distance L from the pivot point, m the maximum pressure head along the sprinkler lateral, m the minimum pressure head along the sprinkler lateral, m pressure head at the inlet to a lateral or pivot, m head loss in a sprinkler riser, m pressure head at a radius x from the pivot, m coefficient of retardation for the Scobey equation pipe length, m radius of the irrigated area, m exponent in the friction loss equation number of sprinkler outlets on the lateral pressure in the pipeline pressure at the distal end of a sprinkler lateral, kPa minimum value for the average pressure on a sprinkler lateral, kPa flow in the pipe, m3/s or L/s sprinkler discharge; L/s discharge of the last sprinkler on a lateral, L/s flow in the lateral at a point along the lateral, L/s or m3/s

556

Re Sf V v VT x

Chapter 15 Hydraulics of Sprinkler and Microirrigation Systems

Reynolds number uniform slope of the lateral from the inlet, m/m velocity of flow, m/s kinematic viscosity, m2/s total volume required for a surge tank, m3 distance from the inlet to a point along a sprinkler lateral, m

REFERENCES Albertson, M. L., J. R. Barton, and D. B. Simons. 1960. Fluid Mechanics for Engineers. Englewood Cliffs, N.J.: Prentice-Hall, Inc. Christiansen, J. E. 1942. Irrigation by sprinkling. Univ. Calif. Agr. Exp. Sta. Bull. 670. Chu, S. T., and D. L. Moe. 1972. Hydraulics of a center pivot system. Trans. ASAE 15(5): 894-896. Hart, W. E. 1972. Subsurface distribution of nonuniformity applied surface waters. Trans. ASAE 15(4): 656-661, 666. Jensen, M. C., and A. M. Fratini. 1957. Adjusted “F” factor for sprinkler lateral design. Agric. Eng. 38(4): 247. Keller, J., and R. D. Bliesner. 1990. Sprinkle and Trickle Irrigation. New York, N.Y.: Van Nostrand Reinhold. Kincaid, D. C., and D. F. Heermann. 1970. Pressure distributions on a center-pivot sprinkler irrigation system. Trans. ASAE 13(5): 556-588. Liou, C. P. 1998. Limitations and proper use of the Hazen-Williams Equation. J. Hydraulic Eng. ASCE 124(9): 951-954. Lytle, W. F., and J. E. Winberly. 1962. Head loss in irrigation pipe couplers. Bull. No. 553. Baton Rouge, La.: Louisiana State Univ. and Agric. Mech. College, Agric. Exp. Station. Meadows, M. E., and T. M. Walski. 1998. Computer Applications in Hydraulic Engineering. Waterbury, Conn.: Haestad Press. Seginer, I. 1978. A note on the economic significance of uniform water application. Irrig. Sci. 1: 19-25. Seginer, I. 1979. Irrigation uniformity related to horizontal extent of root zone: A computational study. Irrig. Sci. 1: 89-96. von Bernuth, R. D. 1983. Nozzling considerations for center pivots with end guns. Trans. ASAE 26(2): 419-422. von Bernuth, R. D., and D. Baird. 1990. Characterizing pressure regulator performance. Trans. ASAE 33(1): 145-150.

CHAPTER 16

DESIGN AND OPERATION OF SPRINKLER SYSTEMS Derrel L. Martin (University of Nebraska, Lincoln, Nebraska) Dennis C. Kincaid (USDA-ARS, Kimberly, Idaho) William M. Lyle (Texas A&M University, Lubbock, Texas) Abstract. The design and operation of sprinkler irrigation systems are presented and discussed in this chapter. Information is provided concerning the components of irrigation systems and the characteristics of the system that affect water application efficiency and uniformity. The interactions of system characteristics with soil and crop properties are evaluated relative to design and operation of irrigation systems to minimize runoff, deep percolation, and excessive evaporation losses. Keywords. Application efficiency, Center pivot, Lateral move, Runoff, Side roll, Solid set, Sprinkler irrigation, Towline, Uniformity.

16.1 INTRODUCTION Sprinkler systems have revolutionized the development of irrigated agriculture. Efficient water application with sprinkler irrigation involves the design and operation of pumps, pipes, and sprinkler devices to match soil, crop, and resource conditions. Thus, sprinkler systems can be designed and operated for efficient irrigation for a wide array of conditions. Sprinkler irrigation also facilitated the expansion of irrigated agriculture onto lands classified as unsuitable for surface irrigation. Initially the labor required to transport the system across the field impeded the adoption of sprinkler irrigation. Through automation the labor required for sprinkler irrigation has been reduced significantly. Reduced labor requirements enabled producers to irrigate more frequently with smaller water applications which diminished unintentional leaching and increased the potential to store precipitation in the crop root zone while satisfying crop requirements. Morgan (1993) indicated that sprinkling devices were used as early as 1873. By 1898 seventeen patents had been issued for sprinkler devices. Since that early begin-

558

Chapter 16 Design and Operation of Sprinkler Systems

ning many developments have occurred. The patent for impact sprinklers as we know them today was issued in 1934. Aluminum pipe with rubber gaskets were first produced in the late 1940s while an early version of the side roll machine was first produced in the 1950s. Self-propelled center pivot and lateral move systems were invented in the late 1940s. Producers quickly recognized that controlling an irrigation system was essential for proper performance. One of the first controllers for sprinkler irrigation was installed in 1924. These early developments laid the foundation for the growth of sprinkler irrigation. In the late 1940s and early 1950s development began in earnest and continued with large increases in the 1960s and 1970s when automated systems were commercialized. The amount of land irrigated with sprinkler systems continues to increase. Most of the development today is devoted to automated and semi-automated sprinkler systems. According to the Farm and Ranch Irrigation Survey (USDA-NASS, 2003), sprinkler irrigation was used on approximately 51% of the land irrigated in the United States (Table 16.1). Currently, center pivots are used on approximately 79% of the land that is irrigated by sprinkler. Slightly more low-pressure center pivots are used than medium-pressure pivots. Relatively few high-pressure pivots were used in 2002. Table 16.1. Distribution of sprinkler irrigated land in the United States in 2002 (source: USDA-NASS, 2003). Land Area (hectares) by Pressure Range System

< 207 kPa 210 to 414 kPa > 414 kPa

Total Area % of Sprinklerhectares Irrigated Land

Center pivot

3,925,883

Lateral move Solid set

8,620,685

79.0

60,553

3,909,860 78,784

784,943

139,337

1.3

112,551

364,353

476,904

4.4

Side roll

--

--

739,231

6.8

Traveler or big gun

--

--

256,351

2.4

--

673,498 10,906,006 21,288,338

6.2

Hand move -Total sprinkler-irrigated land Total land irrigated with all types of systems

16.2 COMPONENTS OF SPRINKLER SYSTEMS Sprinkler irrigation systems generally consist of a pump used to lift and pressurize water, a main line piping system to convey water from the pump to the lateral pipelines used to transport water across the irrigated field, and sprinkler devices to apply the water within the field (Figure 16.1). In some cases the main line is subdivided into submains to convey water to portions of the field. Multiple laterals may be used if the field is large or if the field needs to be irrigated frequently. Sprinkler devices are installed at intervals along the lateral. Often a pipe called a riser is used to adjust the height of the sprinkler device to the desired height to avoid crop interference with the jet of water discharged from the sprinkler. For some applications with center pivots or lateral move systems, the sprinkler device is suspended below the lateral using drop tubes to minimize evaporation and drift of sprinkler droplets.

Design and Operation of Farm Irrigation Systems

559

Figure 16.1. Components and general layout of sprinkler irrigation systems.

The proper spacing of sprinklers along the lateral is crucial to providing enough overlap of water from adjacent sprinklers to achieve uniform irrigation. Sprinkler laterals can be permanently installed as with solid set systems, or the laterals can be moved across the field. Laterals can be periodically moved from one location (often called a “set”) to the next location as with hand move systems or the laterals can move continuously across the field as with center pivot and lateral move systems. When the sprinkler lateral is periodically moved across the field the distance between subsequent sets of the lateral must be narrow enough to provide for adequate overlap. Many kinds of sprinkler devices are available; some are shown in Figure 16.2. All sprinkler devices use a nozzle to control the discharge from the sprinkler. The diameter and shape of the nozzle orifice and the water pressure at the nozzle control the rate of flow. The nozzle converts pressure within the piping system to velocity upon discharge from the sprinkler. The velocity propels droplets through the air to provide a wetted pattern about the sprinkler. The design of the nozzle and the sprinkler device determines the diameter of coverage and the distribution of water within the wetted region. As with any irrigation system, the sprinkler system must be laid out to conform to the boundaries of the field. The dimensions of the field physically or economically often dictate the type of sprinkler system to be used. Ultimately the components of the sprinkler system must be matched to irrigate efficiently and economically. Inappropriate specification of any component will negatively affect the entire system.

560

Chapter 16 Design and Operation of Sprinkler Systems

Figure 16.2. Examples of sprinkler devices used on irrigation systems.

16.3 DESIGN FUNDAMENTALS Sprinkler irrigation systems should be designed to satisfy crop water requirements while applying water at a rate that minimizes runoff and excess leaching. To accomplish these objectives, the pressure distribution in the main line and sprinkler lateral must be appropriate as discussed in Chapter 15. Efforts are underway to design systems that apply variable depths of water within a field for site-specific management. However, emphasis is currently on systems to apply water uniformly throughout the field. General concepts that apply to design and operation of sprinkler systems are developed in this section. Details for specific systems are presented later. 16.3.1 Application Efficiency The application efficiency is the fraction of the applied water that remains in the crop root zone following irrigation. Water that infiltrates and remains in the root zone is the net irrigation. The application efficiency is then the ratio of the net irrigation depth (dn) to the gross depth of water applied (dg). The possible losses of water in sprinkler irrigation are illustrated in Figure 16.3. Some water may evaporate before it reaches the crop canopy or the soil surface. A portion of the water is intercepted by the canopy while the remainder falls through the canopy to the soil surface. Water that is intercepted by the canopy can be retained on the canopy, may run down the stem of the plant to the soil surface, or may drip from the plant. Steiner et al. (1983) showed that for a mature corn crop up to 50% of the

Design and Operation of Farm Irrigation Systems

561

Figure 16.3. Diagram of water losses for sprinkler irrigation.

water applied with a center pivot flowed along the leaves and stems to the ground. Water that is intercepted and/or retained on the canopy can evaporate during and after irrigation. Water that reaches the soil surface can evaporate, infiltrate, run off, or be stored in depressions along the soil surface. When infiltration exceeds the available storage in the crop root zone, the excess application will slowly drain through the soil profile inducing deep percolation. Water stored on the soil surface has been called surface storage, depression storage, or retention storage. This water is available to infiltrate after the rate of application of water decreases below the infiltration rate of the soil. Water that runs off the point of application can infiltrate at a downstream location and may be used. If the runoff leaves the field, the water is lost to that field. If the runoff accumulates in low-lying areas deep percolation may occur. The type of system, its design and operation, and the soil and climatological conditions at the time of irrigation all affect the application efficiency. Design of sprinkler systems is based upon average application efficiencies. Typical efficiencies are summarized in Table 16.2. 16.3.2 System Discharge The discharge required for the irrigation system is one of the first considerations in the design process. The discharge must be sufficient to meet crop water requirements for the climatological conditions of the region. We are using the net system capacity as the indicator of the supply required to meet crop needs. The net system capacity (Cn) is the rate water must be continually supplied to satisfy plant water requirements. Heermann et al. (1974) used a daily soil water balance to simulate the effect of net system capacities on the soil water content. For a given capacity, they determined the maximum soil water depletion reached during a year. They simulated the soil water balance for a 59-year period to develop data for a probability distribution of the net

562

Chapter 16 Design and Operation of Sprinkler Systems

Table 16.2. Characteristics of sprinkler irrigation systems (adapted from Keller et al., 1980). Installation Annual Cost of Labor per Irrigation, Application Coefficient Maintenance (hour ha-1) Type of Cost ($ Efficiency of -1 System hour ) (% of purchase) Preseason In Season (%) Uniformity Hand move

300

Towline

2

0.25`

1.73

65-75

70-85

500

3

0.25

1.00

70-80

70-85

Side roll Side move

675

2 4

0.25 0.49

0.86 0.62

70-80

70-85

Solid set

2000

2

0.25

0.15

70-85

75-85

80-92

85-95

85-92

85-95

60-70

70-80

Center pivot standard with corner Lateral move ditch fed hose fed pipe fed

800 925

5 6

0.12

0.05 0.06

-1250 --

6

0.12 0.15 0.12

0.10 0.15 0.07

Travelers

900

6

0.25

0.62

system capacity requirement. Results from von Bernuth et al. (1984) were used to develop design guidelines for center pivots in Nebraska (Figure 16.4). Howell et al. (1989), and Lundstrom and Stegman (1988) used similar procedures for other areas in the Great Plains. The inflow rate required for the system is the amount of water that must be supplied to avoid water stress. The system flow rate is given by ⎛ d g Ai ⎞ ⎛ C T ⎞⎛ A ⎞ ⎟ QS = 0.116⎜⎜ n i ⎟⎟⎜⎜ i ⎟⎟ = 0.116⎜⎜ ⎟ ⎝ Ea ⎠⎝ To ⎠ ⎝ To ⎠

16.1

where Qs = inflow to the sprinkler irrigation system (i.e. gross system capacity), L s-1 Cn = net system capacity, mm day-1 Ti = irrigation interval, days Ea = application efficiency, decimal fraction. Ai = area irrigated, ha, and To = time of operation per irrigation, days dg = gross depth of irrigation water applied, mm The irrigation interval is the time from the start of the irrigation until the beginning of the subsequent irrigation. The irrigation interval is often a nominal value representing normal practices during the peak water requirement period of the irrigation season. The time of operation is the amount of time that water is applied during the irrigation interval. There are times when the irrigation system may be shut down for maintenance, repositioning sprinkler laterals, farming operations, or convenience. This time can be referred to as the downtime. The downtime is the difference between the irrigation interval and the time of operation. The selection of the irrigation interval and downtime must be coordinated with the operator and should not be arbitrarily selected by the designer.

Design and Operation of Farm Irrigation Systems

563

NET SYSTEM CAPACITY, L/s/ha

1.0

0.9 WESTERN NEBRASKA

0.8

0.7

0.6

EASTERN NEBRASKA

0.5 25

50

75

100

125

150

ALLOWABLE SOIL WATER DEPLETION, mm

Figure 16.4. Net system capacity requirement for center pivots in Nebraska (adapted from von Bernuth et al., 1984).

The total system capacity cannot exceed the available water supply. If the capacity requirements are too high, the amount of area irrigated may have to be reduced, or deficit irrigation will be necessary. 16.3.3 Sprinkler Discharge The discharge required for a sprinkler can be determined by the density of sprinklers and the number of laterals within the field. For solid set and moved-lateral systems the density is determined by the spacing between the sprinklers along a lateral and the distance between sets along the main line (Figure 16.2). The representative area for a single sprinkler is the product of the sprinkler spacing along the lateral (SL) and the distance between laterals or the set width (Sm). The number of sprinklers on a lateral (N) is determined by the length of the field (FL), the number of laterals used along the field length (NL) and the sprinkler spacing (i.e., N = FL/NL SL). The number of sets per lateral (n) is determined by the width of the field (FW), the number of laterals along the width of the field (Ns) and the width of an individual set (n = FW /Ns Sm). The number of sprinklers per lateral and the number of sets per lateral must be integers. The irrigation interval depends on the number of sets per lateral (n), the operational time per set for a lateral (Ts), the time required to move the sprinkler later between sets (Tm), and downtime between successive irrigations: (16.2) Ti = n (Ts + Tm) + Td The combination of the operational time and the time required to move from one set to the next is often referred to as the set time. The set time must be selected in consultation with the owner/operator to match labor availability. The minimum downtime is the time required to reposition laterals from the last set in the field to the beginning set for the next irrigation. Additional time may also be required for maintenance and farming operations.

564

Chapter 16 Design and Operation of Sprinkler Systems

The discharge from an individual sprinkler for such systems depends on (1) the depth of water that must be applied per irrigation, (2) the representative area for one sprinkler, and (3) the time that water is applied for an individual set, as qs =

1 ⎛ CnTi ⎞⎛ S L Sm ⎞ d g S L Sm ⎟⎜ ⎟= ⎜ 3600 ⎜⎝ Ea ⎟⎠⎜⎝ Ts ⎟⎠ 3600Ts

(16.3)

where qs = discharge from a sprinkler, L s-1 SL = spacing between sprinklers, m Sm = distance between laterals, m Ts = operational time per set for a single lateral, hours Part-circle sprinklers can be placed at the end of the lateral to provide uniform irrigation without applying water beyond the field boundaries. The discharge for an end sprinkler should be a fraction of the discharge of sprinklers along the lateral as based on the proportion of a full circle that is watered with the end sprinkler. Some solid set systems are designed so that the area between adjacent sprinklers forms an equilateral triangle. For this arrangement the distance between sprinklers is equal to the spacing along the lateral (SL) and the distance between laterals is Sm = 0.866 × SL. Relationships are developed in later sections for the discharge from individual sprinklers on other types of systems. The pressure available to the sprinkler device and the effective diameter of the nozzle determine the discharge from a sprinkler. For circular nozzles the relationship is given by qs = 0.00111 Cd Dn2 P

(16.4)

-1

where qS = discharge per nozzle, L s Cd = discharge coefficient Dn = inside diameter of the nozzle, mm P = pressure at the base of the sprinkler device, kPa. The discharge coefficient, which incorporates characteristics of the sprinkler head and nozzle, is unique for sprinkler devices and nozzles. Heermann and Stahl (2006) developed equations to estimate the discharge coefficient for selected sprinkler devices: Cd = d0 + d1 Dn2 + d2P (16.5) where d0, d1 and d2 are empirical parameters. Typical parameter values are listed in Table 16.3 for selected sprinkler devices. The discharge from the range nozzle and the spreader nozzle should be added for sprinklers equipped with two nozzles. For some nozzles the orifice is not circular and the effective diameter of the nozzle should be used in Equations 16.4 and 16.5. 16.3.4 Diameter of Coverage The diameter of coverage, also called the wetted diameter, (Wd), is the size of the circular pattern that is wetted by an individual sprinkler. The diameter of coverage has a large influence on sprinkler and lateral spacing and thus the cost of a system. The diameter of coverage depends on the velocity of the sprinkler jet, the angle of the nozzle axis above the horizon and the size of droplets that form when the sprinkler jet breaks up. Up to a point, as pressure increases, the diameter of coverage increases and

Design and Operation of Farm Irrigation Systems

565

more uniform application may result. The pressure should be increased as the nozzle sizes increase to ensure that the spray breaks up adequately. Sprinkler manufacturers provide performance data that includes the discharge and diameter of coverage as a function of operating pressure and nozzle sizes as illustrated in Table 16.4. The diameter of coverage provided by the manufacturer generally specifies the height of the nozzle above the ground when the device was tested. Test results are usually for no-wind conditions and include the recommended operating pressure range for the listed configurations. Table 16.3. Coefficients for the discharge coefficient (adapted from Heermann and Stahl, 2006). Minimum Model Pressure Company d1 d2 (kPa) d0 Number 245 0.954 1.859 × 10-5 5.061 × 10-3 Nelson Irrigation Corp. F32AS 245 0.954 Nelson Irrigation Corp. F32S 5.061 × 10-3 1.859 × 10-5 -3 245 0.976 -1.617 × 10 -2.567 × 10-6 Nelson Irrigation Corp. F33A 245 0.976 -1.617 × 10-3 -2.567 × 10-6 Nelson Irrigation Corp. F33AS -3 245 0.976 -1.617 × 10 Nelson Irrigation Corp. F33S -2.567 × 10-6 -4 196 1.011 -2.267 × 10 -3.441 × 10-4 Nelson Irrigation Corp. F33DN -3 245 1.011 -4.889 × 10 -5.737 × 10-6 Nelson Irrigation Corp. F43A -5 245 0.945 -3.363 × 10 -2.928 × 10-6 Nelson Irrigation Corp. F43AP -3 245 0.975 -1.390 × 10 -8.234 × 10-6 Nelson Irrigation Corp. F43P -3 294 0.996 -2.018 × 10 -1.357 × 10-6 Nelson Irrigation Corp. F70A -3 392 0.991 -1.002 × 10 Nelson Irrigation Corp. F80A 3.213 × 10-5 -3 147 1.035 -2.845 × 10 -1.359 × 10-4 Nelson Irrigation Corp. R30 196 0.294 Nelson Irrigation Corp. R3000 0 0 294 0.969 Nelson Irrigation Corp. R30D4 0 0 294 0.957 Nelson Irrigation Corp. R30D6 0 0 98 0.986 -2.692 × 10-3 -9.696 × 10-5 Nelson Irrigation Corp. SPR1 196 0.966 -7.768 × 10-3 Rain Bird Corp. 20 7.394 × 10-5 -2 245 1.059 -2.341 × 10 -4.154 × 10-6 Rain Bird Corp. L20 -2 245 1.084 -1.585 × 10 -6.466 × 10-6 Rain Bird Corp. L2020 245 0.995 -3.149 × 10-3 -5.449 × 10-6 Rain Bird Corp. 30 -3 245 1.009 -9.980 × 10 -8.572 × 10-6 Rain Bird Corp. L30 -3 245 1.081 -4.962 × 10 -6.456 × 10-5 Rain Bird Corp. L3030 -3 294 0.952 Rain Bird Corp. 35 2.904 × 10 -4.927 × 10-6 -3 245 1.000 -2.656 × 10 4.358 × 10-5 Rain Bird Corp. 40 -3 392 0.984 -3.061 × 10 7.541 × 10-5 Rain Bird Corp. 65 -3 392 1.002 -3.780 × 10 6.085 × 10-5 Rain Bird Corp. 70 -3 245 0.960 -1.278 × 10 3.300 × 10-5 Rain Bird Corp. 85 -3 98 0.922 -9.870 × 10 1.005 × 10-4 Rain Bird Corp. 8X -4 245 0.963 -6.183 × 10 Senninger Irrigation Inc. S4006 1.269 × 10-5 -3 245 0.990 -2.502 × 10 -7.896 × 10-6 Senninger Irrigation Inc. S5006 -4 59 0.983 -6.662 × 10 -8.100 × 10-5 Senninger Irrigation Inc. SSPR 59 0.865 Valmont Industries, Inc. SPR06 0 0 98 0.816 Valmont Industries, Inc. SPR10 0 0

566

Chapter 16 Design and Operation of Sprinkler Systems

Table 16.4. Performance data for an example sprinkler. Nozzle Sizes (mm) 4.37 × 2.38 4.76 × 2.38 4.76 × 3.18 5.16 × 3.18 5.56 × 3.18 Nozzle Diam. Diam. Diam. Diam. Diam. Pressure Flow cover Flow cover Flow cover Flow cover Flow cover (kPa) (L s-1) (m) (L s-1) (m) (L s-1) (m) (L s-1) m (L s-1) (m) 173 0.35 26.8 0.40 27.4 0.47 27.4 0.52 27.7 0.58 28.4 207 0.38 28.0 0.44 29.0 0.51 29.0 0.58 29.9 0.64 30.5 242 0.41 29.0 0.47 29.9 0.56 29.9 0.63 30.8 0.70 31.4 276 0.44 29.6 0.51 30.5 0.60 30.5 0.67 31.4 0.75 32.3 311 0.47 30.2 0.54 31.1 0.63 31.1 0.71 32.0 0.80 32.9 345 0.50 30.5 0.57 31.7 0.67 31.7 0.75 32.6 0.84 33.5 380 0.52 30.8 0.60 32.3 0.70 32.3 0.79 33.2 0.88 34.1

5.95 × 3.18 Diam. Flow cover (L s-1) (m) 0.64 28.7 0.71 30.8 0.76 32.0 0.82 32.6 0.87 33.5 0.92 34.1 0.96 34.8

Several types of nozzles are available for sprinklers. Conventional straight bore (with round orifices) brass or plastic nozzles produce the maximum pattern radius and good patterns at medium to high pressures. Noncircular orifice nozzles (with diffuse jets) produce good patterns at reduced pressures but the wetted radius is usually smaller than for straight bore nozzles. Kincaid (1982) developed an equation to predict the diameter of coverage based on the nozzle pressure and discharge from the sprinkler. Heermann and Stahl (2006) used the diameter of the nozzle and the sprinkler pressure to predict the wetted radius for a range of sprinklers:

( ) ( )

Wr = r0 + r1 Dn2 P + r2 Dn2 P

2

(16.6)

where Wr= radius of coverage, m Dn = nozzle diameter, mm P = nozzle pressure kPa, r0, r1, r2 = empirical constants. Values for the empirical parameters are listed in Table 16.5 for selected devices. 16.3.5 Time of Operation The time of operation per set (Ts) should be selected to satisfy crop water requirements throughout the irrigation interval while avoiding deep percolation. The operational time must also be acceptable to the operator. The appropriate set time depends on the rate of water application. The average application rate is the average rate that water is applied to the crop/soil surface. The application rate depends on the discharge and the representative area for an individual sprinkler. The average application rate for solid set and periodically moved systems is given by Ra =

3600 qS S L Sm

(16.7)

where Ra = the average application rate, mm hour-1. The rate of application is important since runoff may occur when the application rate exceeds the infiltration rate. The recommended maximum and minimum applica-

Design and Operation of Farm Irrigation Systems

567

tion rate for typical sprinkler layouts for periodically moved and solid set systems are listed in Table 16.6. These values represent average conditions and should be adjusted when local information is available or where runoff is prevalent. Table 16.5. Coefficients for the wetted radius for selected sprinklers (adapted from Heermann and Stahl, 2006). Company

Model Number

Nelson Irrigation Corp. Nelson Irrigation Corp. Nelson Irrigation Corp. Nelson Irrigation Corp. Nelson Irrigation Corp. Nelson Irrigation Corp. Nelson Irrigation Corp. Nelson Irrigation Corp. Nelson Irrigation Corp. Nelson Irrigation Corp. Nelson Irrigation Corp. Nelson Irrigation Corp. Nelson Irrigation Corp. Nelson Irrigation Corp. Nelson Irrigation Corp. Nelson Irrigation Corp. Rain Bird Corp. Rain Bird Corp. Rain Bird Corp. Rain Bird Corp. Rain Bird Corp. Rain Bird Corp. Rain Bird Corp. Rain Bird Corp. Rain Bird Corp. Rain Bird Corp. Rain Bird Corp. Rain Bird Corp. Senninger Irrigation Inc. Senninger Irrigation Inc. Senninger Irrigation Inc. Valmont Industries, Inc. Valmont Industries, Inc.

F32AS F32S F33A F33AS F33S F33DN F43A F43AP F43P F70A F80A R30 R3000 R30D4 R30D6 SPR1 20 L20 L2020 30 L30 L3030 35 40 65 70 85 8X S4006 S5006 SSPR SPR06 SPR10

r0 7.66 11.16 9.95 9.95 10.74 9.67 9.25 9.36 12.34 12.62 16.87 9.15 6.40 8.54 7.62 3.73 10.54 8.30 9.48 10.75 10.23 10.27 10.64 12.00 13.09 16.52 16.69 6.47 11.02 11.49 3.83 3.52 2.97

Coefficients r1 1.0188 × 10-3 4.4340 × 10-4 6.7152 × 10-4 6.7152 × 10-4 6.5956 × 10-4 4.2715 × 10-4 4.2364 × 10-4 3.9817 × 10-4 4.5998 × 10-4 2.5634 × 10-4 1.1757 × 10-4 0 0 0 0 7.7219 × 10-4 5.2870 × 10-4 9.5893 × 10-4 4.0092 × 10-4 6.4496 × 10-4 1.8772 × 10-4 1.6917 ×10-4 6.9710 × 10-4 5.1893 × 10-4 3.6063 × 10-4 2.7829 × 10-4 1.9212 × 10-4 3.4306 × 10-4 5.0444 × 10-4 3.8105 × 10-4 4.1793 × 10-4 5.2760 × 10-7 7.6539 × 10-4

r2 -5.9017 × 10-8 -8.8466 × 10-9 -2.0504 × 10-8 -2.0504 × 10-8 -1.7156 × 10-8 -1.1985 × 10-8 -7.8742 × 10-9 -6.0400 × 10-9 -7.2575 × 10-9 -1.9555 × 10-9 -3.0453 × 10-10 0 0 0 0 -4.2217 × 10-8 -3.3920 × 10-8 -8.8110 × 10-8 -2.6504 × 10-8 -1.6816 × 10-8 -4.2494 × 10-9 -3.1560 × 10-9 -1.9120 × 10-8 -9.1826 × 10-9 -3.0844 × 10-9 -1.2432 × 10-9 -6.1428 × 10-10 -6.4353 × 10-9 -1.3677 × 10-8 -3.3509 × 10-9 -1.3894 × 10-8 0 -2.9334 × 10-8

568

Chapter 16 Design and Operation of Sprinkler Systems

Table 16.6. Maximum application rate and depth of infiltration based on soil type and management criteria (partially from Keller and Bleisner, 1990). Maximum Application Rate Available Maximum Depth of Infiltration (mm) for a root zone (mm h-1) Water1 m deep with allowable Holding Soil Slope (%) depletions of: Capacity

0-5 Soil Texture Coarse sand 50 Fine sand 40 Loamy fine sand 35 Sandy loam 25 Fine sandy loam 20 Very fine sandy loam 15 Loam 13 Silt loam 13 Sandy clay loam 10 Clay loam 8 Silty clay loam 8 Clay 5

5-8

8-12

>12

40 32 28 20 16 12 10 10 8 6 6 4

30 24 21 15 12 9 8 8 6 5 5 3

20 16 14 10 8 6 5 5 4 3 3 2

(mm m-1) 50-70 75-85 85-100 110-125 130-150 145-165 150-170 160-200 140-170 150-180 140-180 130-180

35%

50%

65%

21 28 32 42 49 53 56 63 53 56 56 53

30 40 45 60 70 75 80 90 75 80 80 75

39 52 59 78 91 98 104 117 98 104 104 98

The depth of water applied per irrigation (i.e., the volume of water per unit area of land) for solid set and moved-lateral systems can be computed from d g = RaTS =

3600 qS TS S L Sm

(16.8)

The depth of water applied and the corresponding application rate for other types of sprinkler systems will be presented in later sections. The net depth of application (dn = Ea dg) should not exceed the soil water depletion at the time of irrigation; yet, the net depth must be large enough to satisfy the crop water use during the irrigation interval. To satisfy these constraints the time of operation per set must be consistent with the irrigation interval and the net system capacity. If the depth required for the net capacity exceeds the depletion at the time of irrigation more laterals are generally required along the field width. For design purposes the depletion is computed as AD = MAD RD TAW (16.9) where AD = allowable depletion, mm MAD = management allowed depletion, decimal fraction RD = root depth during peak water use period, m TAW = total available water per unit depth of soil, mm of water per m of soil. Representative values for the parameters in Equation 16.9 are available in Chapter 3 and Table 16.6. The net depth of water must then fit within the following range: CnTi ≤ d n =

3600 qS TS Ea ≤ AD = MAD RD TAW S L Sm

(16.10)

where dn is net application depth (mm). This equation can be revised to give the maximum and minimum limits for the operational time per set:

Design and Operation of Farm Irrigation Systems

n Tm + Td 86400 qS Ea −n Cn S L Sm

569

MAD RD TAW S L Sm 3600 qS Ea

≤ TS ≤

(16.11)

The results show that the ability of the soil to hold the net application provides the maximum set time while the time required to satisfy the net capacity provides the minimum set time (Figure 16.5). For conditions between the limits the system can be shut off for more than the one day of downtime as used in the figure. These results illustrate that the amount of downtime increases as the sprinkler discharge increases. If there are no feasible combinations the number of laterals along the field width should be increased. While the equation for the time of operation is complex, it summarizes the considerations required for a given area and soil. The equation is easily analyzed with a spreadsheet program. Solution often involves a trial and error procedure. 16.3.6 Drop Size Distribution The size of sprinkler droplets profoundly affects the performance of sprinkler irrigation systems. The analyses by Edling (1985), Kohl et al. (1987), Kincaid and Longley (1989), and Thompson et al. (1993) show that small droplets evaporate faster than large droplets (Figure 16.6). Individual droplets are often assumed to be spherically shaped. Therefore the surface area is given by (πd2) while the mass of the droplet is given by (πρd3/6) where d is the diameter of the droplet and ρ is the density of water. 20

Sprinkler Spacing 12.2 m Lateral Spacing 18.3 m Applic. Efficiency 0.75 Root Depth 1.2 m MAD 0.50 Sets / Lateral 20 Time to Move Lateral 1h Downtime / set 24 h

TIME OF OPERATION PER SET, hours

16

TOTAL AVAILABLE WATER, mm/m

12

200

160

8 7

6

4

8

120

NET SYSTEM CAPACITY, mm/day

0 0.4

0.5

0.6

0.7

0.8

0.9

1.0

SPRINKLER DISCHARGE, L/s

Figure 16.5. Relationship of acceptable time of operation to sprinkler discharge when net capacity water requirements form the minimum operational time and the soil water holding capacity constrains the maximum operational time.

570

Chapter 16 Design and Operation of Sprinkler Systems 25 Impact Sprinklers 38˚C Air Temperature

EVAPORATION LOSS, %

20

20% Relative Humidity 20˚C Water Temperature

15

10

5

0 0.2

0.4

0.6

0.8

1.0

1.5

2.0

3.0

4.0

5.0

DROPLET DIAMETER, mm

Figure 16.6. Illustration of the effect of droplet size on relative evaporation rates.

Accordingly, the ratio of the surface area to the mass of the droplet varies inversely with the diameter of the droplet. Hence, a higher portion of the water in a small droplet is exposed to the atmosphere which results in larger relative evaporation rates. Devices that produce large drops should reduce evaporation loss and would seem to increase the application efficiency. Research has also shown that small drops are very prone to drift in windy conditions. The drag force caused by the difference in velocity between the droplet and the air is proportional to the projected area of the droplet. The momentum is proportional to the mass of the droplet. Therefore, drag forces are more significant for small drops than large drops. The large influence of drag forces causes small drops to decelerate rapidly in still air. Since momentum is more significant for large drops, they are less affected by drag and decelerate more slowly than small drops. These processes result in a variation of droplet sizes with distance from the sprinkler device. In still air, small drops fall closer to the sprinkler while large drops travel further. In windy conditions the additional drag force from the wind transports droplets downwind. This process is called drift. Drift can reduce irrigation efficiency when the water is deposited outside of the field. Uniformity can be reduced if drift results in significant distortion of the application pattern for a prolonged period. The momentum of large drops resists drift in windy conditions; however, it negatively affects the soil surface. Levine (1952) and others have shown that the impact energy contained in water droplets leads to breakdown of soil aggregates and the formation of a seal on the soil surface which reduces infiltration and can lead to runoff. Levine (1952) showed that sandy soils were less affected by droplet impact than finer textured soils. The kinetic energy of a drop can be expressed as πρd3v/12 where v is the velocity of the droplet.

Design and Operation of Farm Irrigation Systems

571

Stillmunkes and James (1982) summarized literature showing that sealing is related to the amount of kinetic energy per unit area at impact and the accumulation of the energy over time. The kinetic energy per unit area increases with droplet diameter, but reaches a plateau where the kinetic energy per unit area is nearly constant as the drop size increases. Similar results were presented by Kohl et al. (1985). The kinetic energy per unit area is proportional to the diameter of the droplet while the drag coefficient as presented by Seginer (1965) is inversely proportional to the diameter of the droplet. These interactions are such that there is relatively little difference in the velocity of droplets as the size of the drop exceeds 2 or 3 mm (Figure 16.7). Stillmunkes and James (1982) concluded that the kinetic energy per unit area is insensitive to the drop size if the drops are larger than 3 mm and the kinetic energy per unit area is dependent on the application rate and the length of time that water is applied. The equation from Stillmunkes and James (1982) for the kinetic energy per unit area is given by

Ke ρdv 2 ρRTv 2 = = a 2 2

(16.12)

where Ke/a = kinetic energy per unit area ρ = density of water d = depth of water applied v = droplet velocity. R = average application rate T = exposure time 10 20 8 6 5 4

9

VELOCITY, m/s

8 7

3

6

2

5 1 4 3

0.5

2

FALL HEIGHT, m

1 0 0

1

2

3

4

5

6

DROP DIAMETER, mm

Figure 16.7. Effect of drop size and height of fall on the velocity of droplets (adapted from Stillmunkes and James, 1982).

7

572

Chapter 16 Design and Operation of Sprinkler Systems

Data from several sources were used by von Bernuth and Gilley (1985) to estimate the relative infiltration rate for bare soils compared to soils protected with mulch or crop cover (Ir): 0.683 1.271 − 0.353 0.237 I r = 1 − 0.0354 d50 v Sa Si

(16.13)

where Ir = rate of infiltration for bare soil relative to protected soil, decimal fraction d50 = volume median drop size, mm v = velocity of the volume mean size droplet, m s-1 Sa = sand content of the soil, % Si = silt content of the soil, %. The effect of the drop size and velocity on infiltration is illustrated in Figure 16.8. While the evaporation and drift of small drops are significant, the relative volume of water in a particular size range of drops must be considered. For example, the mass of water in a 5-mm diameter drop is 1000 times the volume in a 0.5-mm droplet. The drop size distribution must be considered to determine if losses could be significant.

RELATIVE INFILTRATION RATE

1.0

2 4 6 8

0.8 Droplet Velocity, m/s 0.6

0.4

0.2 SANDY LOAM 0.0 0

1

2

3

4

5

6

7

DROP SIZE, mm

RELATIVE INFILTRATION RATE

1.0 2 0.8

4 6

0.6

8 0.4

Droplet Velocity, m/s

0.2 SILT LOAM 0.0 0

1

2

3

4

5

6

7

DROP SIZE, mm

Figure 16.8. Effect of droplet impact energy on infiltration rates.

Design and Operation of Farm Irrigation Systems

573

RELATIVE FREQUENCY OF DROP SIZE

1.0

STATIONARY PAD - Smooth Plate

0.8

STATIONARY PAD - Medium Grooved Plate

0.6

STATIONARY PAD - Coarse Grooved Plate 0.4 IMPACT SPRINKLER - Straight Bore Nozzle - 4 mm Diameter - 400 kPa

0.2

0.0 0

1

2

3

4

5

6

DROP SIZE, mm

Figure 16.9. Examples of drop size distributions for a stationary pad sprinkler device with three styles of plates and an impact sprinkler with a straight bore nozzle.

This has led to methods to determine the distribution of drop sizes from sprinkler devices operated at varying pressures and heights. Various methods have been used to measure the size of water droplets (see Solomon and Bezdek, 1980; Eigel and Moore, 1983; Kohl and DeBoer, 1984; and Dadiao and Wallender, 1985 for descriptions of the various techniques). Examples of drop size distributions are illustrated in Figure 16.9. Bezdek and Solomon (1983) analyzed various mathematical functions to describe drop size distributions. They concluded that the upper-limit lognormal distribution was adequate for many experimental data sets. The disadvantage of this distribution is that the solution is difficult to derive from experimental data. Until recently, the limitation on considering drop size distributions in design has been the scarcity of experimental data for commonly used sprinklers. Kincaid et al. (1996) developed methods to estimate the distribution of drop sizes for fourteen sprinkler devices. They analyzed four models of impact sprinklers that were equipped with straight bore, flow control, or square nozzles. They also examined ten spray head devices equipped with jets that impinged onto fixed or moving plates. An exponential model was used for the drop size distribution. Li et al. (1994) had indicated that the exponential model was comparable in accuracy to the upper-limit lognormal model but much easier to use. The exponential distribution model is given by η ⎫ ⎧ ⎡ ⎛ d ⎞ ⎤⎪ ⎪ ⎟ ⎥⎬ Pv = 100⎨1 − exp ⎢− 0.693⎜⎜ ⎟ ⎢ ⎝ d 50 ⎠ ⎥⎦ ⎪ ⎪⎩ ⎣ ⎭

where Pv = percent of the total drops that are smaller than d d = drop diameter, mm

(16.14)

574

Chapter 16 Design and Operation of Sprinkler Systems

d50 = volume mean drop diameter, mm η = dimensionless exponent. Data were developed for the values of d50 and η for each nozzle and sprinkler combination for a range of operating pressures and nozzle sizes. The exponential model accurately represents the drop size distribution for drops larger than 3 mm and overestimated percentage volumes for smaller drop sizes. Kincaid et al. (1996) developed a procedure to estimate the parameters required for the exponential model. The procedure used the effective diameter of the nozzle and the nozzle pressure. They showed that the ratio of the nozzle diameter to the pressure head at the nozzle could be used to describe the volume mean drop diameter and the empirical parameter η. The relationships that they developed are given by d50 = ad + bd Ω and η = an + bn Ω

(16.15)

where Ω = ratio of the nozzle diameter to the pressure at the base of the sprinkler device ad, bd, an, bn = regression coefficients. Results were grouped into seven categories as describe in Table 16.7. These results provide a more general method to estimate the effects of the design and operation of irrigation systems on the drop sized generated. Kincaid (1996) presented data on the kinetic energy of spray droplets. Relationships were developed to predict the kinetic energy per unit mass based on the ratio of the nozzle diameter to the pressure head: ⎛ N e2 ⎞ Ek = e0 + e1⎜ de ⎟ + U 1.5 (16.16) ⎜H 3 ⎟ n ⎝ ⎠ where Ek = kinetic energy, kJ kg-1 Nd = diameter of the nozzle, mm Hn = nozzle pressure head, m U = wind speed, m s-1 e0,e1,e2,e3 = regression constants. Values for the regression constants are summarized in Table 16.8. Kincaid (1996) also correlated the kinetic energy to the mean volumetric drop size and the percentage of the drops that exceed a diameter of 3 mm:

Ek = 2.79 + 7.2 d50

and

Ek = 10.41 + 0.249 P3

(16.17)

where P3 = the percentage of the drops that are smaller than 3 mm. Table 16.7. Coefficients for estimating drop size distribution parameters. Type of Sprinkler Device ad bd an bn Impact sprinkler, small round nozzles (3-6 mm) 0.31 11900 2.04 -1500 Impact sprinkler, 1.30 2400 1.82 300 large round nozzles (9-15 mm) Wobbler 0.78 1870 2.08 -630 Rotator, 4 groove 1.07 3230 1.70 -830 Rotator, 6 groove 0.81 1480 2.07 -1300 Concave, 30 groove plate 0.82 620 2.68 -750 Flat smooth plate 0.66 680 2.74 -920

Design and Operation of Farm Irrigation Systems

Table 16.8. Coefficients for estimating energy from sprinkler devices. Sprinkler Device e1 e2 e0 Impact, large round nozzles 14.1 45.1 0.5 Impact, small round nozzles 6.9 132 0.5 Rotators, 4 groove plates 12.1 50.7 0.5 Rotators, 6 groove plates 8.9 38.0 0.5 Spinners, 6 groove plate 6.9 36.9 0.5 Low drift nozzle (LDN) 10.4 0.57 2 Medium groove stationary plate 6.2 0.45 2 Smooth groove stationary plate 5.4 0.40 2

575

e3 1.0 1.0 1.0 1.0 1.0 0.5 0.5 0.5

The wind speed has a major impact on the kinetic energy, increasing the energy per unit mass fivefold for wind speeds of 10 m s-1 compared to still air. Moldenhauer and Kemper (1969) showed that the infiltration rate decreased by one order of magnitude when the cumulative drop energy per unit area exceeded 500 J m-2. Spray devices that produce small drops with 5 J kg-1 could apply 100 mm of water before reduction occurs. For devices that produce larger drops with impact energies of 20 J kg-1 the threshold would be reached with 25 mm of water. These results allow designers to evaluate the potential runoff problems for various sprinkler devices. The concepts presented to this point can be applied to the design of sprinkler systems. 16.3.7 Sprinkler Placement Sprinklers must be properly placed and aligned. Occasionally sprinklers are not placed high enough to provide an unobstructed path for the sprinkler jet. The canopy that interferes with the water jet reduces the diameter of coverage and leads to poor water distribution. For row crops, sprinklers should be at least 0.5 m above the tallest mature crop that will be irrigated. For orchard crops the sprinkler should be placed to provide the wetted soil zone without causing degradation of fruit quality by wetting leaves and fruits. If the sprinkler is to be used for frost control, the sprinklers must be high enough to provide coverage of the crop canopy. On center pivots and lateral move systems the sprinkler devices may be placed below the lateral. The sprinklers should be placed so that the water stream does not impact on structural components of the machine. If the sprinkler devices are placed below the top of the crop canopy, they should be placed close enough along the pipeline to provide plants with equal access to water. This often requires that devices be positioned above alternate furrows. The vertical alignment of the sprinkler riser is also important. Nderitu and Hills (1993) showed that the diameter of coverage is reduced if the sprinkler riser is tilted from vertical. They showed that the precipitation profiles were nearly the same when the sprinkler riser was within 10° of vertical. However, the application uniformity decreased when the riser was tilted by 20°. Sprinklers that are firmly supported produced higher uniformities than unsupported risers.

576

Chapter 16 Design and Operation of Sprinkler Systems

16.4 APPLICATION UNIFORMITY 16.4.1 Single-Leg Distribution The uniformity of application from a sprinkler system depends on the distribution of water from individual devices. The distribution is measured with catch containers placed around a sprinkler as shown in Figure 16.10. The sprinkler is operated long enough to measure the depth of water applied at distances from the sprinkler. Measurement containers should be large enough and deep enough to provide an accurate measurement (Kohl, 1972). Fisher and Wallender (1988) showed that the measurement accuracy was directly related to the diameter of the catch container. If the containers are used in an open field, evaporation should be estimated. Various methods are available including using oil in the containers to suppress evaporation (see Heermann and Kohl, 1980). If the irrigation system will operate in calm and windy conditions, it is best to measure the distribution under both wind conditions. SPRINKLER LOCATION COLLECTOR LOCATION

AREA WETTED BY A SINGLE SPRINKLER

DEPTH OF WATER APPLIED

DIAMETER OF COVERAGE

ELLIPTICAL PATTERN

TRIANGULAR PATTERN

RADIAL DISTANCE FROM SPRINKLER

WETTED RADIUS

Figure 16.10. Arrangement to measure the single-leg water distribution of a sprinkler. Elliptical and triangular distribution patterns for the single-leg distribution are shown.

Design and Operation of Farm Irrigation Systems

577

The water application pattern about a single sprinkler is the single-leg distribution. The most commonly used equations are for a triangular and an elliptical distribution (Figure 16.10). Equations for the distributions are given by d (r ) =

3q S To πWr3

(Wr − r )

for a triangular pattern and (16.18)

d (r ) =

3q S To 2πWr3

Wr2 − r 2 for an elliptical pattern

where d(r) = the depth of water applied at a radial distance r from the sprinkler, Wr = the radius of coverage or wetted radius of the sprinkler To = the duration of the sprinkler operation. 16.4.2 Overlapping (Stationary Laterals) Several sprinklers should apply water to a location in the field to achieve an acceptable uniformity of application. The total depth of application can be estimated by overlapping the depth determined from the single-leg distribution for individual sprinklers. An illustration of the procedure is shown in Figure 16.11. This example is based on an impact sprinkler with 4.76 × 3.18 mm nozzles operated at 350 kPa as listed in Table 16.4. An elliptical distribution and an irrigation set time of 10 hours were assumed for the example. These conditions produce the single-leg distribution given by d (r ) = 3 16 2 − r 2 . A distribution of catch containers is assumed to compute the uniformity. In this case the containers are placed on a 3 m × 3 m grid. The first container is located half the grid spacing from the sprinkler. To compute the depth of water applied at each container the radial distance from a sprinkler to the container (r) is com2 2 puted by r = x + y where x is the horizontal distances from the lateral and y is the vertical distance to the sprinkler. The first four values for each container in Figure 16.11 are arranged by sprinkler starting with the upper-left sprinkler and continuing to the lower-right sprinkler. The top and bottom rows of containers receive water from upstream and downstream sprinklers that are not shown in the figure. The fifth and sixth rows of the data shown in the figure represent the contribution for those sprinklers. The total depth of water applied is shown as the last value in the column for each container. For this example the maximum depth applied was 145 mm while the minimum was 89 mm. The coefficient of uniformity was computed to be 87 using procedures described in Chapter 4. Overlapping provides a means to evaluate sprinkler spacing in the design of moved-lateral or solid set systems. Distribution data are available from the manufacturer or testing organizations for many sprinkler devices. Actual data from a single-leg test conducted for normal wind conditions provide more representative information for assessing the uniformity. The effect of overlapping applications can also be computed for moving systems. The procedure is somewhat more involved and is discussed in a later section for those systems.

578

Chapter 16 Design and Operation of Sprinkler Systems DISTANCE FROM LEFT LATERAL, m 0

1.5

4.5

7.5

10.5

13.5

16.5

18

DISTANCE FROM TOP SPRINKLER, m

0

48 0 37 0 27 0 112

46 27 35 0 24 0 132

43 37 30 20 15 0 145

37 43 20 30 0 15 145

27 46 0 35 0 24 132

0 48 0 37 0 27 112

46 0 43 0 89

45 24 41 15 125

41 35 37 30 143

35 41 30 37 143

24 45 15 41 125

0 46 0 43 89

7.5

43 0 46 0 89

41 15 45 24 125

37 30 41 35 143

30 37 35 41 143

15 41 24 45 125

0 43 0 46 89

10.5

37 0 48 0 27 0 112

35 0 46 27 24 0 132

30 20 43 37 15 0 145

20 30 37 43 0 15 145

0 35 27 46 0 24 132

0 37 0 48 0 27 112

1.5

4.5

12

SPRINKLER LATERAL

CATCH CONTAINER

CU = 87

Figure 16.11. Distribution of water from overlapping the single-leg distribution for each sprinkler at each container location.

16.4.3 Wind Effects Wind has a pronounced effect on the distribution of water from sprinklers. The work by Christiansen (1942), Vories and von Bernuth (1986), and Seginer et al. (1991a) and others showed how the application pattern of a single sprinkler is distorted in the wind. The general effects of wind are illustrated in Figure 16.12. The pattern is transported downwind as would be expected; however, the diameter of coverage perpendicular to the wind is reduced. The narrowing of the pattern perpendicular to the wind has an impact on the layout of sprinkler laterals. As illustrated the narrowing of the diameter of coverage can lead to poor uniformity if laterals are not placed closer together when the wind is parallel to the lateral. To compensate for the smaller diameter, sprinklers must be located closer together in the perpendicular direction. It is more economical to space sprinklers closer together along the lateral rather than space laterals closer together. Thus, the general recommendation is to orient laterals perpendicular to the prevailing wind and to space sprinklers closer together along the lateral than the spacing between the laterals. General guidelines have been developed for the maximum spacing to maintain acceptable uniformities (Table 16.9). Models have been developed to predict the distribution of water about a sprinkler (Vories et al., 1987). The models treat water droplets as ballistic objects and the equations of motion which include the effects of gravity and the drag on the drops are

Design and Operation of Farm Irrigation Systems

579

SPRINKLER

WIND DIRECTION Wd PATTERN WITHOUT WIND PATTERN WITH WIND LATERAL WIND DIRECTION

LATERAL PERPENDICULAR TO WIND

LATERAL PARALLEL TO WIND

Figure 16.12. Effect of wind on sprinkler distribution and resulting water application uniformity. Note that orienting laterals perpendicular to the prevailing wind direction is generally the most economical arrangement. Table 16.9. Maximum sprinkler and lateral spacing as a percentage of the effective wetted diameter for sprinklers operating at the average pressure along the lateral. Wind Condition Sprinkler Spacing Lateral Spacing No wind 45 65% 8 km h-1 40 60% 35 50% 8-16 km h-1 30 30% > 16 km h-1

solved. The increased drag resulting from high wind speeds causes droplets to move down wind. Seginer et al. (1991b) showed that the distribution could be simulated if the drag coefficient was adequately described. The formulation they used for the drag coefficient was sensitive to the type of sprinkler used. The adjustment to the drag coefficient requires outdoor measurements for acceptable accuracy. The coefficient of uniformity was quite sensitive to the adjustment made to the drag coefficient. Han et al. (1994) developed a mathematical model of the effects of wind on the single-leg distribution. They used an ellipse to describe the effect of wind on the horizontal shape of the distribution pattern and shape patterns across four principal sections of the pattern to predict the depth of application. They also conducted tests of

580

Chapter 16 Design and Operation of Sprinkler Systems

numerous combinations of sprinkler devices, nozzle types, operating pressures and wind speeds. While their model is more empirical than the work by Seginer et al. (1991b) it does provide a means to predict the three dimensional distribution of water about a sprinkler in windy conditions.

16.5 SOLID SET SYSTEMS One method to minimize labor and automate sprinkler irrigation is to permanently, or for a single season, install laterals at intervals across the field. This type of system is called a solid set system (Figure 16.13). Solid set systems are adaptable to a wide range of soils, crops, topography and field shapes. Being relatively expensive, they are commonly used for high-value crops (orchards, turf, and seedling establishment) to save labor and for environmental control. Solid set systems can be temporary or permanent. Temporary systems use aboveground aluminum or plastic laterals placed in the field at the beginning of the season, left in place for at least one irrigation and removed prior to harvest. Permanent systems use buried plastic, asbestos cement, or coated steel pipe for main and laterals with risers or riser outlets aboveground. Controllers allow complete flexibility in operation of solid set systems. Sophisticated solid state controllers are available to control normally open diaphragm valves on sprinkler laterals. Two types of controls are commonly used. For one method, lowvoltage (24 VAC) electrical solenoids are used as pilot valves to control water pressure to a diaphragm. The second method uses a hydraulic system with small tubing to supply air or water pressure directly to the diaphragm to close the valve. Solid set systems are frequently designed with a control valve at the inlet to the lateral. The valves are normally closed and are opened by supplying an electrical excitation for the length of time that water is to be applied. This allows each lateral to operate independently. A controller is used to determine how long each lateral is operated. When multiple laterals are operated simultaneously, the combination is called a circuit or a zone. Solid set systems provide excellent control of the amount of water applied. Some characteristics of solid set systems are summarized in Table 16.1. The disadvantages of to solid set systems include: ƒ The high cost to install and maintain—More laterals are required than for a periodically moved system, which increases costs substantially. The electronic valves and the controller also increase the costs. Costs increase substantially when the main line, submains and laterals are buried. All of the working parts also require maintenance for proper operation. ƒ The inflexibility due to installation of the system at specific locations in the field—If production practices change, such as changes in implement width or row spacing, it is very difficult to modify the layout of the solid set system to facilitate the new management practices. ƒ Its inconvenience to farm around—The systems are difficult to farm around unless perennial crops are planted and remain in the same location for prolonged periods. The advantages of the solid set system are that: ƒ The systems apply water uniformly across the field. ƒ The systems provide push-button control. ƒ Properly designed solid set systems can satisfy auxiliary needs such as frost control.

Design and Operation of Farm Irrigation Systems

581

Figure 16.13. Diagram of solid set sprinkler irrigation system.

16.5.1 Sprinkler Selection, Performance, and Spacing In general, solid set systems are designed to use low-flow, medium-pressure sprinklers. However, large sprinklers may be used if they are manually moved or individually valved. Sprinkler spacings will vary from 9 m by 9 m, to 73 m by 73 m. Nozzle sizes can be as small as 1.59 mm to as large as 36 mm, pressures range from 172 to 620 kPa. The sprinkler spacing depends upon the sprinkler and nozzle combination, operating pressure, desired coefficient of uniformity (CU), wind speed, and use of the system. For certain high-value crops, it may be desirable to design for a high CU. A crop of lesser value may not justify the cost of a design with a high CU. Since it is not possible to design for all wind conditions, the system should be designed for average conditions. A system designed for frost and freeze protection may not require as high a CU as a system for soil moisture control. A system that is used to supplement rainfall may not need as high a CU as one in which crop production totally depends upon irrigation.

582

Chapter 16 Design and Operation of Sprinkler Systems

16.5.2 Designing for Constant Sprinkler Discharge The general design procedure for solid set systems where sprinkler nozzle and pipe sizes are uniform has been described. As indicated there is variation of flow along the lateral. The following procedure can be used to design individual sprinkler laterals, or groups of laterals and associated main line or submain sections where there is less variation once the sprinkler spacing and discharges have been specified. The flow within any section of pipe can be determined. The procedure is iterative for selection of pipe sizes and input pressure. The procedure can be used to design complete systems or subsections of large systems and includes the following steps: 1. Assuming the desired sprinkler spacing has been selected, the first step is to lay out a system of laterals and main line on a topographic map of the field, or measure the elevation of each proposed sprinkler location, as well as the location and elevation of the system inlet. The elevation of each sprinkler outlet position is determined. 2. Calculate the flow in each pipe section as the total of all sprinkler downstream of that section. Flows need to be recalculated whenever the sprinkler flow or the number of operating sprinklers changes. 3. Pipe sizes are selected for main line and lateral sections. Initially, a large, uniform pipe size can be selected for main line and laterals, and then pipe sizes can be reduced in certain areas as the design is optimized. 4. Establish an assumed pressure head at an initial point in the system, the inlet being a convenient point. The initial pressure may be set low or equal to the minimum adequate sprinkler pressure. 5. Calculate pressures at all points by working upstream or downstream one pipe section at a time or by using the pressure distribution relationship previously presented. 6. Evaluate the pressure distribution. If some sprinkler pressures are inadequate, increase the input pressure and go back to Step 5 until the minimum sprinkler pressures are obtained. If all sprinkler pressures are within desired limits, the design may be acceptable, but some pipes may be oversized. If the range of pressures exceeds the desired limit, some pipes may be undersized, or elevation differences may be too large. 7. Reduce pipe sizes in selected areas, usually near the ends of laterals, or low elevation areas, and go back to Step 5. Repeat as necessary until pipe sizes and pressure distribution are optimized. When the pressure distribution has been calculated, the required nozzle size for each sprinkler can be calculated using the specified flow and calculated pressure. If the range of pressures is sufficiently narrow, one nozzle size can be used. Alternatively, pressure regulated sprinklers or flow-control nozzles can be used. If only a fraction of the laterals are to be operated at one time, the main line design should be checked with each set operating to ensure adequate pressure for all sets. The main line size(s) can usually be minimized by spreading the operating laterals as uniformly as possible over the whole main line. However, it may be desirable for cultural reasons to concentrate the operating laterals, in which case the set farthest from the inlet will usually dictate the main line design. The variability of topography within each group of sets must be considered. If elevation differences are great in the direction perpendicular to the lateral, pressure regu-

Design and Operation of Farm Irrigation Systems

583

lated sprinklers or flow-control nozzles are desirable. Also, if different numbers of laterals may be operated at different times, inlet pressure may change. The sprinkler flow, input pressure, pipe roughness, or any of the pipe sizes can be changed recomputed when using a spreadsheet. If the spacings are changed, the layout and thus the elevations would need to be changed accordingly. The laterals can be shortened by setting downstream pipe diameters to zero. In the case of uphill flows, large unavoidable pressure differences may be encountered. Pressure regulating valves can be located at lateral inlets or other points to reduce and limit the pressure to a specific value. Such controlled-pressure points can be used as the starting point for pressure calculations. Alternatively, individual pressure regulators can be used on sprinklers to limit the nozzle pressure and the inlet pressure can be increased to maintain the minimum pressure throughout the system. The following examples will illustrate the design procedure. Example 1. A 160-m × 180-m field is to be irrigated as shown in Figure 16.14. A spacing of 17 m along the main line and 15 m along the laterals is selected. The sprinkler flow is 0.8 L s-1. The inlet is at the pump, located one main line section upstream of the first lateral. The pump elevation is 10 m. Figure 16.15 shows the calculations using a spreadsheet program. Numbers in bold type indicate the required input data. Part A gives the pipe roughness (Hazen-Williams C), sprinkler flow, spacings, pump (inlet) pressure and elevation, total flow and computed average and percent variation in sprinkler pressure. WATER SOURCE AND PUMP LATERAL 1 2

SECTION

v

v

MAINLINE

3

4

5

6

7

8

9

10

v

v

v

v

v

v

v

v

VALVE

1 2 3

9

4 5

8

6

7 7 8

6 5

9 10

SPRINKLER

4 ELEVATION CONTOUR, m

3

FIELD BOUNDARY

Figure 16.14. Plan view of field irrigated with a solid set sprinkler system.

584

Chapter 16 Design and Operation of Sprinkler Systems

Figure 16.15. Design of mainline and laterals for a solid set system with 10 laterals, where: Pm is the pressure head at the main line outlet or lateral inlet, Pmin is the minimum pressure and DP is the pressure difference on a lateral, Ns is the number of sprinklers on a lateral (ON:OFF = 0,1), Ql is the total lateral flow, Qtot is the total flow in the mailine section, and EL is the elevation of the inlet of a main line section (m).

Design and Operation of Farm Irrigation Systems

585

Part B gives the number of sprinklers or pipe sections on each lateral, elevations and diameters of the main line section upstream of each lateral, and computed main line pressures, minimum lateral pressures, pressure difference on each lateral, flow in each main line section and lateral, and main line velocities. Laterals are designated as being on or off by entering 1 or 0 in the ON:OFF column (Figure 16.15). Individual laterals can be turned off to simulate smaller sets or movable lateral systems. Parts C to F give the elevations, pipe diameters, sprinkler pressure heads, and computed nozzle sizes in grid form corresponding to the grid layout of Figure 16.15. The sprinklers are located at the downstream end of each lateral section. Following the above procedure, the elevations and pipe sizes were input (Parts C and D). The flows in the main and laterals are computed by summing the sprinkler flows downstream of each section. The assumed input pump pressure of 50 m becomes the inlet pressure for the first mainline. All 10 laterals are to be operated simultaneously, so the flow in the first section is 80 L s-1. The diameter is 200 mm, and friction loss is 0.6 m. The outlet pressure for the first section is 50 – 0.6 + 10 – 9.9 = 49.5 m, which becomes the inlet pressure for the second main line section and also the inlet pressure for the first lateral. Pressures were calculated down the first lateral one section at a time, and the minimum was found to be 39.9 m (in section 8 since the laterals are running down slope). Successive main and lateral sections were calculated to the end of the last lateral. The overall minimum pressure head was 39.6 m in lateral 2. Pipe size selection and readjustment is the main iterative process in this procedure. In this example pipe sizes were adjusted so the pressure difference within the laterals is less than 20% of the minimum pressure. The pipe size on the first section of the first 2 laterals was reduced to decrease the lateral pressures and thus reduce the overall pressure variation to less than 20%. The computed nozzle sizes are thus nearly uniform, and one nozzle size can be selected. If variable nozzle sizes are needed to maintain uniform flows, the designer can select the available nozzle size closest to the computed size. Example 2. This example uses the same field and lateral layout in Example 1, operated by running two adjacent laterals per set. Results are given in Figure 16.16 for laterals 9 and 10, the laterals farthest from the pump. The total flow at the pump decreases to 16 L s-1. The main line sizes are reduced accordingly. The lateral pipe sizes are the same as in Example 1. The average pressure, Pav, and variation Pvar, are for the operating laterals only.

16.5.3 Operation and Maintenance Guidelines The operating mode for a solid set sprinkler system depends upon the design and use of the system, available labor, water supply, and available capital. The system can be designed using the lateral or area (zone) design method. With lateral design individual laterals are controlled by valves and each lateral may be operated as desired. Normally, more than one lateral is operated simultaneously, but the operating laterals usually are widely separated in the field. The lateral design method minimizes the main or supply line pipe size, but it increases the number of valves required and also

586

Chapter 16 Design and Operation of Sprinkler Systems

Figure 16.16. Design of mainline and laterals for a solid set system with 2 laterals o perating, where Pm is the pressure head at the main line outlet or lateral inlet, Pmin is the minimum pressure and DP is the pressure difference on a lateral, Ns is the number of sprinklers on a lateral (ON:OFF = 0,1), Ql is the total lateral flow, Qtot is the total flow in the mailine section, and EL is the elevation of the inlet of a main line section (m).

Design and Operation of Farm Irrigation Systems

587

the time to open and close valves when a manual valve system is used. With the area (or zone) design method, a contiguous portion of the field is irrigated at one time. Usually a submain is installed to supply water to that portion of the field. For frost and freeze protection, the entire system may be operated at one time. Depending upon the crop being protected, the application rate will be 2 to 5 mm per hour. In the eastern U.S., most orchard systems are designed to apply water over the crop. In the western U.S., both undertree and overtree systems are used; however, with saline water only undertree systems can be used successfully. If the system is being used strictly for irrigation, only a portion of the system is normally operated at one time. Where several hours are required for irrigation, control may be manual or automatic. For an irrigation system on a shallow-rooted crop grown on a coarse-textured soil, or in a container nursery operation where daily or frequent irrigation is required, it is best to automatically control the sequencing of the system. Where labor is very limited, automatic control may be desirable regardless of irrigation frequency, however this will increase the initial investment. Conversely, limited capital may require a totally manual system. A limited water supply, such as a well or stream, may mean that only a portion of the system can be operated at once.

16.6 PERIODICALLY MOVED LATERALS Sprinkler systems in this category have laterals that are moved between irrigation settings. They remain stationary while irrigating. The lateral is drained prior to moving to the next set. Set-move laterals are used extensively because of their relatively low cost and adaptability to a wide range of crops, soils, topographies and field sizes. Equipment cost is largely dependent on the number of sets irrigated by each lateral. They are well suited to soils with high water holding capacity, deep-rooted crops, lowgrowing crops, supplemental irrigation, and deficit irrigation management. They can be classified as hand move or mechanical-move systems. Mechanical-move systems are similar to hand move except that pipe types and sizes are often dictated by mechanical rather than hydraulic considerations. The system consists of laterals, a pipeline with outlets for distributing irrigation water to sprinkler devices that are periodically moved across the field. The lateral is made of pipe sections that are from 50 to 150 mm in diameter and 6 to 18 m long (Figure 16.1). A coupler is installed on one end of the pipe section. The other end of the pipe section is inserted into the upstream coupler and fastened with either a hook that latches into the coupler or a ring assembly. New types of couplers are currently being developed that use different mechanisms to connect the sections of pipe. Gaskets are installed in the coupler to prevent leaks when the system is pressurized. Small pipes, called risers, convey water from the lateral to the sprinkler. Water is supplied to the laterals by main lines, or submains which branch from the main line. The most common system has a single center main line with one or more laterals which irrigate on both sides of the main line. If there are multiple laterals, they are spaced equally, so by the time any lateral reaches the starting position of the lateral ahead of it, the entire field has been irrigated once. The spacing of the sprinklers on the laterals and between subsequent sets of each lateral is such that the water distribution patterns from the sprinklers give almost complete overlap. Large systems often require more complex designs with multiple main lines, although simple systems are possible on rectangular fields up to at least 64 ha.

588

Chapter 16 Design and Operation of Sprinkler Systems

Valve-tees are usually placed in the main line at the desired interval for spacing between lateral settings. The valve-tees are controlled by manipulating a valve-opening elbow that makes the connection between the main line and laterals. Common valve spacings are 12.2, 15.2, 18.3, and 24.4 m. Since common pipe lengths are 6.1, 9.1, 12.2, and 15.2 m, the desired valve spacing is obtained by using various combinations of lengths. For many systems buried main lines are preferred. The pipe should be placed safely below plow depth and should also be below frost depth, unless provisions are made for draining the pipeline. The advantages of periodically moved systems are that: ƒ investment costs are minimal; ƒ the systems offer a great deal of flexibility; ƒ the systems are easy to understand and operate; and ƒ the sprinklers and nozzles are generally the same size, which maximizes interchangeability. The disadvantages of periodically moved systems are: ƒ the high labor requirements; ƒ the relatively large applications of water each irrigation; ƒ the reduced uniformity when a uniform size of nozzle is used on long laterals or rough terrain; and ƒ the time required for laterals to drain before moving. The water application and other characteristics of the hand move, towline and side roll systems are very similar. Characteristics of the system are presented in Table 16.1. 16.6.1 Hydraulics The hydraulic design of moved laterals has been described in Chapter 15 and previously in this chapter. With these systems the sprinkler and nozzle sizes are generally constant and the pipe diameter for the lateral is uniform. Thus, the procedures developed in earlier sections apply. The general procedure is to layout the field boundaries, water supply and pump location. The main line is frequently laid down the center of the field to minimize pressure losses in long laterals. The spacing between sprinklers and laterals is selected to fit the field. With an initial layout the discharge for the sprinkler and lateral are determined. The minimum average pressure is determined for the selected size of lateral pipe. From these data the nozzle sizes can be determined. The diameter of coverage of the sprinkler must be large enough to provide adequate overlap. If all components are adequate the inlet pressure for each lateral position should be determined. The main line should be selected to provide the highest feasible uniformity. The cost of design alternatives should be determined. For the final design, product specifications and an operational plan should be developed and discussed with the client. Special considerations for each type of periodically moved system are discussed below. 16.6.2 Hand Move Laterals The earliest periodically moved systems involved laterals that were moved by carrying sections of pipe across the field. Between moves the lateral operates for a period time (the set time) and applies water to a portion of the field (a set) (Figure 16.17). This is called a hand moved system. An extensive amount of labor is required to move laterals from one set to the next set, which discourages frequent movement resulting in large applications of water per irrigation.

Design and Operation of Farm Irrigation Systems

589

Figure 16.17. Picture and operational sketch of a hand move irrigation system. Common pipe diameters range from 51 to 152 mm, and pipe lengths are 6.1, 9.2, and 12.2 m. The most popular aluminum lateral length is 9.1 or 12.2 m. Shorter lengths mean more walking during the move. Longer lengths are more difficult to transport and do not provide proper spacing for the common sprinkler sizes.

Hand move systems are the lowest equipment cost alternative for moved-lateral systems, but are labor intensive. Lack of available labor is the main reason farmers are tending toward center pivot or other automated systems. Most hand move sprinkler systems now use aluminum laterals, although plastic is available. Most lateral pipe couplers contain a chevron-type rubber gasket which seals under pressure, and are designed to latch automatically when the pipes are pushed together. Many couplers contain an optional adjustment for easier unlatching. Hook and latch type couplers can be fixed so they will unlatch automatically when the irrigator pushes and twists the pipe. Ball and socket couplers automatically latch when the pipe is under pressure, and

590

Chapter 16 Design and Operation of Sprinkler Systems

release when the pressure is off. Drop-lock couplers have hooks that engage as the pipe is lowered to the ground. The automatic unlatch saves some walking, but could increase the hazard of an accidental unlatching. In contrast to the main line gaskets, the lateral gaskets are designed to release their tight contact with the pipe when the pressure on the water is reduced. This permits the water to drain from the pipe when the pressure has been turned off so the pipe can be moved easily to the next setting. Each coupler is threaded to receive a sprinkler riser pipe, usually 25 mm in diameter. If both the coupler and the riser are aluminum, it is customary to connect them with a zinc alloy fitting or Teflon tape to avoid thread seizure. The riser should extend at least to the top of the crop canopy, but the uniformity of water distribution is improved if it is extended another 0.5 m. The uniformity of water distribution can usually be improved by using an offset pipe with a 90° elbow every second irrigation. The length of the offset should be half of the spacing between lateral settings. Using the offset pipe permits the lateral to be placed midway between the positions used during the previous irrigation. Thus, considering two irrigations added together, a 12.2-m by 18.3-m spacing, for example, is effectively reduced to 12.2 m by 9.2 m. A good procedure for the irrigator to follow when moving the lateral from one setting to the next is to start by moving the valve opening elbow and the section of pipe connected to it. As soon as these pieces are in place at the new location, the valve is slightly opened so a very small stream of water runs out the end of the first pipe section. As each subsequent section of pipe is put into place, the small stream of water runs through it, flushing out any soil or debris that may have been picked up during the move. The last section of pipe with its end plug in place can be connected before the stream of water reaches the end and builds up pressure. Then the irrigator walks back along the lateral, correcting any plugged sprinklers, leaky gaskets, or tilted risers. After returning to the main line, the valve is opened further until the desired pressure is obtained. A quick check with a pitot gauge on the first sprinkler confirms the valve adjustment. To save time on each lateral move, there is a tendency to completely open the valve and fill the line as quickly as possible. This causes water hammer at the far end of the line, so a surge plug at that end may be needed. The sprinklers commonly used on hand move systems may have either one or two nozzles. Typically, individual sprinkler capacities range from about 0.06 to 0.63 L s-1. Operating pressures range from 240 to 415 kPa. Certain crops, such as orchards, require specially designed sprinklers. When the sprinklers are used over the tops of the trees, conventional models can be used. However, when they are used under trees, sprinklers with low water trajectory must be used. Lowering the trajectory reduces the uniformity, unless the spacing is reduced. Hedge-rowed trees present a difficult problem, especially if it is desirable to irrigate through the skirts of the rows. 16.6.3 Towline Systems Towline or skid-tow systems were developed to reduce labor required to reposition laterals. The most efficient layout for towlines is to divide the field in half so the lateral can be pulled in a zigzag fashion across the field (Figure 16.18). Towline sprinkler laterals have relatively rigid couplers fitted with skids or wheels so the line can be moved by pulling it from the end. The skids consist of a flat metal plate held on the

Design and Operation of Farm Irrigation Systems

591

Figure 16.18. Picture and operational sketch of a towline or skidtow system.

underside of the pipe by one or more clamps. In one type, the skid is placed under the coupler and clamped at both ends. This makes the skid take the major part of the end thrust at the coupler when the pipe is towed. If relatively long sections of pipe are used, a second skid may be needed under the center of each section to reduce abrasion from soil contact. Stabilizers, outriggers, or wheel-mounts were used to keep the lateral from tipping. Two or three outriggers along the line are needed to keep the pipe oriented with the skids on the bottom and the sprinklers upright. For wheel-type units, a pair of wheels mounted on a simple U-frame is clamped to each section of pipe. The wheels are oriented so the entire length of lateral pipe can be pulled endwise. The pipe itself stands only 0.3 to 0.5 m above the ground. The flexibility of the pipe and the articulation of the couplers permit the lateral to curve slightly while being moved to a new position. In one type, however, the lateral stays straight. The wheels are fixed so they shift to a 45° angle from the lateral when pulled in one direction, and then shift back to a 45° position the other way when pulled from the

592

Chapter 16 Design and Operation of Sprinkler Systems

other end of the lateral. Thus, by pulling alternately from both ends, the entire length of lateral is shifted the desired distance to the next set position.A coupler and hitch are attached to each end of the lateral so that it can be towed in either direction. An end cap is used to plug the downstream end of the pipeline. Drain plugs are installed along the lateral to empty water from the pipeline before moving. Often a flexible hose is used to connect the lateral to the main line. If more than one lateral is used in a field, provisions need to be made to split the main line while moving the lateral. A telescoping coupler can be used for this purpose. When the lateral reaches the edge of the field, the lateral must be disassembled and transported to the starting location. If the surrounding property is amenable, the lateral can be towed to the starting position. The traditional way of moving a towline lateral is to snake it past the main line in an S-shaped curve to a new position on the other side (Figure 16.18). For the next setting, the lateral is pulled in the other direction past the main line in an opposite Sshaped curve. With this procedure, each move needs to advance the lateral only half the distance between adjacent sets. Towline systems are the least expensive of the mechanically moved systems. However, they are not used extensively because the moving process is tedious, requires careful operation, and damages many crops. Towline systems have been used successfully in some forage crops and in row crops. The moves are made easier if the main line is buried. 16.6.4 Side Roll Systems The side roll, or wheel-move, system is a third type of periodically moved lateral. With this system, wheels are mounted on the sprinkler lateral to carry the pipeline above the crop (Figure 16.19). A cart provides power to rotate the wheels. The cart and the water feed to the side roll can be positioned anywhere along the lateral. Frequently the pipeline is used as the axle for the torque; however, a separate drive shaft can be used to rotate the pipe. Multiple laterals are often used in a field. A special apparatus with a swivel and weight is used to keep sprinklers vertical when the pipeline rotation is not precise. The mainline may be above or below ground. Flexible hoses are often used to connect the side roll lateral to the mainline. Rigid couplers permit the entire lateral to be rolled forward by applying torque at the center while the pipe remains in a nearly straight line. Aluminum pipe having a 100 or 125 mm diameter is commonly used. To have sufficient strength, the aluminum pipe wall thickness should be at least 1.8 mm. A motorized drive unit, usually near the center of the lateral, provides torque to move the lateral and holds the pipe in place during operation. Normally, the drive unit contains a gasoline engine and a transmission with a reverse gear. Electric motors or hydraulic motors are also used. Typical lateral length is 400 m, but longer lengths are made with two drive units spaced about ½ the lateral length apart, and connected by a drive shaft. Since the greatest torque is applied to the pipe near the drive unit, 125-mm pipe is sometimes used near the center of the lateral for greater strength. The pipe is flexible enough so that these systems can be used on rolling topography with mild slopes. The most popular sprinkler spacing (and pipe length) is 12.2 m. The wheels are usually placed in the center of each length of pipe, with sprinklers located halfway between the wheels. Thus, a standard 400-m lateral contains 32 pipe lengths and 36 wheels because four wheels are required for the drive unit. Sometimes an extra wheel is provided for the last pipe section at each end.

Design and Operation of Farm Irrigation Systems

593

Figure 16.19. Picture and operational sketch of a side roll irrigation system.

The rigid couplers can be quickly disconnected to shorten the lateral in the case of odd shaped fields. Often the sprinklers are provided with self-levelers so they will right themselves if the lateral is not stopped where the riser would be exactly upright, i.e. on variable topography, and to aid in aligning the laterals with main line valves. Labor requirement is approximately five minutes per lateral per move. Some side roll drive units can be controlled from the end of the lateral, eliminating the need to walk to the center drive unit. At least one manufacturer has developed an automated side roll system that can be programmed to drain and move itself and irrigate up to five sets. Water is supplied through a flexible hose. The wheel diameter must be large enough so the pipe will pass over the crop without damaging it, and the crop will not prevent the lateral from being rolled to the next position. Common wheel diameters commonly are 1.17, 1.47, 1.63, and 1.93 m. A spring-loaded drain valve also is located about midway between the wheels, near the pipe coupler and near the sprinkler. This valve opens automatically when the pressure is off, so the pipe will drain quickly and permit moving the lateral forward to the next set without much time loss. (Attempting to roll the pipeline when it is full of water will damage the equipment).

594

Chapter 16 Design and Operation of Sprinkler Systems

The most popular side roll spacing along the main line is 18.3 m. Two popular operating schemes are used. In one, the lateral is connected to every outlet valve along the main line, and when the lateral reaches its destination and completes its last set, it is rolled back to the starting point. In the other, the lateral is connected to every evennumbered outlet valve on the main line while the lateral is moved across the field, and then connected to the odd- numbered valves while the lateral is moved back to the starting position. For the latter case the lateral interval between irrigations is longer at the two ends of the field than in the center. As with hand move laterals, there is a tendency to completely open the hydrant valve and fill the line as quickly as possible, causing a water hammer at the far end of the line. Therefore, a surge plug at the closed end is recommended. The use of offsets, especially for the 12.2 m by 18.3 m spacing, is also recommended. Side roll laterals are highly susceptible to wind damage when they are empty, and should be staked down during the off season. Special braces, which allow the lateral to roll in one direction only, help protect the laterals during the irrigation season. A lateral with 32 sprinklers is commonly designed with 100-mm diameter pipe, even when the water is introduced into it from one end and the friction loss is 55 to 60 kPa. However, if the water is admitted to it at the center of the lateral, the friction loss is reduced to about 1/5 as much. A 125-mm pipe would have only about 1/3 the friction loss of the 100-mm pipe when the water is admitted from one end. The best method will depend on the future price and availability of energy. End-feed laterals are usually preferred because a drive-through roadway can be maintained along the main line for easy accessibility to the valves. 16.6.5 Side Move with Trail Lines Side move laterals with trail lines are supported on wheel-mounted A-frames. The pipe does not serve as the axle of the wheels and can be higher above the ground. Each A-frame carriage is driven from a drive shaft that extends the length of the pipeline. The drive shaft can be turned from the center of the line or from one end. One version uses a continuous-move (center pivot) type lateral operating in stationary set-move mode. The wheels are powered by electric or hydraulic motors supplied from an onboard generator or hydraulic pump. The small diameter trail lines can each carry several sprinklers. Usually, short sprinkler risers are used as they are easier to keep upright than tall risers. Outriggers at the last sprinkler on each trail line are used to keep the risers upright; however, these may damage some crops. This system is sometimes called a “movable solid set.” It greatly reduces the number of moves necessary to cover the field, thus saving labor. Sprinkler spacings and nozzle sizes that give low application rates at an acceptable uniformity can be used. With the low rates, 24-h set times may be practical for some soils and crops, thus permitting a normal daytime work schedule for the irrigator. When a trail line system reaches the end of the field, the trail lines are uncoupled, and the lateral is moved to the opposite ends of the trail lines, which are then recoupled to irrigate on the way back across the field. The wheels on most side move systems can be turned 90°, permitting the lateral to be pulled endwise to another field.

Design and Operation of Farm Irrigation Systems

595

16.7 CENTER PIVOTS In 1948 Frank Zybach invented the self-propelled irrigation system. His invention used towers with wheels to carry a pipeline around a pivot point in the field. Even though his invention has undergone numerous changes, the basic concept is still used. A span of pipe is supported by a tower and a truss system (Figure 16.20). Today most systems are propelled by electrically or hydraulically powered motors mounted on each tower. A system of switches on each tower energizes the motor when the tower needs to move. The depth of water application is generally controlled by selecting the speed of the last or end tower. For many electric systems a one-minute timer is used to control the velocity. If the timer is set to 100%, the motor on the end tower is energized the entire minute causing the end tower to move at a constant velocity equal to the maximum speed of the system. If the timer is set to 50% the motor is only energized for 30 seconds; thus, the end tower is stationary for 30 seconds and then moves for 30 seconds at the maximum speed for the end tower. Some hydraulic and electric systems provide continuous movement of the end tower at variable speeds to apply the desired depth of application. The interior towers are controlled by switches or valves mounted on the tower. One switch or valve is set to energize the motor if the downstream tower has moved far enough to exceed a start angle. The switch or valve energizes the motor and causes the tower to move at a constant velocity. The tower moves until the angle between adjacent spans exceeds a stop angle. For continuously moving systems the interior towers move continuously at varying speeds to maintain alignment. The interior towers may be designed to move at a faster velocity than the end tower allowing them to catch up with the end tower to maintain alignment. Controllers are presently available to change the speed and/or direction of rotation of the pivot lateral as the system circles the field. This is useful if different crops are planted under a pivot or if obstructions are located in the field. Small center pivots are also made that can be moved within a field, or from field to field. This allows for semi-automated irrigation of irregularly shaped fields and small tracts of land. Center pivots can be equipped with an end gun to increase the portion of a field that is irrigated (Figure 16.20). The end gun is a large sprinkler similar to that used on a traveler that is mounted on the end of the pivot lateral. The gun throws water a long distance thereby increasing the amount of land irrigated for a given length of lateral. A valve is attached to the end gun so that the end gun only operates in the corners of the field. When the pivot lateral reaches a preset angle of rotation, the valve opens and water is supplied to the end gun. In some cases a booster pump is attached to the valve to increase the pressure for the end gun. A corner system can be used to irrigate an even larger portion of a square field. A special span is attached to the end of a typical system (Figure 16.21). The corner span revolves about the end of the pivot lateral. The corner span is tucked behind the main lateral when the boundary of the field is close to the end of the pivot lateral. The sprinklers on the corner span begin to irrigate when the pivot rotates to an angle where the sprinkler package on the primary lateral does not reach to the field boundary. Sprinklers on the corner span are attached to a series of valves. As the pivot lateral rotates toward the corner of the field, the corner lateral extends and opens valves as the corner lateral is extended. An end gun can be attached to the corner lateral to throw

596

Chapter 16 Design and Operation of Sprinkler Systems

Figure 16.20. Components and field layout for typical center pivot irrigation systems.

water further into the corner. Many corner systems use an underground cable and an antenna on the corner tower to follow a desired path. The cable can be positioned to irrigate irregular shapes and to move around permanent obstructions. Center pivots have many advantages including: ƒ Automated operation—Center pivots can be operated with minimal labor often making several revolutions without stopping. They can also be controlled remotely from farm vehicles or computers. ƒ Ability to apply small irrigation depths—Since the systems are automated, they can apply small irrigations to match crop needs without leaching nutrients. ƒ Very high uniformity—Since the lateral moves slowly and because there is a great deal of overlap of water application from successive sprinklers on the lateral, center pivots apply water very uniformly. ƒ Chemigation—Pivots can be managed to quickly and uniformly irrigate a field with small amounts of water, which provides the opportunity to apply fertilizer, herbicides, and insecticides.

Design and Operation of Farm Irrigation Systems

597

PRIMARY PRIMARY LATERAL LATERAL CORNER CORNER SPAN SPAN

CORNER CORNER TOWER TOWER FIELD BOUNDARY

+ UN M G TE ND S E SY D IC AN S M BA R S A IU ER D N RA OR C

RADIUS OF BASIC SYSTEM

Figure 16.21. Picture and operational sketch for a center pivot equipped with a corner watering system.

ƒ Little annual setup is required—Once the system is constructed, the pivot can be operated anytime that water is needed. This is advantageous for germination of crops or for preparing seedbeds. Some disadvantages of center pivot systems are its: ƒ Cost—Depending on the reference, the cost of a center pivot system is moderately high. The cost per unit of land for a typical center pivot is approximately 20 to 30% of that for solid set and microirrigation systems. However, the cost exceeds that for systems requiring more labor such as hand move or towline systems. The cost per unit area decreases as the length of the lateral increases; however, several factors limit the maximum length of the lateral. ƒ High application rates—The rate of water application at the outer end of the lateral is quite high, which can cause runoff.

598

Chapter 16 Design and Operation of Sprinkler Systems

ƒ Circular pattern—Center pivots irrigate about 80% of a square field. Lost production in the corners may be a consideration when land is expensive or highvalue crops are produced. The self-propelled irrigation system as named by inventor Frank Zybach in 1948 has transformed irrigation. Many lands are now efficiently irrigated that were once labeled as unsuitable for irrigated agriculture. Center pivot systems as developed by the irrigation industry have reduced labor requirements, improved water application efficiency and contributed to the economic viability of many regions. Nevertheless there have been failures with center pivots. Problems have arisen due to poor design and improper management. In some locations excessive development has led to the overdraft of water resources causing declines in ground water supplies. Without proper management leaching of agricultural chemicals can occur especially under shallow rooted crops grown on sandy soils. Other chapters discuss the planning and operation of irrigation systems to fit within environmental, economical and resource constraints. The focus in this section is on the design of center pivot systems to efficiently and uniformly apply irrigation water. 16.7.1 Sprinkler Discharge The discharge from each sprinkler on a center pivot must be determined to apply water uniformly. The area midway between the adjacent upstream and downstream sprinklers defines the representative area for a sprinkler on a pivot (Figure 16.22). FIELD BOUNDARY

RS

R

SL

SPRINKLER SPACING

SL

SL

PIVOT LATERAL

RADIAL DISTANCE TO SPRINKLER (R)

SL R – SL / 2

R + SL / 2

Figure 16.22. Diagram of representative area for determining the discharge for sprinklers on a center pivot lateral.

Design and Operation of Farm Irrigation Systems

599

This area is area is given by

AR = 2π R SL

(16.19)

where AR = representative area for the sprinkler at distance R from the pivot point SL = local spacing between sprinklers on the pivot lateral, m. It is common for the spacing between sprinklers to vary in intervals along the pivot lateral. Often the spacing near the center of the pivot is double the spacing near the outer end of the pivot. The average discharge per unit area for the pivot is the ratio of the flow into the pivot system divided by the circular area irrigated by the primary system (i.e. where the end gun or corner system is not operating). Combining these relationships provides the discharge required for the sprinkler located at radial distance R from the pivot point: 2QS RS L qR = (16.20) Rs2 where qR is the discharge required for a sprinkler located a distance R from the pivot point and QS is the total discharge into the center pivot system. The expression can be revised to include the gross system capacity as

qR = 2 × 10-4π Cg R SL

(16.21)

To design the sprinkler package the nozzle sizes for each sprinkler must be determined. This requires information on the distribution of pressure along the lateral. For level land the relationship developed by Chu and Moe (1972) can be used for the distribution: 3 5 ⎫⎤ ⎡ ⎧ 15 ⎪ R 2 ⎛ R ⎞ 1 ⎛ R ⎞ ⎪⎥ ⎟⎟ ⎬ ⎟⎟ + ⎜⎜ PR = PS + PL ⎢1 − ⎨ − ⎜⎜ (16.22) ⎢ 8 ⎪ Rs 3 ⎝ Rs ⎠ 5 ⎝ R s ⎠ ⎪⎥ ⎩ ⎭⎦ ⎣ where PR = the pressure in the lateral at point R PS = pressure at the distal end of the pivot lateral PL = pressure loss from the inlet into to the distal end of the pivot lateral. Keller and Bleisner (1990) and Scaloppi and Allen (1993) presented methods to compute the pressure distribution along pivot laterals. Their results produce the same pressure distribution along the lateral as that from Chu and Moe (1972). The relative distribution of pressure loss along the lateral is shown in Figure 16.23. The difference between the methods by Chu and Moe (1972) and Scaloppi and Allen (1993) is in the procedures used to compute the friction loss from the inlet to the distal end of the lateral (i.e., PL). Chu and Moe used the Hazen-Williams equation. Scaloppi and Allen used the Darcy-Weisbach equation. Scaloppi and Allen considered the variation in the velocity head in the lateral as well while Chu and Moe considered it to be negligible. In Chapter 15 the Darcy-Weisbach method was recommended for hydraulic calculations. We concur with that recommendation for applications using computer spreadsheets and/or programs to compute friction loss. Applying the DarcyWeisbach method for sprinkler laterals results in changing friction factors along the pipeline. This is difficult to represent in this chapter. Thus, the hydraulics for center pivot pipelines is illustrated using the Hazen-Williams equation for this chapter.

600

Chapter 16 Design and Operation of Sprinkler Systems FRACTION OF PRESSURE LOSS ALONG LATERAL

1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0 0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

FRACTION OF DISTANCE ALONG LATERAL

Figure 16.23. Distribution of pressure loss along a center pivot lateral.

The pressure loss along pivot laterals is traditionally computed by the industry using the Hazen-Williams equation (see Chapter 15). Typical C values for center pivot materials are provided in Table 16.10. Since the discharge from sprinklers varies along the lateral for a center pivot the F value for pivot laterals is different than for other sprinkler laterals. Results from Chu and Moe (1972) show that the F value for pivots without end guns is 0.54. Pair et al. (1983) show that the F value is 0.56 when the end gun is operating. These values have been compared with a procedure that computes the pressure loss between each sprinkler along the pivot lateral as used by Kincaid and Heermann (1970). Comparisons show that the method by Chu and Moe (1972) is sufficiently accurate. Based on these procedures the pressure distribution along pivot laterals can be computed using Table 16.10. These results provide the friction loss along laterals that only have one diameter of pipe. Figure 16.23 shows that over half of the total loss along the lateral occurs within the first third of the lateral and that 80% of the pressure loss in the lateral occurs in the first half of the pipeline. Experience has shown that operating pressures and therefore operating costs can be reduced by using larger diameter pipe for the first portion of the lateral. Computing the friction loss for laterals that contain two diameters of pipe can be computed for R ≥ Rc as ⎡ ⎧⎪ R 2 ⎛ R ⎞3 1 ⎛ R ⎞5 PR = PS + PLs ⎢1 − 1.875 ⎨ − + Rs 3 ⎜⎝ Rs ⎟⎠ 5 ⎜⎝ Rs ⎟⎠ ⎢ ⎩⎪ ⎣

and for R ≤ Rc:

⎡ PC = PR + PLs ⎢1 − 1.875 ⎢ ⎣ ⎡ PR = PC + PLL ⎢ 1.875 ⎢ ⎣

⎧⎪ Rc 2 ⎛ Rc ⎞3 1 ⎛ Rc ⎞5 − + ⎨ Rs 3 ⎜⎝ Rs ⎟⎠ 5 ⎜⎝ Rs ⎟⎠ ⎩⎪

⎫⎪ ⎤ ⎬⎥ ⎭⎪ ⎥⎦

(16.23)

⎫⎪ ⎤ ⎬⎥ ⎭⎪ ⎥⎦

⎧⎪ R ⎛ Rc3 R 3 ⎞ 1 ⎛ Rc5 R5 ⎞ ⎫⎪ C − R − 2 ⎜ 5 − 5 ⎟⎬ ⎜ 3− 3⎟ + ⎨ ⎜ ⎜ ⎟ 3 ⎝ Rs Rs ⎠ 5 ⎝ Rs Rs ⎟⎠ ⎪ ⎪⎩ RS RS ⎭

⎤ ⎥ ⎥ ⎦

Design and Operation of Farm Irrigation Systems

601

Table 16.10. Friction loss (m/100 m) in galvanized steel pipe for center pivot and lateral move systems. Results assume a C value of 140 for the Hazen-Williams equation and a single pipe size. To compute losses for a pivot multiply values times the lateral length and the appropriate F factor. System Outside Diameter of Pipe (mm) Discharge 102 114 127 141 152 168 178 203 219 254 (L s-1) 5 0.57 10 2.06 1.12 15 4.36 2.38 1.39 20 7.42 4.06 2.37 1.38 25 11.2 6.13 3.58 2.09 1.42 30 15.7 8.60 5.02 2.92 1.99 1.21 35 20.9 11.4 6.68 3.89 2.65 1.61 1.22 40 26.8 14.6 8.56 4.98 3.40 2.06 1.56 45 33.3 18.2 10.6 6.19 4.22 2.56 1.94 1.00 50 22.1 12.9 7.53 5.13 3.12 2.36 1.21 60 31.0 18.1 10.6 7.20 4.37 3.31 1.70 1.16 70 24.1 14.0 9.57 5.81 4.41 2.26 1.55 80 30.9 18.0 12.3 7.44 5.64 2.89 1.98 90 22.4 15.3 9.25 7.02 3.59 2.47 1.18 100 27.2 18.5 11.3 8.53 4.37 3.00 1.43 120 38.1 26.0 15.8 12.0 6.12 4.20 2.01 140 34.6 21.0 15.9 8.14 5.59 2.68 160 26.9 20.4 10.4 7.16 3.43 180 33.4 25.3 13.0 8.91 4.26 200 30.8 15.8 10.8 5.18 220 36.7 18.8 12.9 6.18 240 22.1 15.2 7.26 260 25.6 17.6 8.42 F factor for pivots: Without end gun: 0.54 With end gun: 0.56 Adjustment for pipes with C value 100 110 120 130 140 150 different roughness Multiplier 1.86 1.56 1.33 1.15 1.00 0.88

where RC = location along the lateral where the pipe diameter changes PLS = pressure loss from the inlet to the distal end for the small diameter pipe PLL = pressure loss from the inlet to the distal end for the large diameter pipe. An example of the use of this procedure is included in the example shown in Figure 16.24. The example illustrates that the investment in larger pipe for the first 40% of the lateral would reduce the inlet pressure by approximately 20% while using larger pipe along the entire lateral would only reduce the pressure by 29%. A cost analysis is required to determine which alternative is optimal. The above computations for pressure distribution along the lateral neglected the effects of elevation changes. If there is a uniform slope along the lateral the pressure can be adjusted appropriately and the proper nozzle sizes can be selected to provide the desired discharge. However, the performance of the pivot across a sloping field varies once a specific set of nozzles is installed. Often design along a flat slope is a reasonable compromise for sloping fields. For many fields irrigated with pivots the terrain is

602

Chapter 16 Design and Operation of Sprinkler Systems 450 168 mm DIAMETER PIPE

PRESSURE IN LATERAL, kPa

400

MIXED PIPE SIZES

350 300 250 200

203 mm DIAMETER PIPE

150 100 80 L/s FLOW RATE

50 0 0

50

100

150

200

250

300

350

400

450

500

RADIAL DISTANCE FROM PIVOT, m

Figure 16.24. Distribution of pressure along center pivot laterals with external diameters of 168 and 203 mm, and for a mixed lateral with 200 m of 203-mm and 300 m of 168-mm pipe.

neither uniform nor flat. Since the design of sprinkler packages is usually accomplished with computer programs that start at the end of the lateral and determine pressure loss and nozzles sizes for each outlet along the lateral, the elevation at each location can be used in the calculation. Pressure regulators are commonly used if the elevation differences in the field lead to pressure variations that would reduce the uniformity of application below an acceptable level. The procedures from Chu and Moe (1972) are useful in understanding the hydraulics of sprinkler laterals and for simple analysis of pivots. However, center pivot systems are typically designed using computer programs that start at the distal end of the pivot and sequentially compute the pressure available at each upstream sprinkler outlet (Heermann and Stahl, 2006). In these programs the friction loss in the portion of the lateral between outlets is computed using either the Darcy-Weisbach or HazenWilliams equations for a pipe without outlets. The pressure losses through fittings used to connect the sprinkler to the lateral are also computed. This has become more important as the operating pressure for sprinkler devices continues to decrease to save energy and as the sprinkler device is installed further from the pivot later using drops. Using the sprinkler discharge equation for the required sprinkler discharge gives the required flow at each outlet. With this information and the available pressure at the base of the sprinkler device, the most appropriate nozzle size can be determined. Since a finite number of nozzle sizes are available, it is not possible to exactly match the required discharge. Many designers maintain the cumulative error between the actual total flow in the system and the required flow. At each outlet the nozzle is selected that closely matches the required discharge for the outlet and that minimizes the cumulative flow error. These types of programs provide very detailed specification of system characteristics and allow for consideration of the terrain of the land.

Design and Operation of Farm Irrigation Systems

603

16.7.2 Uniformity of Application As with other types of sprinkler systems the uniformity of application beneath center pivots can be computed by considering the overlapped depth of application from individual sprinklers. This analysis provides guidance for the most appropriate spacing of sprinklers devices along the lateral. The procedure is more complex for pivots since individual sprinklers are not stationary but travel in circular paths about the field. Bittinger and Longenbaugh (1962) developed the framework for analyzing the distribution of water from moving sprinkler systems. They considered the case of sprinklers moving in a straight line and in a radial pattern. They showed that sprinklers that travel on an arc produce a skewed application pattern that is somewhat difficult to analyze. However, the effect of the skewed pattern is negligible when the distance from the pivot point to the sprinkler is more than five times the wetted radius of the sprinkler. For center pivots the inaccuracy of uniformity calculation incurred from assuming linear travel of sprinklers will be small, thus only the solutions for linear travel are presented here. The framework for the analysis by Bittinger and Longenbaugh, which considered triangular and elliptical water application patterns for an individual sprinkler, is illustrated in Figure 16.25. The precipitation rates for the triangular and elliptical application rates are given by Ppt (Wr − s ) for triangular patterns with 0 ≤ s ≤ Wr P= Wr

and

P=

Ppe Wr

Wr2 − s 2 for elliptical patterns with 0 ≤ s ≤ Wr

(16.24)

where P = precipitation rate at distance s from the sprinkler, mm h-1 Pp = peak precipitation rate at the sprinkler location, mm h-1 Wr = wetted radius of the sprinkler, m s = distance from the observation point to the sprinkler, m t, e = subscripts denoting triangular and elliptical patterns. The peak application rate for the triangular and elliptical patterns can be computed from the discharge of the sprinkler located at radial distance R and the wetted radius of that sprinkler: Ppt =

3q R πWr2

for triangular patterns and Ppe =

3q R 2πWr2

for elliptical patterns

(16.25)

The solutions for the depth of water applied at a point from an individual sprinkler as the pivot lateral moves over the point were developed by Bittinger and Longenbaugh (1962). Heermann and Hein (1968) used this procedure to compute the depth of water along a radial line from the pivot point to the end of the center pivot. The depth applied at a point is the summation of water applied by all sprinklers that apply water to the point and is given by for triangular patterns: d =

⎛ 2 Wri Ppti ⎡⎢ 2 2 ⎜ 1 + 1 − ui − − u u 1 ln ∑ i i ⎜ ω i =1 Ri ⎢ ⎜ ui ⎢⎣ ⎝

1

N

⎞⎤ ⎟⎥ ⎟⎥ ⎟ ⎠⎥⎦

604

Chapter 16 Design and Operation of Sprinkler Systems

APPLICATION RATE

TRIANGULAR

ELLIPTICAL

AVERAGE

DISTANCE FROM SPRINKLER (s)

OBSERVATION POINT

PATH OF SPRINKLER

ET ANC IST D IAL RAD

ER NKL PRI S O

(R)

Wr

s x



PIVOT POINT

Wr

u Wr

PIVOT LATERAL

SPRINKLER

AREA WETTED BY A SPECIFIC SPRINKLER

Figure 16.25. Illustration of parameters used to compute uniformity of water application for center pivot systems.

(

2

π N Wri Ppei 1 − ui and for elliptical patterns: d = ∑ Ri 2ω i =1

)

(16.26)

where d = depth of water applied at a point u = ratio of the distance from the sprinkler to the point relative to the wetted radius of the sprinkler,

Design and Operation of Farm Irrigation Systems

605

ω = angular velocity of the pivot lateral, Ri = radial distance from the pivot point to sprinkler i, N = number of sprinklers that apply water to the point of interest. The angular velocity can be determined from the system capacity and the average depth of water applied (da): 2πCg ω= (16.27) da In using these equations a series of points along a radial line is selected that represents the density of observations desired for computing the uniformity. At each point Equation 16.26 is used to compute the aggregate depth of application. Heermann and Hein (1968) used this technique to compute the coefficient of uniformity for pivots as given by ⎡ ⎛ ∑ d j X j ⎞⎟ ⎤⎥ ⎜ ⎢ j ⎟⎥ ⎢ ∑⎜ X j d j − X j ⎟⎥ ∑ ⎢ j ⎜ ⎟⎥ ⎜ j ⎢ ⎠ (16.28) UC p = 100 ⎢1 − ⎝ ⎥ d X ∑ j j ⎥ ⎢ j ⎥ ⎢ ⎥ ⎢ ⎥ ⎢ ⎦ ⎣ where UCp = the uniformity coefficient for center pivots Xj = distance from the pivot point to the point that the depth is computed dj = depth of water applied at point j. Field results and simulation models have shown that the uniformity of application is very high for well-designed and operated center pivots. It is common to find uniformities greater than 90 for center pivots. This is higher than most other types of sprinkler systems or alternate irrigation methods. The previous analysis assumed that the angular velocity was constant. Of course center pivot lateral do not rotate at a constant angular velocity. Some pivots are designed so that the towers move at a constant velocity for a period of time and then the tower is stationary until a signal is received for the tower to move again. Other pivots are designed to move continuously but not at a constant velocity. If a pivot tower is stationary at one point the area watered during that time receives a larger application than areas irrigated when the lateral is moving. If sprinkler devices are used that have a small wetted radius, the start-stop motion of the later can lead to nonuniformity, as illustrated by Hanson and Wallender (1985). With this program they were able to simulate the performance of the sprinkler package for any terrain and for the adjustments of the pivot system that control the start-stop nature of the pivot. This procedure allows for efficient evaluation of many design alternatives. The development presented is based on triangular or elliptical application patterns for a single sprinkler. Devices that have been introduced recently have a different application pattern that those analyzed by Bittenger and Longenbaugh (1962). The procedure presented here should be combined with the single-leg distribution for the packages that are of concern to assess the performance of sprinkler packages.

606

Chapter 16 Design and Operation of Sprinkler Systems

At this point the nozzle size for each sprinkler along the lateral can be computed for alternative sprinkler packages. The operating pressure at the pivot point can be estimated for the field terrain and each sprinkler package. Ultimately, the uniformity of application can be computed (Heermann and Stahl, 2006). However, two additional features, runoff and evaporation losses, have a bearing on the suitability of a design.

LIM ITE D

40

60

CLAY

D

AN TIC IPA EX TE CE DC LL EN EN T TE RP IVO GO O T

30

70

SAND

90

50

50

40

SILTY CLAY

SANDY CLAY

SILTY CLAY LOAM

CLAY LOAM

60

70

80

SANDY CLAY LOAM

20

LO AM Y

20

80

30

10

10

90

ILT TS EN RC PE

PE PE RF RC OR MA EN R MA TC GI NA LA NC L Y E

16.7.3 Runoff Avoidance For center pivots to operate efficiently the applied water must be available for crop use. If the water runs off sloping lands or evaporates while in the air, the design efficiency will not be achieved. Keller and Bleisner (1990) provide a simple diagram of the suitability of center pivots based on the soil texture (Figure 16.26). While this helps with the general suitability, it is not adequate for design of the center pivot. The water application rate beneath the outer spans of a center pivot is very high. The application rate may exceed the ability of the soil to infiltrate water. Some of the water applied at rates exceeding the infiltration rate can be stored temporarily on the soil surface. This is called surface or retention storage. Once the local surface storage is filled, the excess application begins to flow across the field. Some infiltration occurs as the water flows across the field. Ultimately, the runoff water either accumulates in low areas in the field leaves the field. In either case the distribution of water that infiltrates is different than the distribution of water applied and some water is lost as runoff or deep percolation. Thus, both uniformity and efficiency are reduced when runoff becomes significant. Kincaid et al. (1969) investigated the potential for runoff from center pivot irrigation systems. They developed a procedure to adjust the infiltration rate measured with systems that pond water on the soil surface for conditions that occur under center pivots. Dillion et al. (1972) were the first to develop design procedures that considered the infiltration characteristics of the soil. A combination of these procedures was used by Gilley (1984) to develop guidelines for the selection of sprinkler packages based on

LOAM SANDY LOAM

90

SILT LOAM SILT

SA ND

80

70

60

50

40

30

20

10

PERCENT SAND

Figure 16.26. Anticipated performance of center pivot systems based on soil texture (adapted from Keller and Bleisner, 1990).

Design and Operation of Farm Irrigation Systems

607

70 INFILTRATION RATE PEAK APPLICATION RATE

RATES, mm / hour

60

50 SURFACE STORAGE

40

APPLICATION RATE

POTENTIAL RUNOFF

30

20

10

0 0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

TIME, hours

Figure 16.27. Illustration of the runoff potential when the application rate of the pivot exceeds the infiltration rate of the soil.

the potential for runoff. With this method an elliptical pattern is assumed for the sprinkler package. The combined distribution of water from the overlap of individual sprinklers provides an elliptical application rate perpendicular to the lateral (Figure 16.27). This application rate is given by P (t ) =

Pp tp

2 t t p − t2

(16.29)

where P(t) = rate of water application as a function of the time (t) that water has been applied Pp = peak application rate for the elliptical pattern of the sprinkler package tp = time after initial wetting that the peak application rate is reached. With this distribution for the sprinkler package the peak application rate can be computed as ⎛ Q ⎞ R Pp = 4 ⎜ S ⎟ ⎜ π R 2 ⎟ Wr S ⎠ ⎝ (16.30) This demonstrates that the peak application rate for the sprinkler package depends entirely on the design of the system. Once a flow rate is established for the irrigation system and the sprinkler package is installed on the pivot, the peak application rates for points along the system are established. The quantity within the parentheses in Equation 16.30 is the gross system capacity. The peak application rate increases linearly with the system capacity and the distance from the pivot point and inversely with the wetted radius of the sprinkler package. The time required to reach the peak application rate (tp) can be computed from:

608

Chapter 16 Design and Operation of Sprinkler Systems 100 INFILTRATION RATE

RATES, mm/hr

75

50

APPLICATION DEPTH

25 20 mm

30 mm

40 mm

60 mm

0 0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

TIME, hr

Figure 16.28. Illustration of the effect of the application depth on potential runoff.

tp =

Wr d g ⎛ Q s 2π R ⎜ ⎜ π R2 S ⎝

⎞ ⎟ ⎟ ⎠

(16.31)

Thus, the time to reach the peak application rate is linearly related to the wetted radius of the sprinkler package and the gross depth of water applied, and inversely related to the distance from the pivot point and the system capacity. The total time that water is applied to a point is twice the time required to reach the peak application rate for the elliptical pattern which is symmetrical about the peak rate. Thus, the time that water is applied to a point is affected by design variables (i.e., Wr, R, QS, and RS) and by management (dg). The example in Figure 16.28 shows how the potential for runoff increases for increasing water application depths. Gilley (1984) used the infiltration rates provided by the intake family of curves provided by the Natural Resources Conservation Service to estimate the maximum depth of water that could be applied before runoff began. The surface storage volumes from Dillion et al. (1972) were used to incorporate the effects of slope on the runoff process (Table 16.11). Results from Gilley are presented in Figure 16.29 for the four soils that he analyzed. These results provide a means to begin assessment of the suitTable 16.11. Allowable surface storage as a function of slope (adapted from Dillion et al., 1972). Slope Range, % Allowable Surface Storage, mm 0-1 12.7 1-3 7.6 3-5 2.5 >5 0.0

Design and Operation of Farm Irrigation Systems MAXIMUM APPLICATION DEPTH, mm

MAXIMUM APPLICATION DEPTH, mm

0.1 INTAKE FAMILY 40

30

20

12.5 7.5

10 2.5

0

SURFACE STORAGE, mm

0.3 INTAKE FAMILY 40

30 12.5

20 7.5

10

2.5

0

SURFACE STORAGE, mm

0

0 0

50

100 150 PEAK APPLICATION RATE, mm/hr

200

0

250

50

100

150

200

250

PEAK APPLICATION RATE, mm/hr

50

50 0.5 INTAKE FAMILY

MAXIMUM APPLICATION DEPTH, mm

MAXIMUM APPLICATION DEPTH, mm

609

50

50

40

30 12.5

20 7.5

10

2.5

0

SURFACE STORAGE, mm

0 0

50

100

150

PEAK APPLICATION RATE, mm/hr

200

250

1.0 INTAKE FAMILY 40

30

12.5

20

7.5 2.5

10

0

SURFACE STORAGE, mm

0 0

50

100

150

200

250

PEAK APPLICATION RATE, mms/hr

Figure 16.29. Maximum depth of application to avoid runoff as a function of the peak application rate for the sprinkler package installed on a center pivot for four NRCS intake families and surface storage values shown in Table 16.11.

ability of sprinkler packages on fields with various soils and slopes. The method by Gilley was extended to consider an entire field by Wilmes et al. (1994). Von Bernuth and Gilley (1985) and Martin (1991) adapted the method from Hachum and Alfaro (1980) to use the Green-Ampt model instead of the family method to simulate infiltration. These developments allow for the consideration of surface sealing and initial soil water distribution on the performance of sprinkler packages. The variables given above for the description of the performance of center pivot sprinkler packages shows that five system variables are involved. Considering these variables shows that runoff increases as the system capacity, depth of application, and the length of the lateral increase. The runoff is inversely related to the wetted radius of the sprinkler package. Problems can occur if the length of the pivot lateral is excessive, especially for soils with low infiltration rates and steep slopes. Unfortunately, some operators attempt to rectify runoff problems by reducing the system capacity. This increases the potential for water stress during dry periods and encourages management that applies excess water during time of the year when water demands are low in an effort to guard against periods of inadequate supply. The methods such as those by Heermann et al. (1974), von Bernuth et al. (1984), and Howell et al. (1989) should be employed to determine the system capacity required for a specific location and soil. These methods included the utilization of stored soil water to carry crops through periods of high evaporative demand. Users should not arbitrarily reduce capacity to avoid runoff problems.

610

Chapter 16 Design and Operation of Sprinkler Systems

The sprinkler package and the depth of application are the most feasible adjustments once a system is in place. The depth of application per irrigation can be easily adjusted by operators to manage applications to reduce runoff. This can be an iterative procedure based on field observation of runoff problems. The second most feasible method to solve runoff problems is to replace the sprinkler package (perhaps only on the outer portions of the pivot) to sprinkler devices that provide a larger wetted radius for the same pressure as the previous package. Boom systems are available to spread water over a wider distance to also mitigate problems. There have also been a series of special tillage systems developed to increase the surface storage to reduce runoff. One implement that has been used creates implanted reservoirs. The machine used to create the implanted reservoirs uses a subsoiler followed by a paddlewheel assembly that creates the reservoirs. Research by Oliveria et al. (1987), Kranz and Eisenhauer (1990), Cuelho et al. (1996), and others has helped to quantify the benefits of systems that develop implanted reservoirs. The implanted reservoirs provide 5 to 10 L of storage per reservoir. For the reservoir density and row spacing used by Cuelho et al. (1996) the storage volume is equivalent to the water use for two days during the middle of the season if water is applied to every furrow and all furrows are specially tilled. The ability to sustain the reservoirs throughout the growing season varies greatly depending on the cohesiveness of the soil, the slope in the field and rainfall patterns. Others have used basins within the furrows to store water. The basins are generally from 1 to 3 m long. Generally basins can store more water than implanted reservoirs but they are not well suited to sloping soils. In the end, the designer must weight the potential savings of using low-pressure sprinkler devices that have smaller wetted radii than devices requiring more pressure against the costs of special tillage to store water that may run off. In some locations there is an additional benefit of special tillage in reduction of runoff and erosion from rainfall. 16.7.4 Devices to Irrigate the Corners Many center pivots are equipped with special sprinklers attached to the end of the lateral to increase the amount of land irrigated in the corners of the field. The end gun is operated when the lateral reaches an angle where the end gun throw will stay within the boundaries of the property (Figure 16.30). The central angle (β) during which the pivot operates was presented by von Bernuth (1983): ⎛R ⎞ S ⎟ ⎜R ⎟ ⎝ E⎠

β = cos −1 ⎜

(16.32)

where RE is the total radial length irrigated when the end gun operates. The area irrigated in each corner (AE) is given by

(

)

⎤ ⎡π AE = RE2 − RS2 ⎢ − cos −1 (β )⎥ ⎣4 ⎦

(16.33)

These relationships show that there is a trade-off between the radial length of the area irrigated with the end gun and the central angle that is irrigated. When the coverage of the end gun is short, the central angle is larger but the area gained per unit rotation of the lateral is small. The area irrigated in one corner relative to the area irrigated with the primary system is shown in Figure 16.31. These results show that the area in

Design and Operation of Farm Irrigation Systems

611

FIELD BOUNDARY

AREA IRRIGATED WITH ENDGUN

END GUN RADIUS

RE

β

RADIUS OF AREA IRRIGATED WITH ENDGUN ON

Rs

RADIUS OF AREA IRRIGATED WITH ENDGUN OFF

PIVOT POINT

Figure 16.30. Illustrations of the parameters used to describe the amount of land irrigated in pivot corners.

End Gun Discharge / System Discharge

0.5

D

r ha isc

ge 5

0.4

4

0.3

3

Area 0.2

2

0.1

1

Area in One Corner as % of System Area

6

0.6

0

0.0 0

0.05

0.10

0.15

0.20

0.25

End Gun Radius / System Radius

Figure 16.31. Discharge required for an end gun as a percentage of the flow for the main system and the area in one corner of the field relative to the area in the main field.

612

Chapter 16 Design and Operation of Sprinkler Systems

the corners is maximized when the length of the end gun coverage (RE –RS) is approximately 18% of the system radius (RS). The discharge for an end gun depends on the length of the coverage by the end gun: q E = QS

(R

2 E

− RS2

RS2

)

(16.34)

where qE is the discharge required for the end gun. The discharge for the end gun is approximately 35% of the discharge required for the primary system if the optimal area is irrigated (Figure 16.31). While the results in Figure 16.31 show that the area is maximized at a ratio of 18% of the system length, it is difficult to find end guns that provide the wetted radius required at the needed discharge. The area in the corner does not change appreciably for radii larger than 12% of the system length; therefore, the radius of the end gun can be less than the optimal with little loss in area. Solomon and Kodoam (1978) provided data on the pattern for end sprinklers. They show that the depth of application from typical end guns decreases near the edge of the pattern wetted by the end gun. Because the application tapers off at the end, it may be necessary to apply some water beyond the edge of the field to ensure that crops planted there receive an adequate water supply. In windy climates this is especially important. Application of water beyond the boundary of the planted cropland reduces the application efficiency of the end gun below that for the primary system and affects the central angle if the overthrow cannot be applied to the adjacent land. Many center pivots used today utilize a low-pressure sprinkler package for the primary system. Although end guns have been developed that require less pressure than earlier, end guns often require more pressure for adequate operation than is available at the end of the lateral for many sprinkler packages. A booster pump is generally needed on these systems to provide for adequate operation. The booster pump is usually located near the end of the lateral, often at the last tower. As special supply line is provided from the lateral to the booster pump and ultimately to the end gun. The power required for the end gun is computed from the discharge for the end gun and the difference in pressure between that for the sprinkler lateral at the distal end and that required for the end gun. Any pressure losses in the supply system to the end gun must also be considered. As the above information illustrates the discharge through the end gun represents a substantial portion of the flow for the primary system. This results in more pressure loss along the lateral while the end gun operates, which can cause the discharge from the sprinkles on the lateral to decline. Ultimately, two systems curves need to be constructed to determine how the center pivot interacts with the pump used to supply irrigation water. A pump should be selected that provides for efficient operation for both conditions. It would be most desirable for the difference in the discharge for the primary system when the end gun operates and when it is off to be small. To irrigate a larger portion of the area in the corners of a field a corner span can be attached to the end of the primary lateral. When the lateral has rotated to the proper angle, the corner span begins to move into the corner of the field. As the corner span rotates into the corner, a series of sprinklers turns on. More sprinklers operate as the corner span rotates further into the corner. The hydraulics of the corner system is very complex requiring computer modeling to predict the performance of corner machines.

Design and Operation of Farm Irrigation Systems

613

16.8 LATERAL MOVE SYSTEMS The components of center pivot systems have been used to develop a system that travels in a straight line. These systems are called linear or lateral move machines. The towers, pipe material and the truss systems are very similar to center pivots. The difference is that instead of pivoting about a permanent base where water is supplied, the water supply to the lateral is available across the field. Water application amounts and the frequency of irrigation for lateral move systems are similar to that for center pivots. Thus, guidelines and limitations for design and operation of pivots generally apply for lateral move systems. Water can be supplied to the lateral move by one of three methods. A supply canal can run parallel to the direction of travel of the lateral move (Figure 16.32). A pump mounted on the supply tower lifts water from the canal and pressurizes the water for the system. Although a portable dam can be attached to the suction system to block flows on sloping fields; the canal system is limited to fields that are relatively flat in the direction of travel. A second choice is to drag a hose across the field similar to that for a traveler (see Section 16.10). Water is supplied from the main pump to a riser along the travel path of the supply tower. A hose is attached to the riser and to the inlet on the supply tower. As the machine moves it pulls the hose across the field. Many

Figure 16.32. Examples of lateral move systems that are supplied by a ditch and a hard-hose system.

614

Chapter 16 Design and Operation of Sprinkler Systems

times the valve is placed in the middle of the field and the supply hose is long enough to irrigate the entire field without stopping. If the field is too long, several risers may be needed. Some lateral moves were designed to automatically connect to valves connected to an underground main line. This type of system is equipped with carts that lead and trail the lateral. The carts automatically connect to valves installed on risers from the buried main line. Water is supplied by one or both carts. This system is expensive and complex; thus, few of these systems have been produced. The lateral move requires a guidance system to control the direction of travel. Three types of systems have been used. One uses an aboveground cable that runs across the field parallel to the direction of travel. An assembly on the cart follows the cable and keeps the lateral move on course. A second option uses a signal from a lowvoltage buried cable with an antenna guidance system on the control tower. The third option uses a trench cut across the field to provide the direction. A guide follows the trench to steer the lateral move system across the field. The guidance and alignment system on lateral move machines is interfaced to maintain proper alignment and to ensure that the system follows the intended path. The speed of the last tower is controlled by the irrigator to apply the desired depth of water. The alignment of individual towers for lateral move systems works similar to that for center pivots. However, this does not ensure that the lateral move progresses parallel to the guidance cable. The guidance assembly is designed to reduce the speed of the interior towers if the lateral move begins to travel away from the guidance cable. If the lateral move system is progressing on an angle toward the guidance cable, the velocities of the interior towers are increased causing the lateral move to change the direction of travel. Lateral move systems have characteristics similar to pivots. They also have the following advantages: ƒ a larger portion of a square field is irrigated than for pivots; ƒ a rectangular field can be irrigated; and ƒ the rate of water application is less than for pivots, leading to fewer problems with runoff. The disadvantages of lateral move systems are that: ƒ the cost per unit area irrigated is substantially higher than for pivots; ƒ more labor is required to move the system to the starting point or to reverse the system so it can irrigate in the opposite direction; ƒ the hose used to supply the system can be difficult to move and attach; and ƒ aboveground guidance and supply systems interfere with farm operations. Utilization of lateral move systems lags significantly behind the use of center pivots. The higher investment costs and increase labor required to arrange the water supply and to reposition the lateral move system deter wider use. However, the systems do offer substantial promise. Low energy precise application (LEPA) systems and other types of sprinkler packages that are capable of applying water below crop canopies are well suited to lateral move systems. Such systems essentially become moving microirrigation systems. If the development of site-specific irrigation develops, lateral move systems are logical choices for supplying irrigation water, crop nutrients, and other agricultural chemicals. Thus, there are good reasons to assume that the use of lateral move irrigation systems will grow.

Design and Operation of Farm Irrigation Systems

615

16.8.1 Layout Since the cost per unit length of the lateral move usually exceeds the linear cost of the water supply system, investment costs are minimized by laying out the system so that the lateral pipeline is parallel to the shortest side of a rectangular field. This orientation results in longer travel distances that produce longer irrigation intervals. The length of the field, water supply capacity, and the depth of application determine the irrigation interval. If the lateral move is repositioned to a starting location after each irrigation the downtime required to move the lateral and water supply system must be included in the irrigation interval. In some cases irrigators simply reverse the direction of travel at one edge of the field and irrigate in the opposite direction for the subsequent irrigation. There is some downtime to reposition the water supply system for this mode of operation but it is not as long as required to return the lateral move to the starting position. With this type of operation the depth of water applied and the irrigation interval at the ends of the field is essentially twice that at the center of the field. Care must be taken to avoid deep percolation and/or water stress at the edges of the field when operating in such a manner. Lateral move systems are not capable of negotiating steep lands or severely rolling terrain. The maximum recommended slope along the lateral can be as high as 6%, but should generally be less than 2%. The type of water supply system and the direction of operation determine the maximum slope in the direction of travel. For canal-fed systems the maximum slope is about 0.5%, while it is 1% for canal-fed systems with a movable dam in the canal and 3% for hose-supplied systems. The pressure loss in the lateral can be minimized by placing the water supply system in the center of the field. However, this may present obstacles to farming operations that are less severe if the water source is placed along the boundary of the field. The consequences of each alternative should be discussed with the producer. As with all sprinkler systems the main line must be designed after the lateral move system is designed. The head-discharge curve for matching pumps to the sprinkler system can be developed by treating the lateral move as a single large sprinkler or the pressure distribution function can be used directly. The head-discharge relationship depends on the losses in the conveyance system, supply carts, and canal suction system. The losses in various components for typical systems are shown in Figure 16.33. Information should be obtained for specific models for final design. 16.8.2 Water Application Lateral move systems have characteristics similar to center pivots and periodically moved laterals. The discharge from sprinklers should be constant similar to moved laterals. The discharge is given by Q S qS = S L FW (16.35)

where the field width (Fw) is parallel to the lateral. The pressure distribution along the lateral is similar to that for moved laterals, yet the pipe materials are the same as for center pivots. The friction per unit length along the lateral can be computed from Table 16.10; however, the F value for lateral move systems should be determined as for moved laterals.

616

Chapter 16 Design and Operation of Sprinkler Systems 50

64

45

76

89

102

114

127 NOMINAL DIAMETER OF HOSE,

H EA D LOSS, m

40 35 30

HARD HOSE

25

152

20 15 10 5

LOSS IN SUPPLY CART

0 0

10

20

30

40

50

60

70

80

90

100

INFLOW, L/s 50 64

45

76

89

102

114

40 127

HEAD LOSS, m

35 30 25

NOMINAL DIAMETER OF HOSE, mm

20 15 10

SOFT HOSE @ 1000 kPa

5 0 0

10

20

30

40

50

60

70

80

90

100

INFLOW, L/s

Figure 16.33. Head loss in hard and soft hoses used on lateral move and traveler irrigation systems. Head loss in the supply cart for travelers is also shown.

The pressure and required discharge for a sprinkler are used to select the sprinkler package and the nozzle sizes. The lateral move is different than moved laterals in that the size of nozzle may vary along the lateral as the pressure changes. If pressure regulators are used along the entire machine, the nozzle sizes would be constant. The application rate is uniform along the lateral of a lateral move system because the representative area is the same for all sprinkler The application rate can be computed using the elliptical or triangular patterns as previously presented. The peak application rate for a lateral move system is given by 4QS Pp = (16.36) πWr FW and the time to the peak application rate is tp = Wr/v

(16.37)

Design and Operation of Farm Irrigation Systems

617

where v is the average linear velocity of the lateral move and Wr is the radius of coverage for the sprinkler package. For equal net system capacities and application efficiencies, the peak application rate for a lateral move is equal to the application rate at a point 65% of the way along the lateral of a pivot. The depth of water applied per irrigation is given by dg =

QS vFW

(16.38)

Runoff is less of a problem with lateral move systems than for center pivots because field slopes are less than sometimes found for center pivots and the peak water application rate is less than for pivots. The runoff relationships for center pivots in Figure 16.29 can be utilized to determine if runoff is a potential problem. Lateral move systems provide the opportunity of very high uniformity of application. Part-circle sprinklers can be used to improve the uniformity at field boundaries. The effect of sprinkler spacing on the uniformity can be estimated using the overlapping procedures developed for center pivots. Automated sprinkler systems are increasingly replacing surface irrigation systems. When systems are replaced irrigation management must change. A frequent problem is that irrigators attempt to apply the same depth of water per irrigation with automated sprinkler systems as was applied with surface systems. This negates the potential of the automated system and is generally unsuccessful. Runoff and traction problems generally occur with such management. Applying smaller depths per irrigation capitalizes on the potential of automated sprinkler systems to provide high application efficiencies. The maximum depth of water applied with lateral move systems should be less than 50 mm with 25 mm per application being typical. 16.8.3 Water Supply System Design considerations vary depending on the type of water supply used for the lateral move. For systems that drag a supply hose, considerations involve the friction loss in the hose and other elements of the water supply system and the force required to drag the hose across the field. The pipe roughness for the Hazen-Williams equation (i.e., the C value) is approximately 150 for the hard or soft hoses used to supply water. The inside diameter of soft hoses varies with the pressure inside the pipe and additional information is required for operating pressure other than shown in Figure 16.33. There are limits on the shortest bending radius for each hose. Short bends cause the soft hose to kink, which temporarily blocks the flow, but there is no long-term damage to the hose. The hard hose can be damaged if it is bent too severely. The force required to drag the hose limits the maximum diameter and length of hose and may require a special design of the tower that pulls the hose. Hose lengths of 200 m are common as that length matches well with the property dimensions common in the U.S. Hoses longer than 200 m are not common and should only be specified if in consultation with manufacturers. The normal towers used to pull the hose are not usually capable of pulling 200 m of 203-mm diameter hose. Even for smaller diameter hose, the supply tower can be equipped with four wheels that are supplied power rather than the normal two. This improves traction for wet or slippery surfaces.

618

Chapter 16 Design and Operation of Sprinkler Systems

Systems that are supplied from a ditch require that the channel capacity, system capacity, water supply source, and the onboard pump and power unit be matched. Once the topography is known the ditch can be designed using procedures described in earlier chapters on conveyance systems. The minimum depth of the ditch is usually about 1 m. The minimum bottom width is about 0.3 to 0.6 m. Erosion, weed control, and trash in the ditch are issues that must be considered in the design and operation. If the soils at the site are highly permeable, the canal may need to be lined with flexible membranes or concrete. Either type of lining significantly increases the cost of the system. The sideslope of the canal should not be so steep that erosion occurs or that humans and wildlife cannot climb from the canal should they enter the waterway. Outflow structures may be required in locations where storm water enters the canal. Special features are needed for canal supply systems to ensure that the ditch does not overtop if the lateral move stops. Controls are also needed to stop the lateral move if the water supply is interrupted. Systems that automatically connect to valves attached to buried main lines require special considerations in design. When the water flow changes between the supply carts some water hammer may develop in the supply system. The amount of pressure surge depends on the flow and the time required for valves to open and close. Pressure surges problems are most prevalent on long lateral moves that require a large inflow. Pipeline protection is essential for these systems. The supply system for these lateral moves is very complex and is usually designed by the manufacturer.

16.9 LOW ENERGY PRECISION APPLICATION (LEPA) SYSTEMS LEPA irrigation systems are either center pivots or lateral move systems that are modified with extended length drop tubes and application devices designed to apply small frequent irrigations at or near ground level to individual furrows (Lyle and Bordovsky, 1981). The primary purpose of LEPA systems is to minimize evaporation from spray droplets, foliage, and soil surfaces. The LEPA concept involves soil surface management to increase surface storage (Lyle and Bordovsky, 1983). Selected tillage methods and/or crop residue management are used to increase retention of both rainfall and irrigation. Nozzle pressure requirements for LEPA are low since wide dispersal of water is not necessary. Much of the pressure head at the nozzle (located near ground level) is obtained by the elevation head differential between the lateral and the nozzle. This results in reduced lateral design pressures as compared to overhead sprinkler packages and reduced energy requirements. LEPA systems discharge water beneath the plant canopy and therefore offer an opportunity to irrigate with poor-quality water that might cause leaf burn when applied through sprinkler systems. LEPA systems are also advantageous for crops susceptible to fungal diseases that may thrive with frequent wetting of the foliage. 16.9.1 General Considerations There are numerous considerations in the design, installation and management of LEPA systems that are not required for sprinkler systems (Lyle, 1994). Since water is applied as a narrow band or stream it is important that the LEPA drop tubes and discharge devices be positioned so that each plant within a field has equal opportunity for irrigation water delivery. This is best accomplished if water is applied in the furrow

Design and Operation of Farm Irrigation Systems

619

between crop rows. It is, therefore, recommended that plant row position for both lateral move systems and center pivots be established using the system’s tower tire tracks as a guide for row crop establishment. This guideline results in circular rows for center pivots and linear rows for lateral move systems. It is highly desirable that the system span length be an even multiple of the row spacing and also of equipment width to facilitate consistency in row establishment and applicator placement along each span. For alternate furrow application, the drop tubes and applicators should be positioned in the “soft” furrows or those not compacted by tractor or equipment wheel traffic so that infiltration is maintained. This requires that equipment gage wheel location in relation to tractor wheels be such that every other furrow remains free of traffic. LEPA systems are designed to be essentially independent of soil intake rates. LEPA systems are best suited to soils that maintain structural integrity throughout the season to retain surface storage formed by enhancing tillage practices. Topography (slope) is the primary limiting factor in choosing LEPA irrigation. Although system speed can be adjusted to accommodate slopes of up to 2% without irrigation surface redistribution, slopes should be limited to 1% for circular rows in climates with high-intensity rainfall to minimize rainfall runoff and potential soil erosion. In all instances, normal practices to prevent erosion from rainfall runoff, such as terracing and/or grassed waterways, should be used in conjunction with LEPA. 16.9.2 Surface Storage and Water Application Devices The establishment of improved soil surface storage and its maintenance throughout the irrigation season is a function of both tillage methods and type of applicator chosen for irrigation. The combination should maintain seasonal surface water storage capable of holding the entire water volume of each irrigation pass without surface water redistribution. Recommended surface modification practices include basin tillage, reservoir tillage, interfurrow chiseling or subsoiling, or any combination of these practices with thick standing stubble or residue. These practices may also beneficially increase the infiltration rate. To insure surface containment of applied irrigation, LEPA systems should be operated at a sufficiently high speed so that the application volume is less than or equal to the surface storage. For some situations and locations this might require daily irrigation. The surface storage should be minimally diminished during LEPA irrigation. In addition, the design and installation of the nozzle and regulator system should deliver maximum uniformity for all operating conditions within the field and throughout the irrigation season. Ideally, LEPA applicators should have a narrow profile to minimize crop contact and drag. It is desirable for applicators to also possess spray capability in addition to single furrow discharge, although this is not the primary mode of operation. Examples of spray application needs include temporary use for seed germination, herbicide application, and when close-seeded crops are included in cropping rotations. Currently there are two basic LEPA applicator designs that minimize erosion to furrow dikes or other enhanced soil storage conditions (Figure 16.2). One consists of a nozzle and shroud assembly which causes water to be discharged as a continuous sheet or bubble. The other design uses removable drag socks, designed to minimize furrow dike erosion as water is delivered directly to the soil surface.

620

Chapter 16 Design and Operation of Sprinkler Systems

Nozzle selection and hydraulic design are similar to that of other pressurized irrigation systems but careful consideration of elevation head gain and friction losses in each drop tube between the lateral and the nozzle is required to accurately determine the pressure available to the applicator. 16.9.3 Drop Tube Design The drop tube should be sufficiently long to position the application device when irrigating from 10 to 45 cm above the soil surface depending on type of applicator and topography. The diameters of all materials making up the drop tube should be sufficient to supply necessary operating pressure to an applicator or pressure regulator with an overhead lateral pressure at the end of the system of no greater than 70 kPa. A pressure of 21 kPa above the regulator rating is normally necessary at the regulator inlet for correct operation of LEPA systems. Drop tube diameters normally range from 15 to 20 mm. Drop tubes are attached to the overhead pipeline outlets by means of furrow arms (extended length gooseneck connectors) which are normally 30 to 50 cm in length. Although pivots and lateral move systems may be ordered with outlets spacings the same as the desired furrow spacing, they are often not in satisfactory alignment. The furrow arm is therefore rotated to center the drop over the furrow. When determining friction loss between the pipeline and nozzle, entrance and friction losses in the furrow arm and associated fittings as well as the various components of the drop tube must be calculated using appropriate head loss equations. Several material options are available for drop tube construction and configuration. The major (largest) portion of the drop tube, normally the upper portion, must be sufficiently rigid to prevent significant movement in high winds during preplant irrigation and prior to full canopy establishment. Rigidity should also be sufficient to insure that the LEPA applicator remains within the crop canopy after full canopy development. The entire drop tube should exhibit sufficient flexibility to allow travel over terraces or soil mounds, or encounters with other objects without breakage. The upper rigid section of drop tubes may consist of galvanized steel, UV-protected PVC, or extruded polyethylene. Galvanized steel may also be used for a lower section (approximately 70 cm in length) located above the applicator for weight and rigidity to help maintain applicator position within the canopy. Slip weights of galvanized metal, concrete, or poly material may also be slipped over light and/or flexible materials to make up the lower section of drop tubes. A short section (0.5 to 0.67 m) of flexible reinforced or non-reinforced vinyl-covered PVC hose or tubing should be used to couple the more rigid upper and lower section to provide necessary flexibility to the drop. It also allows adjustment of applicator height above the soil surface by modifying the hose length after filling the system with water. 16.9.4 Evaluation and Performance The coefficient of uniformity is used for evaluating overhead irrigation systems but it is not applicable to the individual furrow application of LEPA systems. Instead, LEPA nozzle package design and performance should be evaluated by nozzle discharge uniformity. The nozzle discharge uniformity describes the uniformity of LEPA nozzle discharge rate converted to equivalent application depth along the length of the system. The nozzle discharge uniformity is calculated for pivots with the HeermannHein equation and with the Christiansen’s equation for lateral move systems. How-

Design and Operation of Farm Irrigation Systems

621

ever, the depth of applied water in the equation is replaced with equivalent application depth, which equals the individual nozzle discharge (measured by timed volumetric catchment) divided by coverage area per nozzle. LEPA nozzle package design should result in nozzle discharge uniformities of 96 or greater. A measured drop in uniformity, to 94, due to elevation and/or flow rate changes in the field is the lowest value recommended. Modification of design parameters (operating pressure, pipe size, pressure regulator use, etc.) should be considered if uniformity drops below 94 over significant portions of an irrigated field.

16.10 TRAVELERS Another method to automate and save labor is the traveler irrigation system. Traveler systems consist of a large sprinkler, commonly referred to as a “gun,” mounted on a moving cart (Figure 16.34). Water is supplied to the cart by a hose. With early designs the cart was connected to a cable and winch system which was attached to an anchor. The winch, which was powered by water pressure, retracted the cable pulling the cart across the field. Many current designs use the water supply hose to pull the cart across the field (Figure 16.34). A large spool is anchored at one end of a travel lane. The spool rotates winding up the hose and pulling the cart toward the spool. Whether connected by cable or hose, the spool assembly is designed to provide a nearly constant travel velocity. This requires that the spool rotate faster when small amounts of cable or hose have been retracted. The traveler system is operated as shown in Figure 16.34. The cart is positioned at one edge of the field and is pulled to the center. The system then shutdowns and the cart is moved to the opposite side of the field. Once a strip is irrigated, the system is moved to the next lane. Occasionally, the system can be pulled across the entire field to avoid reconnecting the hose at the middle of the field. The gun on the cart discharges a large volume of water and produces a large wetted radius. Therefore, the travel lanes are often spaced up to 100 m apart. The disadvantages of traveler systems include: ƒ High pressure requirements—Travelers require higher pressures than other sprinkler systems. The pressure at the cart can exceed 700 kPa. Pressure loss in the hose and water supply components add to the pressure requirement. Therefore, traveler irrigation systems have high operating costs. ƒ Labor required to move the cart—Traveler systems should be considered semiautomated since the cart, hose, and spool must be manually moved. ƒ Loss of land in the travel lanes—The carts generally have a low clearance, so if tall crops are grown, the lane must be planted to a low-growing crop. ƒ Nonuniform application at the edges and center of the field—The gun is operated to irrigate part of a circular pattern (Figure 16.34). The exact angle depends on the lane spacing, the overlap of adjacent passes of the traveler, and the design of the gun. However, it is common to require more than 180° rotation, as shown in Figure 16.34. If the traveler cannot apply water beyond the boundaries of the property, there are sections along the edges of the field that may receive less water than the bulk of the field. A series of dry diamond-shaped regions may also occur near the center of the field. If too much overlap occurs at the center there can be areas of excess irrigation.

622

Chapter 16 Design and Operation of Sprinkler Systems HOSE REEL AND CART

SUPPLY HOSE

TRAVELING GUN

ZONE WETTED WHEN STATIONARY

AREA WETTED IN ONE PASS

TRAVELING GUN

SUPPLY HOSE

HOSE REEL/CART

TRAVEL LANE

FIELD BOUNDARY

MAINLINE WELL & PUMP

Figure 16.34. Picture and operational sketch of traveler or big-gun irrigation system.

ƒ Poor uniformity in windy conditions—Since the traveler throws water a long distance, wind can distort the application pattern. To maintain uniformity in locations with high winds the travel lanes must be closer together, which increases labor requirements and leads to longer times between irrigations. ƒ Large droplets—Traveler systems often produce large droplets with a high velocity producing a large amount of energy when reaching the soil/crop surface. The impact energy can exacerbate runoff and soil compaction. Therefore, travelers are best suited to cropping and farming systems that provide plant or residue material to absorb the energy in the droplets before reaching the soil surface. There are several advantages for traveler systems, including their:

Design and Operation of Farm Irrigation Systems

623

ƒ Flexibility—Since the lanes can vary in length, traveler systems can be used to irrigate nearly any shape of field. The systems can be moved from field to field to irrigate several tracts of land. This is especially attractive in semi-humid areas where the system may not be necessary for one field for the entire season or where cropping rotations require that irrigation be shifted from one field to another. ƒ Wide range of irrigation depths—Since the sprinkler moves automatically, the depth of water applied per irrigation can be set to the desired amount. If a small irrigation is needed the system can be set to move quickly across the field. Conversely, for deep-rooted crops the cart can move slowly. ƒ Factory assembly—The system is assembled during manufacturing, thus the system can be quickly installed. This allows traveler systems to be operational with a minimum amount of preparation. Traveler irrigation systems are mostly used in semi-arid and semi-humid climates where the needs for irrigation are not as large and consistent as for arid lands. The systems are more popular in locations where the size and shape of fields are not amenable to center pivot or lateral move systems, or where multiple fields can be irrigated from the same water source. The high pressure requirements for travelers cause operating costs to be high. The major considerations for travelers are the hydraulic design of the water supply system and the arrangement of travel lanes to match field boundaries while achieving acceptable uniformities of application. Users should be very careful around traveler systems. Travelers operate at high pressures and large amounts of tension develop in the hose or cable system used to move the traveler. Both of these conditions contribute to the danger. Special care should be taken to avoid applying water onto electrical power lines. Kay (1983) recommends travel lanes be located at least 30 m from electrical power lines. 16.10.1 Field Layout As with any irrigation system, the initial design step is to lay out the field boundaries and arrange the system on the landscape. Topographical information is needed to ensure that the desired uniformity is achieved. Small elevation changes in the field are of less concern with travelers because of the high operating pressure. Slopes only slightly affect the ability of the traveler to maintain the uniform speed of travel required for acceptable uniformities. Therefore, the topography within a field is not of major concern to the operation of the traveler. The pressure required to reach the highest elevation in the field should be the primary interest. The arrangement of travel lanes is the principal consideration in laying out the system. The maximum length of the supply hose is generally 200 m. If the field is wider than 200 m the traveler will generally have to be pulled toward the center of the field requiring that the main line be positioned through the center of the field. The distance between lanes should be the same across the field. Also, the lanes must be placed close enough together to provide the overlap needed for acceptable uniformity. The spacing of the travel lanes depends on the wind speed and the diameter of coverage of the sprinkler. Recommended maximum spacings are given in Table 16.12. Iteration may be needed to find a travel lane spacing that provides uniform irrigation while fitting within the field dimensions. It is generally recommended that the travel lanes be oriented perpendicular to the prevailing wind direction. The first travel lane may be

624

Chapter 16 Design and Operation of Sprinkler Systems

Table 16.12. Maximum spacing of lanes for traveler systems (adapted from Addink et al., 1980 and Kay, 1983). Wind Speed, m s-1 Diameter of 0 0 to 2.5 2.5 to 5 >5 Coverage, m (80% of Wd) (70% of Wd) (60% of Wd) (50% of Wd) 50 40 35 30 25 60 48 42 36 30 70 56 49 42 35 80 64 56 48 40 90 72 63 54 45 100 80 70 60 50 120 96 84 72 60 140 112 98 84 70 160 128 112 96 80 180 144 126 108 90

located a half of the spacing from the boundary of the field if the water from the sprinkler can be applied beyond the edge of the field. Otherwise the first travel lane should be a full spacing from the field boundary. The sprinkler can be positioned at the edge of the field if the water can be applied beyond the edges of the field that are perpendicular to the travel lanes. Otherwise the sprinkler cart should be positioned to avoid throwing water beyond the property boundaries. If multiple travelers are used in one field it is desirable to supply water to the center of the field and to place a traveler on each half of the field. This minimizes the difference in pressure available to each sprinkler. 16.10.2 Hydraulic Design The discharge from the sprinkler on the travelers should be equal to the flow required for the field divided by the number of travelers used in the field. The guns used on travelers can be equipped with either tapered or ring nozzles. Tapered nozzles generally produce larger diameters of coverage but do not provide as much breakup of the sprinkler jet. The depth of water applied per irrigation can be determined from dg =

qS vWT

(16.39)

where WT is the width of the travel lane (i.e. the distance between travel lanes) and v is the linear velocity of the sprinkler cart. The discharge from the sprinkler and the spacing between travel lanes determines the pressure required for the sprinkler. The pressure must be high enough to provide the discharge and diameter of coverage required. The total pressure loss in the systems consists of losses in the main line, fittings, supply hose, and the sprinkler and hose carts. The pressure loss in the supply hose can be determined from Figure 16.33. The diameter of the soft hose increases as the pressure of the system increases. Thus, if the pressure is different than shown in Figure 16.33 the head loss will vary accordingly. There is additional loss within the sprinkler and reel carts. Losses increase when the hose is wrapped around the reel rather than

Design and Operation of Farm Irrigation Systems

625

when fully extended. Oakes and Rochester (1980), Rochester et al. (1990), and Rochester and Hackwell (1991) provided methods to compute the friction loss for varying sizes of travelers. The head loss in the components of the traveler was computed from

hl = kV 2/2g

(16.40) where hl = the loss k = the loss coefficient V2/2g = the velocity head. The loss coefficients in the sprinkler cart and hose-reel cart were found to be 1.76 and 3.91, respectively. The coefficient for the loss due to coiling on the reel was represented by a bend coefficient which was found to be 0.09 m-1. The bend coefficient is multiplied by the length of hose coiled on the reel. The loss coefficients vary with the size of the system and the design of the equipment; therefore, they should be used as an initial estimate. Portable pipe or underground pipelines with risers and valves can be used for the main line used to supply water to the traveler. The friction loss in the main line is similar to that for any other type of sprinkler system. Special care is required to protect pipelines against pressure surges with traveler systems. Travelers require high operating pressures so pumps used to supply the water are usually designed for high pressures. When more than one traveler is supplied by the same main line or if the supply hose should kink or some other form of obstruction occurs, the pressure in the conveyance system can increase rapidly. The pipeline should be protected to avoid damage under such conditions. In many cases it is best to select pumps where the shutoff head is less than the pressure rating of the supply system. Water hammer can also be a problem since the pressure builds quickly when water reaches the sprinkler. The elasticity of the hose helps to minimize some of the pressure surge problem but water hammer should be evaluated in designing the supply system. 16.10.3 Uniformity The uniformity of application with traveler systems depends on a constant velocity of travel, constant sprinkler discharge and proper spacing to provide adequate overlap. Travelers are now specially designed to vary the speed of the reel used to retract the sprinkler cart so that variations in travel speed along the travel lane are not severe. Long supply hoses exert considerable resistance when fully extended as compared to when only a portion of the hose must be moved. The increased resistance can cause variations in the speed of travel. Thus, designers should check with manufacturers for the maximum length of hose for local conditions. The pressure at the inlet to the sprinkler varies with the distance of travel of the sprinkler cart, thus calculations should be made for varying amounts of retracted hose. The uniformity can be estimated using the overlapping procedure for a constantly moving sprinkler. The single-leg distribution of the guns used on travelers will not usually fit the elliptical or triangular functions that were presented, thus the procedure must be modified for the appropriate distribution (Rochester et al., 1989). The overlapping process can be used to assist in selecting lane spacings. The single-leg distribution for average wind conditions should be used if available. However, the overall field uniformity and application uniformity are difficult to estimate due to the variability along field boundaries and at the center of the field.

626

Chapter 16 Design and Operation of Sprinkler Systems

The operator must check to ensure that the sprinkler device is operating at the appropriate pressure. Low pressure leads to an inadequate breakup of droplets which produces a doughnut-shaped water application pattern. Excessive pressure causes the sprinkler jet to break into to small drops that do not travel as far as intended and that are subject to evaporation and drift.

16.11 AUXILIARY USES OF SPRINKLER SYSTEMS Sprinkler irrigation systems can perform several auxiliary uses in addition to supplying crop water requirements. Permanently installed systems can be operated quickly to meet these additional demands. If the system must be positioned in the field and periodically moved the ability to perform auxiliary uses is diminished. The application of effluent and agricultural chemicals are two common uses. Since the sprinkler system is designed for high uniformity and operates during the rapid growth period of crops, sprinkler systems provide excellent capabilities for these applications. However, these applications are regulated in most states to ensure protections of soil and water resources. In addition, special hydraulic designs are required to guard against the backflow of chemicals into the water source. Details of the use of sprinkler systems for chemigation are discussed in more detail in Chapter 19. Due to the high heat of fusion for ice, sprinkler systems can be used for frost and freeze protection in some applications. Generally water must be applied frequently to the crops during periods of frost or freeze danger. This may require a higher flow rate for the field than needed for crop water requirements. The requirement also eliminates those systems that cannot irrigate the entire field in a very short time. Very careful management is required to ensure that the objectives are accomplished for frost or freeze protection. Successful practices vary considerably depending on the water source, wind speeds, and other local conditions. Local guidelines should be followed for success. Care must also be taken to avoid damage to the irrigation system when applying water during cold periods. Ice may form on structural members of the irrigation system and the added weight may cause components to fail. In some locations wind erosion, before plants are large enough to shield the soil surface, is a major concern. Irrigation during such periods may increase the cohesion between soil particles, which increases aggregate stability and reduces erosion. Care must be taken, however, because the impact from water droplets can dislodge soil particles and contribute to increased wind erosion when the soil dries. Irrigation with small applications is usually adequate to stabilize the soil surface for a period of time. The efficiency of water use for erosion control is low. The small application does not wet the soil to a very large depth which leads to evaporation of water from the soil surface with little storage of water in the soil profile. In many ways irrigating to control wind erosion is a last-ditch effort, as it is better controlled through residue management and other production practices.

16.12 SAFETY Producers, service technicians, and other individuals that work around sprinkler irrigation equipment must be very careful. Sprinkler irrigation equipment is often connected to high-voltage electrical supplies, has numerous moving parts, requires high water pressures, and operates in a wet and slippery environment. The systems are occasionally used to apply chemicals that could be toxic. While many standards and operational guidelines have been developed for the proper design, manufacturing, instal-

Design and Operation of Farm Irrigation Systems

627

lation, and operation of this equipment, not all systems adhere to the intended practices. Anyone designing or operating these systems must be aware of the appropriate laws, codes, and engineering standards that apply to the equipment and the intended use. The standards from ASABE are continually updated and should be consulted routinely for proper practices. Local ordinances and other regulations should be determined before the system is designed.

16.13 SUMMARY This chapter describes the fundamentals of sprinkler irrigation, performance of sprinkler systems including uniformity and efficiency of application, types and characteristics of sprinkler systems currently used, and design and management procedures for specific types of sprinkler systems. Information is provided to enhance design and management of sprinkler systems which are the most rapidly growing form of irrigation today.

LIST OF SYMBOLS AD AE Ai AR Cd Cn d d50 dg Dn Ea Ek FL FW Hn Ir Ke/a MAD n N Nd NL NS P P(i) P3 PL PLL PLS Pp PR PS

allowable depletion area irrigated in a corner of a center pivot area irrigated representative area for a sprinkler on a center pivot discharge coefficient net system capacity depth of water applied or effective diameter of water droplet volume mean drop diameter gross depth of irrigation water applied inside diameter of a nozzle application efficiency kinetic energy length of the field width of the field nozzle pressure head rate of infiltration for bare soil relative to protected soil kinetic energy per unit area management allowed depletion number of sets per lateral number of sprinklers on a lateral diameter of the nozzle number of laterals along the field length number of laterals along the width of the field pressure in a sprinkler lateral or at a sprinkler nozzle rate of water application as a function of time percentage of drops smaller than 3 mm pressure loss from the inlet into to the distal end of the pivot lateral pressure loss from the inlet to the distal end for the large diameter pipe pressure loss from the inlet to the distal end for the small diameter pipe peak precipitation rate at the sprinkler location pressure in the center pivot lateral at a point R from the pivot point pressure at the distal end of the pivot lateral

628

Pv qE qs Qs R Ra RC RD RE s Sa Si SL Sm T TAW Td Ti Tm To tp Ts u U UCp v Wr WT

β ρ ω Ω

Chapter 16 Design and Operation of Sprinkler Systems

percent of the total drops that are smaller than a specified size discharge required for an end gun discharge from a sprinkler flow into the sprinkler system (equal to the gross system capacity) average application rate or radial distance from pivot point average application rate location along the lateral where the pipe diameter changes root depth during peak water use period total radial length irrigated when the end gun operates distance from the observation point to the sprinkler sand content of the soil silt content of the soil sprinkler spacing along the lateral distance between laterals along the main line (equal to the set width) exposure time total available water per unit depth of soil downtime between successive irrigations irrigation interval time required to move the sprinkler later between sets time of operation per irrigation time after initial wetting that the peak application rate is reached. operational time per set for a lateral distance from a sprinkler to a point relative to the wetted radius of the sprinkler wind speed the uniformity coefficient for center pivots droplet velocity or linear velocity for a traveler system radius of coverage or wetted radius of the sprinkler distance between lanes for a traveler system angle of operation for center pivots and end guns density of water angular velocity of a pivot lateral ratio of the nozzle diameter to the pressure at the base of the sprinkler

REFERENCES Addink, J. W., J. Keller, C. H. Pair, R. E. Sneed, and J. W. Wolfe. 1980. Design and operation of sprinkler systems. In Design and Operation of Farm Irrigation Systems, 621-660. M. E. Jensen, ed. St. Joseph, Mich. ASAE. Bezdek, J. C., and K. Solomon. 1983. Upper limit lognormal distribution for drop size data. J. Irrig. Drain. Eng. 109(1): 72-88. Bittinger, M. W., and R. A. Longenbaugh. 1962. Theoretical distribution of water from a moving irrigation sprinkler. Trans. ASAE 5(1): 26-30. Christiansen, J. E. 1942. Irrigation by sprinkling. Univ. Calif. Agr. Exp. Sta. Bull. 670. Chu, S. T., and D. L. Moe. 1972. Hydraulics of a center pivot system. Trans. ASAE 15(5): 894-896. Cuelho, R. D., D. L. Martin, and F. H. Chaudry. 1996. Effect of LEPA irrigation on storage in implanted reservoirs. Trans. ASAE 39(4): 1287-1298.

Design and Operation of Farm Irrigation Systems

629

Dadiao, C., and W. W. Wallender. 1985. Drop size distribution and water application with low-pressure sprinklers. Trans. ASAE 28(2): 511-514, 516. Dillon, Jr., R. C., E. A. Hiler, and G. Vittetoe. 1972. Center pivot sprinkler design based on intake characteristics. Trans. ASAE 15(5): 996-1001. Edling, R. J. 1985. Kinetic energy, evaporation and wind drift of droplets from low pressure irrigation nozzles. Trans. ASAE 28(5): 1543-1550. Eigel, J. D., and I. D. Moore. 1983. A simplified technique for measuring raindrop size and distribution. Trans. ASAE 26(4): 1079-1084. Fisher, G. R., and W. W. Wallender. 1988. Collector size and test duration effects on sprinkler water distribution measurement. Trans. ASAE 31(2): 538-541. Gilley, J. R. 1984. Suitability of reduced pressure center pivots. J. Irrig. Drain. Eng. 110(1): 22-34. Hachum, A. Y., and J. F. Alfara. 1980. Rain infiltration into layered soils: Prediction. J. Irrig. Drain. Eng. 106(4): 311-319. Han, S., R. G. Evans, and M. W. Kroeger. 1994. Sprinkler distribution patterns in windy conditions. Trans. ASAE 37(5): 1481-1489. Hanson, B. R., and W. W. Wallender. 1986. Bidirectional uniformity of water applied by continuous-move sprinkler machines. Trans. ASAE 29(4): 1047-1053. Heermann, D. F., and P. R. Hein. 1968. Performance characteristics of self-propelled center pivot sprinkler irrigation system. Trans. ASAE 11(1): 11-15. Heermann, D. F., and R. A. Kohl. 1980. Fluid dynamics of sprinkler systems. In Design and Operation of Farm Irrigation Systems, 583-618. M. E. Jensen, ed. St. Joseph, Mich.: ASAE. Heermann, D. F., H. H. Shull, and R. H. Mickelson. 1974. Center pivot design capacities in eastern Colorado. J. Irrig. Drain. Eng. 110(2): 1127-141. Heermann, D. F., and K. M. Stahl. 2006. CPED: Center Pivot Evaluation and Design. Available at: www.ars.usda.gov/Services/docs.htm?docid=8118. Howell, T. A., K. S. Copeland, A. D. Schneider, and D. A. Dusek. 1989. Sprinkler irrigation management for corn-southern Great Plains. Trans. ASAE 31(2): 147160. Kay, M. 1983. Sprinkler Irrigation Equipment and Practice. London, UK: Batsford Academic and Educational Ltd. Keller, J., and R. D. Bliesner. 1990. Sprinkle and Trickle Irrigation. New York, N.Y.: Van Nostrand Reinhold. Keller, J., F. Corey, W. R. Walker, and M. E. Vavra. 1980. Evaluation of irrigation systems. In Irrigation: Challenges of the 80’s. Proc. Second Nat’l Irrigation Symp., 95-105. St. Joseph, Mich.: ASAE. Kincaid, D. C. 1982. Sprinkler pattern radius. Trans. ASAE 25(6): 1668-1672. Kincaid, D. C. 1996. Spraydrop kinetic energy from irrigation sprinklers. Trans. ASAE 39(3): 847-853. Kincaid, D. C., and D. H. Heermann. 1970. Pressure distribution on a center pivot sprinkler irrigation system. Trans. ASAE 13(5): 556-558. Kincaid, D. C., D. F. Heermann, and E. G. Kruse. 1969. Application rates and runoff in center pivot sprinkler irrigation. Trans. ASAE 12(6): 790-794. Kincaid, D. C., and T. S. Longley. 1989. A water droplet evaporation and temperature model. Trans. ASAE 32(2): 457-463.

630

Chapter 16 Design and Operation of Sprinkler Systems

Kincaid, D. C., K. H. Solomon, and J. C. Oliphant. 1996. Drop size distributions for irrigation sprinklers. Trans. ASAE 39(3): 839-845. Kohl, R. A. 1972. Sprinkler precipitation gage errors. Trans. ASAE 15(2): 264-265, 271. Kohl, R. A., and D. W. DeBoer. 1984. Drop size distributions for a low pressure spray type agricultural sprinkler. Trans. ASAE 27(6): 1836-1840. Kohl, K. D., R. A. Kohl, and D. W. DeBoer. 1987. Measurement of low pressure sprinkler evaporation loss. Trans. ASAE 30(4): 1071-1074. Kohl, R. A., R. D. von Bernuth, and G. Heubner. 1985. Drop size distribution measurement problems using a laser unit. Trans. ASAE 28(1): 190-192. Kranz, W. L., and D. E. Eisenhauer. 1990. Sprinkler irrigation runoff and erosion control using inter-row tillage techniques. Applied Eng. Agric. 6(6): 739-744. Levine, G. 1952. Effects of irrigation droplet size on infiltration and aggregate breakdown. Agric. Eng. 33(9): 559-560. Li, J., H. Kawano, and K. Yu. 1994. Droplet size distributions from different shaped sprinkler nozzles. Trans. ASAE 37(6): 1871-1878. Lundstrom, D. R., and E. C. Stegman. 1988. Irrigation scheduling by the checkbook method. Extension Circular No. AE-792. Fargo, N.D.: North Dakota State Extension Service. Lyle, W. M. 1994. LEPA defined: More control with less consumption. Irrig. J. 44(6): 8,11. Lyle, W. M., and J. P. Bordovsky. 1981. Low energy precision application (LEPA) irrigation system. Trans. ASAE 24(5): 1241-1245. Lyle, W. M., and J. P. Bordovsky. 1983. LEPA irrigation system evaluation. Trans. ASAE 26(3): 776-781. Martin, D. L. 1991. Effect of frequency on center pivot irrigation. In Proc. Nat’l Conf. Irrigation and Drainage, 38-44. ASCE. Moldenhauer, W. C., and W. D. Kemper. 1969. Interdependence of water drop energy and clod size on infiltration and clod stability. Soil Sci. Soc. America Proc. 33: 297301. Morgan, R. M. 1993. Water and the Land: A History of American Irrigation. Washington, D.C.: The Irrigation Association. Nderitu, S. M., and D. J. Hills. 1993. Sprinkler uniformity as affected by riser characteristics. Applied Eng. Agric. 9(6): 515-521. Oakes, P. L., and E. W. Rochester. 1980. Energy utilization of hose towed traveler irrigators. Trans. ASAE 23(5): 1131-1138. Oliveira, C. A. S., R. J. Hanks, and U. Shani. 1987. Infiltration and runoff as affected by pitting, mulching and sprinkler irrigation. Irrig. Sci. 8: 49-64. Pair, C. H., W. H. Hing, K. R. Frost, R. E. Sneed, and T. J. Schiltz. 1983. Irrigation. 5th ed. Washington, D.C.: The Irrigation Association. Rochester, E. W., C. A. Flood, Jr., and S. G. Hackwell. 1990. Pressure losses from hose coiling on hard-hose travelers. Trans. ASAE 33(3): 834-838. Rochester, E. W., S. G. Hackwell, and K. H. Yoo. 1989. Pressure vs. flow control in traveler irrigation evaluation. Trans. ASAE 32(6): 2029-2034. Rochester, E. W., and S. G. Hackwell. 1991. Power and energy requirements of small hard-hose travelers. Applied Eng. Agric. 7(5): 551-556.

Design and Operation of Farm Irrigation Systems

631

Scaloppi, E. J., and R. G. Allen. 1993. Hydraulics of center pivot laterals. J. Irrig. Drain. Eng. 119(3): 554-567. Seginer, I. 1965. Tangential velocity of sprinkler drops. Trans. ASAE 8(1): 90-93. Seginer, I., D. Kantz, and D. Nir. 1991a. The distortion by wind of the distribution patterns of single sprinklers. Agric. Water Mgmt. 19: 341-359. Seginer, I., D. Nir, and R. D. von Bernuth. 1991b. Simulation of wind-distorted sprinkler patterns. J. Irrig. Drain. Eng. 117(2): 285-306. Solomon, K., and J. C. Bezdek. 1980. Characterizing sprinkler distribution patterns with a clustering algorithm. Trans. ASAE 23(4): 899-906. Solomon, K., and M. Kodoma. 1978. Center pivot end sprinkler pattern analysis and selection. Trans. ASAE 21(5): 706-712. Steiner, J. L., E. T. Kanemasu, and R. N. Clark. 1983. Spray losses and partitioning of water under a center pivot sprinkler system. Trans. ASAE 26(4): 1128-1134. Stillmunkes, R. T., and L. G. James. 1982. Impact energy of water droplets from irrigation sprinklers. Trans. ASAE 25(1): 130-133. Thompson, A. L., J. R. Gilley, and J. M. Norman. 1993. A sprinkler water droplet evaporation and plant canopy model: I. Model development. Trans. ASAE 36(3): 735-741. USDA-NASS (U.S. Department of Agriculture National Agricultural Statistics Service). 2003. Farm and Ranch Irrigation Survey and the 2002 Census of Agriculture. National Agricultural Statistics Service (NASS), Agricultural Statistics Board, USDA. von Bernuth, R. D. 1983. Nozzling considerations for center pivots with end guns. Trans. ASAE 26(2): 419-422. von Bernuth, R. D., and J. R. Gilley. 1985. Evaluation of center pivot application packages considering droplet induced infiltration reduction. Trans. ASAE 28(6): 1940-1946. von Bernuth, R. D., D. L. Martin, J. R. Gilley, and D. G. Watts. 1984. Irrigation system capacities for corn production in Nebraska. Trans. ASAE 27(2): 419-424, 428. Vories, E. D., and R. D. von Bernuth. 1986. Single nozzle sprinkler performance in wind. Trans. ASAE 28(6): 1940-1946. Wilmes, G. J., D. L. Martin, and R. J. Supalla. 1993. Decision support system for design of center pivots. Trans. ASAE 37(1): 165-175.

CHAPTER 17 MICROIRRIGATION SYSTEMS Robert G. Evans (USDA-ARS, Sidney, Montana) I-Pai Wu (University of Hawaii, Honolulu, Hawaii) Allen G. Smajstrala (University of Florida, Gainesville, Florida) Abstract. Microirrigation, the slow rate of water application at discrete locations at low pressures, includes trickle or surface drip, subsurface drip, microsprinklers and bubblers. It has made tremendous strides in the past three decades, and has become the modern standard for efficient irrigation practices for water conservation and optimal plant responses. Microirrigation is an extremely flexible set of technologies that can be economically used on almost every crop, soil type and climatic zone, but it requires a high level of management. These particular systems and their unique equipment and components have specialized needs and problems. This chapter discusses many of the advantages as well as disadvantages of various microirrigation technologies and their applications to horticultural and agronomic crops. Water quality concerns, filtration and management are addressed in detail Keywords. Bubblers, Design, Drip, Irrigation, Management, Microsprinklers, Trickle.

17.1 INTRODUCTION The development of modern drip irrigation technologies in the 1960s marked a significant step in the history of irrigation science and technology. The first attempts were plagued with problems; however, most of these have been solved and almost all aspects of microirrigation have greatly matured since that time, especially in the areas of filtration, water treatment, and emitter technology. Bucks (1995) has provided a knowledgeable and concise summary of the history of microirrigation for those who would like additional information. Microirrigation includes any localized irrigation method that slowly and frequently provides water directly to the plant root zone. Drip irrigation, trickle irrigation, bubblers, localized small microsprinklers, microspinners, and microsprayers are included in the general term. The slow rate of water application at discrete locations with associated low pressure and the irrigation of only a portion of the soil volume in the field can result in water delivery systems at relatively low cost, as well as reduce water diversions, compared to other irrigation methods.

Design and Operation of Farm Irrigation Systems

633

Microirrigation offers the potential for precise, high level management and is an extremely flexible irrigation method. It can be adapted to almost any cropping situation and climatic zone. Microirrigation can be used over a wide range of terrain conditions, and it has allowed expansion of irrigated crop production into areas with problem soils (such as either very low or very high infiltration rates) and poor water quality that could not be used with other irrigation methods. It can be installed as either a surface or subsurface water application system. Microirrigation can be used on most agricultural crops, although it is most often used with high-value specialty crops such as vegetables, ornamentals, vines, berries, olives, avocados, nuts, fruit crops, and greenhouse plants. In many cases, it can be economically used for field crops, golf greens, fairways, cotton, and sugarcane. However, the requirements for appropriate designs and management in humid areas can be considerably different from those in arid areas and the technology and techniques suitable in one area may not work in the other. Microirrigation will not be the most appropriate or economical irrigation method in all situations. The use of microirrigation is rapidly increasing around the world, and it is expected to continue to be a viable irrigation method for agricultural production in the foreseeable future. With increasing demands on limited water resources and the need to minimize environmental consequences of irrigation, microirrigation technology will undoubtedly play an even more important role in the future. Microirrigation provides many unique agronomic and water and energy conservation benefits that address many of the challenges facing irrigated agriculture. Farmers and other microirrigation users are continually seeking new applications, such as wastewater reuse, that will continue to provide new challenges for designers and irrigation managers. Any irrigation system must be compatible with cultural operations associated with a specific crop. Adoption of microirrigation may require new or innovative adaptations to various cultural practices and even the development of new harvest and tillage equipment. For example, surface lateral lines can hinder traditional harvest operations, requiring pre-harvest removal of the tubing or development of a new harvester and harvesting techniques. Lateral lines can be buried but this generally requires moving to minimal-tillage or permanent bed systems. An in-depth understanding of the unique benefits and limitations of microirrigation systems is needed to successfully design and manage these systems. As with all other irrigation methods, there are trade-offs with both positive and negative impacts on irrigation scheduling, efficiency, and uniformity, as well as environmental impacts, crop responses, and economics. 17.1.1 Advantages and Disadvantages of Microirrigation Microirrigation has advantages as well as disadvantages to consider and understand before adopting the technology. Advantages include water conservation and reduced deleterious water quality impacts due to high application efficiencies, automation capabilities, improved or increased yields, ease of chemical applications, and potential sustainability. Disadvantages include a high potential for emitter plugging, high system costs, and required high levels of management. 17.1.1.1 Advantages. Microirrigation is commonly used in areas with limited water and high water costs, but it has great value in other areas as well. Properly designed, installed, and managed microirrigation systems can eliminate surface runoff and associated soil erosion, efficiently and uniformly apply water-soluble fertilizers,

634

Chapter 17 Microirrigation Systems

and achieve high uniformity and efficiency of water application. They generally tend to have smaller wetted areas, reduced deep percolation, and lower evaporation losses than other irrigation methods. There can be water and chemical savings because of increased efficiency, reduced weed control costs because a limited surface area is wetted, and better productivity can be achieved due to improved control of water and nutrients in the root environment. Microirrigation generally has high production efficiencies whether expressed as yield per unit water, yield per unit nutrient input, or yield per unit land area. Advanced cultural practices such as the use of plastic or sheet paper mulches to reduce weed growth, heat soils, and decrease soil evaporation are also facilitated by drip irrigation. Due to relatively small pipe and valve sizes, microirrigation systems are easily and inexpensively automated, which reduces labor costs and improves general management flexibility. Because microirrigation methods can apply water in small amounts that nearly match evapotranspiration, soil characteristics such as hydraulic conductivity and water-holding capacity are usually not limiting. Less salt may be applied with the irrigation water because less water is needed with these potentially highly efficient systems. Low soil matric potentials reduce salinity hazards, improve the ability to manage saline or sodic soils, and permit the use of poorer-quality water than can be used with other irrigation methods. Because of its potential to be highly efficient, microirrigation is often specified as a best management practice for reducing groundwater contamination due to irrigation. The ability to precisely manage soil water deficits and to make prescription applications of nutrients and other chemicals through the irrigation system often leads to increased yields. In general, a body of research has consistently shown increases in vegetative growth and yields compared to more traditional irrigation techniques as a function of scheduling and management. Perennial crops may also experience more rapid growth and earlier production under microirrigation because water and nutrient stresses on young plants with small root zones may be greatly reduced. Uniformity in plant growth across a field, due to uniform water and nutrient distributions, also contributes to overall yield increases. Microirrigation must be managed as both a water and a nutrient application system. Fertilizers and other water soluble chemicals such as pesticides (e.g., nematicides, systemic insecticides, herbicides) and soil amendments (e.g., acids, polymers, powdered gypsum) can be efficiently and effectively applied through microirrigation systems. Buried drip irrigation systems are particularly amenable to the application of soil fumigants as well as other chemicals that tend to be fixed by the soil particles (e.g., some pesticides and phosphorus fertilizers). If designed and managed properly, microirrigation systems can reduce off-site impacts of irrigation on wildlife habitat and aquatic ecosystems compared to other methods. Pesticide use is often reduced because the efficacy of systemic pesticides is enhanced. In arid areas, herbicide expenses are usually less because only a portion of the area is wetted so weed germination is reduced; and because the soil and plant canopy are generally drier, there is often lower fungal disease pressure and fungicide use is generally less (Scherm and van Bruggen, 1995). Plastic films (biodegradable and nonbiodegradable), large sheet paper, and other mulches often work very well in drip irri-

Design and Operation of Farm Irrigation Systems

635

gated crop culture to control weeds (and eliminate herbicide use) and reduce soil evaporation losses. Microirrigation systems may enhance long-term sustainability of an agricultural operation because of their potential for maximizing water application efficiencies and minimizing chemical applications. Limited water supplies (quality and quantity) can be utilized more efficiently for agricultural production, thus providing more water for competing uses or reducing withdrawals from aquifers. In addition, microirrigation methods are low-pressure systems that typically use less total energy compared to sprinklers. 17.1.1.2 Disadvantages. Because of their relatively small orifice sizes, microirrigation emitters can be easily plugged due to physical, chemical, and biological factors. Clogging adversely affects uniformity, and can negate the benefits and effectiveness of microirrigation. Microirrigation systems are generally expensive to install and maintain but are similar in costs to most other advanced irrigation methods. For larger systems (e.g., >10 ha) with relatively close plant spacing, their cost is comparable to permanent solid set sprinkler systems covering the same area although the filtration and chemical treatment systems are major expenses that can vary widely depending on conditions and system size. High-density plantings requiring large amounts of tubing may not be economical. Operational costs will be high due to the need for chemical treatment, filtration, and labor for routine flushing of lines, although lower energy costs and water savings may offset some of this increase. There can also be significant costs associated with the retrieval and disposal of tape/tube and non-biodegradable plastic mulches. A high level of management is required to operate and maintain a microirrigation system. Managers require a greater level of training and proficiency than for surface or sprinkler systems. They command higher salaries and are usually employed yearround because of the need to retain their skills, however, they can generally cover three to four times as much cropped area as an irrigator using more traditional methods, primarily due to automation. The higher level of management also requires adoption of ancillary technologies (with their associated costs), such as irrigation scheduling, soil water monitoring, and frequent detailed plant tissue nutrient analysis for fertigation programs. As a general rule, microirrigation systems are less forgiving of mismanagement or poor design than methods that irrigate a much larger portion of the root zone. These problems range from overirrigation and excessive leaching of chemicals to severe drought, salinity, or nutrient stresses. Uneven distributions of water, nutrients, and roots across a field can create problems unique to microirrigation. The restricted wetted soil volume may affect the extent of the rooting system and the physical stability of a plant. Smaller rooting volumes also limit the amount of soil water available to buffer the plant against drought in the event of an irrigation system failure. In addition, the small wetted soil volumes increase the difficulty of maintaining an optimally balanced soil nutritional status because access to nutrients stored in adjacent nonirrigated soils is limited. Pest problems may change because the frequent irrigation may create environmental and moisture conditions favorable to fungal diseases or pests that may not be concerns under other irrigation technologies (e.g., mites that favor dry, dusty condi-

636

Chapter 17 Microirrigation Systems

tions). Frequent irrigations may also create optimal conditions for some plant diseases requiring special management (e.g., fumigation before and/or after the growing season to minimize inoculum). Polyethylene microirrigation tubing can be physically damaged by a number of mechanical and natural causes. Damage by farm equipment commonly occurs. Coyotes, rodents, and other animals may damage tubing, especially when looking for water in arid areas. Gophers and other rodents may cut through buried tubing as they burrow. Woodpeckers have been reported to peck holes in tubing. Insects and spiders can plug emitters, but may also enlarge orifices when searching for water. Tall grass, weeds, spider webs, and large insects can stop the rotation of microspinners. 17.1.2 Soil, Water Quantity, and Water Quality Considerations 17.1.2.1 Soils. A microirrigation system must be designed and managed to match the soils on which it is used. Deep sandy soils often have little lateral spreading of water requiring several small irrigations each day and/or microsprinklers to expand the wetted root volumes. Improper scheduling due to poor system design or management can result in excessive deep percolation and leaching of nutrients. When application rates exceed infiltration capacity, soils become saturated, weeds and other problems may be enhanced due to large wetted areas, and runoff may even occur. In addition to environmental pollution considerations, soil waterlogging can result in increased plant disease and induce plant physiological disorders. Soil salinity will affect system design and management because salts accumulate at the edges of wetted areas and on the soil surface. Deficit irrigation may lead to excessive salt levels in the soil profile. These salts need to be periodically leached, which can be complicated by the development of preferential flow paths. The use of plastic film mulches that reduce soil evaporation have also been found to reduce soil salinity directly under the mulches. Injection of acids (e.g., sulfuric) may sometimes be required to increase the solubility of salts to facilitate leaching. Maximum leaching of salts will occur near the emitters, with effectiveness decreasing with distance from the emitters. Seed germination of some crops (e.g., lettuce) may require sprinkler irrigation to move salts below the seed bed. In areas of low rainfall, it may be necessary to sprinkler irrigate periodically to drive salts below the tubing depth. Another salinity-control technique is to irrigate during rains, pushing salts to the outside of the wetted volume. Bed shaping—forming beds higher than necessary—has also been used. The drip system is then operated to push the salts to the surface. The salty surface soils are then scraped to the side and the crop planted in the less-saline soil lower in the bed. A complete soil chemical analysis should be conducted as an initial part of the planning process, for water application as well as for cultural decisions. Salination and/or changes in soil pH may also develop because of the quality of the water supply or as a result of various water treatment and chemical/fertilizer management programs. Soil pH can have major effects on the availability of soil nutrients to the plants and in some cases can cause toxicity. Soil amendments (e.g., gypsum or lime) should be applied prior to planting to ameliorate existing or anticipated problems, although some supplemental gypsum can be injected through the microirrigation system. Producers should annually monitor the soil chemistry in the rooting volume throughout the life of the irrigation system.

Design and Operation of Farm Irrigation Systems

637

17.1.2.2 Water quantity. Timing, availability, and reliability of long-term water supplies must be determined. Seasonal irrigation requirement depths for many crops can range from 100 to as much as 2000 mm. Canal and on-farm delivery systems in many areas of the world are often designed to satisfy the requirements of surface or sprinkler irrigation (e.g., 1.0 L s-1 ha-1 based on the farm’s total irrigated area) or deliveries may be based on calendar rotations (e.g., every 7 days) that are inappropriate for microirrigation designs. ET may be higher with microirrigation due to reduced drought stress, although most of the time it will be less than under other irrigation methods due to reduced soil evaporation losses. Supplemental wells and storage ponds may be required to utilize microirrigation technologies effectively under some conditions. 17.1.2.3 Water quality. Physical, biological, and chemical water quality, including salinity, is a major concern in the management of all microirrigation systems (Nakayama and Bucks, 1991; Lamm et al., 2000). The physical, biological, and chemical characteristics of the water supply from all sources (e.g., wells, canals, reuse ponds) must be considered. The potential for emitters to become plugged by physical, chemical, or biological contaminants can create significant problems. Success hinges on filtering and treating the water to match actual water quality conditions throughout the year with both surface and groundwater. It is sometimes not economically feasible to treat a water source to make it suitable for microirrigation and other irrigation methods should be considered. A successful water treatment program must accommodate worst-possible conditions while meeting high microirrigation water quality standards. The potential for soil salination due to the water supply must be assessed. Fertigation and injection of other chemicals require knowledge of the water chemical constituents to ensure compatibility between injected chemicals as well as to help determine suitable chemical water treatment needs and procedures. Laboratory tests are necessary to determine the nature and composition of inorganic contaminants, as well as the relative proportions of each that may create significant problems in the long-term management of the system or affect the crop’s utilization of water and nutrients. Specific concerns may be pH, salinity (electrical conductivity), calcium, magnesium, sodium, iron, manganese, carbonates, bicarbonates, and sulfur. Organic contaminants may sometimes be problematic, but these are usually controlled by good filtration and chlorine treatments. A suitable treatment program may consist of several progressive steps or phases including: settling basins, gravity screens, centrifugal separators, screen filters, disk filters, and/or media filters, plus the injection of chlorine, acids, or other water treatment chemicals. Appropriate design and management of each stage of the treatment system provides the capability of maintaining high water quality standards throughout the life of the project despite variations in the physical and chemical properties over time. More specific and detailed information on water quality concerns related to microirrigation is discussed in several sections later in this chapter 17.1.3 Environmental Considerations Substantial environmental advantages can result from properly designed, maintained, and managed microirrigation/chemigation systems. Environmental advantages result from reduced diversions of water, reduced chemical usage, and reduced groundwater contamination by reducing leaching of salts and other chemicals below the root zone.

638

Chapter 17 Microirrigation Systems

Although microirrigation can achieve highly uniform water applications, it is in fact always less than perfectly uniform. As a result, some portions of the field will be overirrigated and some will be underirrigated. Overirrigation will waste water and cause deep seepage which may contaminate underlying groundwater resources. Deep seepage losses occur due to overestimating ET causing excessive applications; nonuniformity of irrigation; overirrigation because of poor scheduling or lack of automation and feedback; and leaching for soil salinity management. Underirrigation will result in a yield reduction and may cause undesirable soil salinity accumulations. Improper application of some pesticides and fumigants may negatively affect beneficial soil biota, including earthworms, bacteria, fungi, and insects. Disposal of flush water from filters and lateral lines may sometimes be a problem, especially if pesticides or fertilizers are in the effluent. 17.1.4 Economic Considerations The ultimate goal of any agricultural activity is to achieve maximum net economic return. The irrigation system uniformity, irrigation scheduling practices, cost of water, yield price, yield reductions by deficit irrigation, and damage caused by overirrigation, including possible groundwater contamination, are all very important factors affecting the economic return from a microirrigated production system. Expected economic returns and the required system uniformity can also affect emitter selection.

17.2 MICROIRRIGATION SYSTEMS 17.2.1 Methods Microirrigation methods are generally defined by the water emission device. Emission devices range from thin-wall plastic tube with simple orifices, microsprinklers, orifices and long-path laminar flow emitters and microtubing, to more elaborate and efficient turbulent-path and pressure-compensating emitters. Some emission devices are manufactured as an integral part of the plastic tubes and tapes while others are attached during installation. Surface applicators include emitters (drippers), microsprinklers/ microsprayers, and bubblers, all of which apply water on or above the surface of the soil. Subsurface drip involves the use of point-source emitters or line-source emitter tubing and tapes to apply water below the soil surface at depths depending on the soil type and crop. Surface and subsurface drip have also been used for water table control in some humid areas as a variation of subirrigation, primarily on vegetable crops. Drippers and bubblers are designed to apply water at or slightly above atmospheric pressure, whereas microsprinklers apply water from about 70 to more than 250 kPa. Two general categories of microirrigation laterals are polyethylene tape and tubing. Tapes are collapsible, thin-walled, low-pressure polyethylene tubes with built-in emitters or orifices. Tubing is more rigid and more expensive than tape, has thicker walls, and may or may not have pre-installed emitters. Tapes and tubing may also be divided into five classes depending on use: (1) disposable, thin-walled surface tape (1-year life); (2) shallow, buried tapes (1-5 year life); (3) reusable/retrievable surface tapes (1to 3-year life); (4) retrievable surface tubing (multi-year life); and, (5) buried tubing (multi-year life). Tapes are most commonly used on annual or seasonal row crops while tubing is used more often on perennial crops. ASAE Standard S435, Polyethylene pipe used for microirrigation laterals, presents manufacturing and testing requirements for tubing (ASAE, 2005c).

Design and Operation of Farm Irrigation Systems

639

Tapes typically have wall thicknesses ranging from 0.1 mm to 0.64 mm and inside diameters may range from 9.5 to 28.6 mm. Emitters usually have close, uniform spacing (e.g., 3.5-60 cm) along a lateral line and emitters are simple orifices, long path, labyrinth flow paths. or a combination. Emitters may be embossed within the welded seam of the tape or they may be separate, pre-molded devices installed during fabrication. Currently, emitters on tapes are not pressure compensating, and water discharge rates of individual tape emitters range from 0.2 L h-1 to over 7.6 L h-1. Tape operating pressures range from 20 kPa to about 140 kPa. They should always be installed with the emitters facing up because of plugging problems due to sediment accumulations in the bottom. Tubing has wall thicknesses typically ranging from 0.25 mm to over 0.9 mm with inside diameters from 9 mm to over 35 mm. Pre-installed emitters on tubing have uniform spacing, however, in contrast to tapes, point-source emitters and microsprinklers can be installed in the field at any spacing on the tubing to meet specific irrigation requirements. Emitters are either non-pressure compensating or pressure compensating with water discharge rates from 1.5 to over 20 L h-1. Microsprinkler discharge rates range from about 5 to over 40 L h-1. Operating pressures range from about 40 to over 250 kPa. 17.2.1.1 Drip and microsprinkler emitters. Water distribution by drip and microsprinkler emitters can be characterized as line-source or point-source applications for both tapes and tubing. Line sources apply water in a continuous or near-continuous pattern along the length of the lateral. In this category are soaker hoses or porous pipes (line-source emitters) in which the entire pipe wall is a seepage (and filtration) surface, as well as drip tapes with closely spaced (e.g., 15-30 cm) emission points whose water application patterns overlap. Point-source emitters can be grouped based on their flow characteristics into long-path emitters (microtubing, laminar-flow, and turbulent-path emitters), short-path emitters (microsprayers and other orifice emitters), orifice-vortex emitters, and pressure-compensating emitters. These devices apply water at discrete points and overlap between wetting patterns may or may not occur, depending on emitter spacing, irrigation duration, and emitter flow rate. Orifices and microtubing emitters are the two simplest emission devices. They were common in the early development of drip irrigation, but are currently only used on tapes. Plugging is usually a serious problem for orifice emitters due to small (less than 0.3 mm) outlet diameters and low discharge rates. Orifice-vortex emitters are orifice emitters that have been modified so that water enters the emitter with an angular velocity such that the circular vortex motion provides additional energy loss so that the orifice can be larger and less prone to plugging. Microtubing is a long-path emitter inserted into the lateral line. Different lengths of microtubing can produce various flow rates depending on their dimensions and water pressure. The size of microtubing typically ranges from 1 to 10 mm in diameter, and the flow characteristics can be either laminar or turbulent, as a function of tube size. The laminar-flow (small-diameter) microtubing tends to accumulate small deposits and is quite susceptible to partial plugging. The long-path or spiral emitter is basically a microtubing emitter that is wrapped around a short, larger plastic tube to make a more compact unit. Larger-sized microtubing with turbulent flow produces a bubbler effect and usually has few plugging problems (Rawlins, 1977).

640

Chapter 17 Microirrigation Systems

The labyrinth emitter, designed with long intricate passageways, will create turbulent flow at normal operating pressures and is often called a turbulent or turbulent-path emitter. The turbulent flow resists plugging by allowing the flow path to be as large as possible and preventing small particles from settling or becoming lodged in the passageway. Flow rates from turbulent path emitters are also relatively insensitive to temperature fluctuations (Wu and Phene, 1984; Rodriguez-Sinobas et al., 1999), thus avoiding a major cause of nonuniform water application under field conditions. Currently, most point-source emitters utilize turbulent-flow paths to control the application of water from tubing and some tapes. Emitters can be inserted or molded into the tubing or tape during the manufacturing process. With internal “in-line” emitters, there are no protrusions to interfere with mechanical installation or retrieval of the tubing or tape. Alternatively, emitters (and microsprinklers) can be attached to the outside of the tubing when the system is installed, usually by manually punching a hole and inserting the barbed end of the emitter. This procedure requires more labor but it allows a system to be customized to match the needs of widely or unevenly spaced plants. Microsprinkler or minisprinkler emission devices are generally simple orifices and include small, low-pressure minisprinklers, foggers, spitters, jets, and sprayers that are installed in the field on tubing. These typically apply water (at 35 to 70 L h-1) to larger areas than drip emitters, but do not uniformly cover the entire cropped area. They are used to irrigate tree crops, shrubs, widely spaced plantings, and localized grass areas with extensive root systems, especially on sandy soils where lateral movement of soil water is limited by soil hydraulic properties or other areas with greatly restricted root zone depths. Nozzle sizes typically range from 0.5 mm to 2 mm; plugging problems are greatly reduced with nozzle sizes larger than 0.75 mm (Wu et al., 1991) combined with adequate filtration and chemical treatment of the water. Microsprinklers are installed after the lateral tubing or pipe has been laid in the field. They may be inserted with barbed fittings directly into the tube but are more commonly mounted on stake assemblies and connected to the lateral lines with 4-6 mm tubing. Some may also be mounted directly on threaded PVC fittings on the lateral. The state of the art for microsprinklers is advancing rapidly and improved microsprinklers (e.g., pressurecompensating and self-cleaning) are being developed and tested. A variation of the microsprayer pulses the water jet in short bursts of up to 60-70 cycles per minute, which serves to minimize application rates while maximizing the wetted radius. These can be an advantage on heavy soils with low infiltration rates or soils where poor lateral water movement may be a concern. Both drip emitters and microsprinklers are available as pressure-compensating devices. These use a flexible orifice that changes its diameter depending on the pressure, thereby regulating the flow. Pressure-compensating devices are used to provide uniform flows from each emitter along a lateral whenever elevation differences or excessive pressure losses to long lateral lengths cause flow variations to exceed design standards. However, these devices are more costly than standard emitter devices. 17.2.1.2 Low-head bubbler irrigation systems. Bubblers are large-orifice, lowpressure emitters that apply water at discrete points but at considerably higher rates than common drip or microsprinkler emitters. Filtration requirements are greatly reduced, but flow rates are often so high that basins or very flat terrain may be required to prevent runoff. Some bubblers are designed to operate on gravity flow or low-head,

Design and Operation of Farm Irrigation Systems

641

high-flow pumps (1 to 8 m of head), while others use pressurized (e.g., 200 kPa) pipelines to distribute water. The higher-pressure bubbler systems use special devices with large openings where flow rates can be mechanically adjusted at each location. Low-head bubbler irrigation systems use microtubing emitters to deliver water to the crop. These systems consists of a main line connected to a water source, a constanthead device, manifolds, laterals, and small-diameter (e.g., 4 mm to 26 mm inside diameter microtube) delivery hoses. Laterals are usually buried and laid between rows. A sufficiently long, large-diameter (e.g., 5 to 25 mm) delivery hose is inserted into the lateral pipe to deliver water to a plant. The delivery hoses are anchored to a tree or stake, and the outlet elevations are adjusted to the hydraulic energy gradient so that water flows or “bubbles” from all hoses at equal rates. Bubblers are well-suited for economical irrigation of trees and vine crops and are being developed for turf and landscape applications. Bubbler systems do not usually require elaborate pumping and filtration systems, but are not widely adopted (Yitayew et al., 1995). Design considerations and installation of low-head bubbler systems are discussed by Yitayew et al. (1995), Yitayew et al. (1999), Thorton and Behoteguy (1980), and Rawlins (1977). 17.2.2 Wetting Patterns The applied water moves through the soil largely under unsaturated flow conditions at the wetting front. The distribution of water and the shape of the wetted volume can be predicted from the physical laws of capillary movement for either point sources or line sources (Warrick and Lomen, 1983; Clothier et al., 1985; Philip, 1991; Or and Coelho, 1996; and many others). A point-source emitter will provide a wetted volume in the soil, which is affected by the initial soil water content, emitter flow rate, irrigation frequency and duration, capillary movement of water and the water-holding capacity of the soil. In arid areas, the emitter creates wetting patterns in the soil that determine the size and shape of the crop root zone. “Point sources” refer to individual emitters with discrete application points. A point-source emitter or a group of emitters forming a point source are generally used for tree crops or other widely spaced plantings. Microspray emitters with large spacing such that their wetting patterns do not overlap are also point-source emitters. Even groups of emitters with overlapped wetting patterns but designed as a unit, such as around individual trees, can be considered to be a point source. “Line source” wetting patterns develop when emitter applications along a lateral merge and form a half-cylinder wetting pattern or trough of wetted soil in the field. High-density row crops are usually irrigated with line sources by lateral lines with closely spaced emitters or microsprayers. The wetting pattern for a point-source emitter in a homogeneous soil is a threedimensional hemispherical shape with a water gradient from the center (point source) to the edge of the sphere. The wetting pattern for a line-source application by closely spaced emitters will form a two-dimensional half-cylinder shape in homogeneous soil. In layered soils, the wetting patterns will tend to be confined within the top layer so the bottom of the hemisphere or half-cylinder will be relatively flat and form a wetting pattern shaped like a disk or rectangle. Hardpans in many soils have smaller pore spaces than the material above or below. The hard pan serves as a barrier because capillary water movement does not readily

642

Chapter 17 Microirrigation Systems

occur from smaller to the larger pores below the hardpan. Likewise, where a fine soil is underlain by a coarser material, irrigation must saturate the upper soil before water will enter the coarser layer.

17.3 DESIGN FACTORS 17.3.1 General Considerations The design of a water application system will determine the maximum potential performance level for any proposed crop use, whereas management dictates the actual benefits received and the magnitude of any ecological impacts. High-quality installations are more easily maintained and are much less expensive to operate over time than a substandard design that requires frequent repairs and has a shorter operational life. Minimum requirements for the design, installation, and performance of microirrigation systems are presented in ASAE Engineering Practice EP405.1, Design and Installation of Microirrigation Systems (ASAE, 2005b). The first rule of microirrigation design is the same as for all irrigation systems: keep it as simple as possible. The system must be designed to meet the users’ level of expertise and it must fit within their perceived needs and cultural practices. It must be reliable and sustainable, able to manage salts, easy to maintain, and allow for needed tillage and harvest operations. The design and installation must be site-specific. They are governed by soil type and depth distributions, topography, climate, water quality, water quantity, the proposed crops and cropping systems, as well as the preferences of the irrigator. However, the fundamental aspects of high-frequency irrigations, limited wetted rooting volumes, filtration and chemical treatment of the applied water, and extraordinary consideration of the spatial uniformity of water applications per emission device are common to all microirrigation systems. Designs should facilitate maintenance. Ponds and chemigation installations should be fenced for safety of workers, children, and animals. Water treatment, filtration, and lateral line flushing must be high priorities. Due to low operating pressures and chemigation requirements, hydraulic variables are more rigorous for microirrigation systems than for other types of systems operating at higher pressures. Total system pressures should normally not be permitted to vary by more than 20% unless pressure-compensating emitters are used. The total allowable pressure loss of the whole system, which provides the desired design uniformities, is selected at the start of the design process and depends on the preferred pressure regulation strategy (e.g., optimal combinations of valves, pressurecompensating emitters, and topographic layouts). Lateral, subunit, submain, mainline, and control head system pressure losses are assigned so that the sum does not exceed the total system design criteria. As much as possible, the systems should be designed based on anticipated actual installed emitter discharges, which are often different from the manufacturer’s literature due to factors such as unit-to-unit variation in manufacturing, system elevation changes, system pressure variations, emitter wear, pressure losses in stake assemblies (microsprinklers), and varying lengths of small-diameter (e.g., 4- to 6-mm diameter) supply tubing from the lateral to the emission devices. The coefficient of flow variation of the emission devices should always be less than 10%. Distribution uniformities should normally be greater than 90%, especially when chemigation will be used. Pipelines (usually PVC) should be placed at sufficient depths to avoid damage from farm and construction equipment. Concrete blocks to prevent pipeline movement

Design and Operation of Farm Irrigation Systems

643

(thrust blocks) need to be placed at appropriate locations to prevent failure of pipelines, valves, and other components. Adequate air relief, vacuum breakers, and pressure relief valves must be sited appropriately to ensure proper operation. Information and procedures for installation of PVC pipe can be found in ASAE Standard S376.1, Design, Installation and Performance of Underground, Thermoplastic Irrigation Pipelines (ASAE, 2005a). Some emitters are also designed to facilitate air relief in the laterals used on highly variable topography. Designing for ease of maintenance is critical. It is important to ensure that there is easy access to all equipment and various system components for maintenance and replacement whether buried or on the surface. All aboveground components are typically at least 0.4 m but not more than 1.5 m above concrete or soil surfaces for ease of maintenance. All pipe nipples should have sufficient length for wrenches. Fenced areas should have provisions for equipment access. Valves should be installed to hydraulically isolate components requiring frequent cleaning, repair, or replacement, such as inline filters. Likewise, it should be possible to isolate blocks within a field for maintenance without shutting down the entire system. Unions, flanges, or Victaulic couplings should be provided for easy removal of the affected components. In areas of cold climate, there must be the capability to winterize the entire system, including drain valves, infiltration pits and, if necessary, provisions for using compressed air to remove water when pipelines do not have appropriate slopes for gravity drainage. Pumps, filters, flow meters, gauges, tanks, and valves may also require special fittings or removal for cold-temperature protection. Pumps and electrical panels should be protected from exposure to the sun by covers or shading to reduce heating and maximize their useful life. A lack of understanding of the fundamental benefits and limitations of microirrigation has resulted in many systems that are unintentionally under-designed. The most common signs of an under-designed system are the inability to fully provide for the water needs of the crop during peak water use periods and inadequate line flushing velocities. Operational flexibility may also be limited by a poor design. External factors such as soil salinity, soil hydraulics, crop sensitivity, water quantity, water quality, and any environmental concerns must be addressed from the beginning of the design process. These concerns will guide the selection of the tubing, emitters, and emitter spacing. 17.3.2 Field and Crop Considerations Microirrigation distributes water directly into the root zone of crops, so the selection of emitters must consider the rooting characteristics of the crop, the expected volume of soil to be wetted in the field, the total amount of water to be applied, and the estimated total allowable time per irrigation per day. The selection of emitters and spacing will be based on the maximum application amounts, estimated irrigation times, water supply considerations, and hydraulic capacities. Perennial crops may require one to five emitter laterals per plant row to adequately supply water needs depending on soil types, water emission device, size of plants and climate. Established, widely spaced plantings, such as pecan trees, should have at least two lines, 2 to 3 m on either side of the row. More closely spaced perennial crops, such as asparagus, grapes, and hops, may need only one lateral per row or bed. Plantings in humid regions may require more laterals and microsprinklers due to the extensive rooting systems that are stimulated by recurrent rainfall to ensure adequate deliv-

644

Chapter 17 Microirrigation Systems

ery of water and nutrients to the roots. Crops using line-source drip tapes, such as row crops and shallow-rooted or closely spaced perennials, are usually designed based on flow rate per unit lateral length. Widely spaced permanent crops using tubing with discrete emitters are usually designed using discharge per outlet. Point-source emitters with non-overlapping wetting patterns, including microsprinklers, are usually selected for low-density plantings of trees, vines, or vegetables. Emitter locations should ensure that the wetting patterns are within the plant’s root zone and that 25% to 50% of the potential root zone is irrigated. High-density row crops are generally irrigated by tapes or line-source lateral lines. Line-source systems should be designed so that the entire root system of high-density crops is within the wetting pattern. Microsprinklers may be required to increase the wetted area to maximize soil water availability and avoid leaching on light, highly permeable soils or areas with shallow root zones. Widely spaced permanent crops, such as citrus, may require microsprinklers to irrigate a relatively large fraction (e.g., > 50%) of the root zone for peak productivity, especially in humid areas. For the same lateral diameter, emitters with higher flow rates will have a larger pressure variation per lateral length compared to low-flow emitters. High flows per unit length of tubing will also limit lengths of runs. It is generally desirable to have the highest discharge rates that meet soil hydraulic conditions, because higher-flow tapes and emitters have larger orifices and are less subject to plugging and thus provide higher uniformity and maximum operational flexibility in scheduling. Buried drip systems are strongly affected by the saturated hydraulic conductivity of the soil and emitter flow rates are selected depending on whether the grower wants to be able to wet the soil surface. 17.3.3 System Considerations System hydraulic capacity for irrigation should be based on peak evapotranspiration demands for the most critical period for a mature planting, usually in the range of 5 to 10 mm per day depending on the crop, climate, and application efficiencies. If economical and practical, a design should aim to supply about 120% of peak ET to provide the capacity to catch up in the event of maintenance down times, line breakage, equipment failures, electrical outages, or other problems. 17.3.3.1 Pipe systems. Main and submain lines utilizing PVC pipe that is not UVprotected should be buried. It is advisable to keep control valves above ground to facilitate maintenance and keep submains full of water to minimize system drainage and decrease startup times. Each block should have isolation valves so that it can be maintained without shutting down the entire system. Pipes, fittings, and valves should have sufficient pressure ratings to withstand waterhammer surges and static pumping heads. The proper size of mainlines, submains, headers, manifolds, and valves, as well as operating pressures, may be dictated by flushing requirements. Most systems will not have sufficient hydraulic capacity to flush the entire system at once. More often, the system will be flushed in zones, with other zones shut off so that sufficient pressure and flow will be available to flush each zone. Even small changes in elevation at the low pressures common to microirrigation can cause large flow variations. Pressures can be managed by proper pipe sizing, special valving, and/or carefully controlling elevation differences within blocks. Pressure

Design and Operation of Farm Irrigation Systems

645

regulation must be designed to ensure uniform water distribution to various parts of the microirrigation system and to reduce fitting failure due to excess pressure buildups. Non-pressure-compensating emitters and tapes must operate at consistent inlet pressures and be properly sized with appropriate run lengths and lateral diameters while considering field slope to optimize application uniformities. Inlet pressure regulation with very low-pressure systems (e.g., 25% in arid regions, >50% in humid regions). For high-density plantings such as row crops, the root systems merge along the row, and closely spaced emitters or line-source systems should be used to apply water uniformly along the row length. 17.3.6 Emitter Plugging Potential The characteristic low application rates, low pressures, and small orifice openings that are unique to microirrigation systems also create emitter plugging problems. Selection of the proper type and size of emitters will reduce the potential for plugging, although all emitter types can completely or partially plug. Less plugging will occur when irrigation water is applied with proper filtration and water treatment. When a number of emitters are considered as a unit, such as several emitters grouped to irrigate a single tree, the uniformity of water application will be much improved as compared to using a single emitter per tree (Wu et al., 1988b). System variability can be controlled within 10% when plugging is zero and at least two emitters are used in a group. The effect of plugging on water application uniformity in the field can be minimized by closer emitter spacing or emitter grouping (Bralts et al., 1987b). A study of contiguous, random plugging (Wu et al., 1991) showed that even with as much as 20% to 30% total plugging, only 1% of the plugging consisted of four to five consecutively

Design and Operation of Farm Irrigation Systems

647

plugged emitters. When the coefficients of hydraulic and manufacturer’s variation are both less than 10% and emitters are placed in groups of four per tree, 10% to 20% random plugging will still maintain an overall coefficient of variation (CV) of 17% to 25%, respectively (Wu, 1993a). For high-density row crops when the coefficients of hydraulic and manufacturer’s variation are less than 10% and the emitter spacing is half of the wetted diameter, a 10% to 20% plugging will produce an overall CV of about 20% to 30%, respectively. A superposition technique (Wu et al., 1989) was used to evaluate the spatial uniformity along the lateral by adding soil-water patterns from all emitters at various specified spacings. This work showed that plugging, followed by emitter spacing, were the most significant factors affecting the spatial uniformity. Wu (1993a) showed that emitter grouping was as significant as spacing. Both the hydraulic design and manufacturer’s variation were less significant than plugging, grouping, and emitter spacing as long as their individual coefficients of variation were less than 10%. Neither the soil-wetting patterns nor the plugging distributions were highly significant when soil-wetting patterns overlapped by 50% along a lateral. When the hydraulic design of a drip irrigation system is based on a 20% emitter flow variation, a 90% and 70% uniformity coefficient can be achieved for 0% and 20% plugging, respectively, as long as the emitter spacing is designed for 0.5 of its wetted soil diameter (Wu, 1993b).

17.4 HYDRAULICS OF EMITTERS AND EMITTER DESIGN VARIATION The basic relationship between emitter flow rate and pressure is given as: q = chx

(17.1)

where q = flow rate c = emitter discharge coefficient h = pressure head x = emitter discharge exponent. The x value is used in emitter selection because it characterizes the flow type in the emitter and varies from 1 to near zero. When x = 1 the emitter is a laminar flow emitter (capillary) whereas x will have a value around 0.85 for microtubes, 0.65 for long or spiral path emitters, 0.5 for fully turbulent flow emitters, about 0.4 for a vortex emitter, and near zero for a fully pressure-compensating emitter. For a given hydraulic variation, less flow variation will occur with turbulent flow emitters than laminar flow emitters. When the hydraulic design is based on a 20% pressure variation in the microirrigation system, fully turbulent flow emitters (x = 0.5) will produce only about 10% emitter flow variation while the laminar flow emitters (x = 1.0) will produce 20% emitter flow variation. 17.4.1 Hydraulics of Orifice Emitters The discharge rate from an orifice or a short-length nozzle is determined by the hydraulic pressure inside the lateral line at the orifice and the orifice dimensions. When the flow path is fixed and the flow cross-section area is constant, the emitter flow rate will be affected by only the hydraulic pressure. The flow rate from an orifice or nozzle type emitter can be theoretically expressed as: q = c1a 2 gh

(17.2)

648

Chapter 17 Microirrigation Systems

where q = emitter flow rate c1 = discharge coefficient a = cross-sectional area of flow g = acceleration due to gravity h = pressure head at the base of the emitter. When the cross-sectional area is constant, Equation 17.2 can be rearranged as a simple power function similar to Equation 17.1: q = C h0.5

(17.3)

where C is the emitter discharge coefficient and equal to c1a 2 g . 17.4.2 Hydraulics of Microtube Emitters A microtube is a small pipe. The hydraulics of microtubes is the same as for pipe flow. Thus, the equation for energy drop by friction in the microtube can be expressed by a simple power function: qm h=K n L (17.4) D where h = head loss due to friction and is also the pressure head at the inlet of a microtube K = emitter discharge coefficient q = emitter flow m = flow rate (q) exponent (m is 1 for laminar flow, 2 for complete turbulent flow and 1.75 for turbulent flow in a smooth pipe) D = inside diameter n = diameter exponent L = length of microtube. Microtube discharge can be determined by rearranging Equation 17.4 as: 1

⎛ n ⎞m q=⎜ D h⎟ (17.5) ⎝K L ⎠ For a given microtube emitter in which the length, L, and diameter, D, are fixed, the emitter flow and hydraulic pressure head is often presented as a simple power function (Equation 17.1) where C is a coefficient and is a constant, x = 1/m and is 0.5 for turbulent flow and 1 for laminar flow. Depending on turbulent flow conditions, x will be between 0.5 and 1.0. 17.4.3 Hydraulics of Long-Path and Labyrinth Emitters The relationship between flow rate, q, and pressure head, h, can be expressed by the same relationship given above for long-path (including spiral-path) and turbulent-flow emitters. The two power function coefficients, C and x for individual long-path and labyrinth emitter, are determined by hydraulic laboratory testing. Values for x will range between 0.5 and 1.0 but are typically 0.65 to 0.85. 17.4.4 Hydraulics of Pressure-Compensating Emitters When an emitter is designed so that the cross-section area decreases with respect to pressure, a = b h− y (17.6)

where a is the cross-sectional area of the emitter flow path, a nozzle, or microtube; b

Design and Operation of Farm Irrigation Systems

649

and y are coefficients. By introducing Equation 17.6 into Equation 17.2 and rearranging, the emitter flow and water pressure can be shown as: q = C2 h0.5− y

(17.7)

where C2 is a coefficient. Equation 17.7 shows that the exponent, 0.5 – y, can be smaller than 0.5, which indicates a reduced effect of pressure head on emitter flow as compared to a turbulent flow emitter. If the y value is 0.5, the exponent will be zero and no changes in emitter flow will occur. When this occurs, an increase of water pressure will cause a decrease in cross-sectional area of flow that exactly compensates for the increase in pressure, and the emitter will be fully pressure compensating. However, if y is greater than 0.5, the exponent will be negative and flow rate will decrease with increasing pressure. The concept of pressure compensation can also be applied to microtube emitters in a similar manner. 17.4.5 Emitter Flow Variation Microirrigation is characterized by frequent water applications at low application rates. Thus, even small variations in the magnitude of the emission device’s flow rate may cumulatively represent relatively large changes in the total seasonal water applications. Factors that may affect emitter flow rate include manufacturing variation, temperature effects, plugging, emitter wear with time, elevation changes, and microsprinkler stake assembly losses. 17.4.5.1 Manufacturer’s flow variation of emitters. The basic emitter flow, q, and water pressure, h, relationship (Equation 17.1) shows that if there is no pressure variation in the microirrigation system, all emitter flows should be constant and the emitter flow variation will be zero. However, in an actual field situation, there will always be emitter flow differences even under constant water pressure conditions. This variation is caused by small errors in the manufacturing process that result in flow differences from one emitter to the next. Any deviation in the flow passage area or shape from a standard size will cause emitter flow variation. The manufacturer’s variation is the variation in emitter flows from a random sample of emitters operated at the same pressure, and is expressed statistically as the coefficient of variation of emitter flow, CVM, which is the standard deviation of emitter flow, S, divided by the mean value of emitter flows, q . CVM =

S × 100 q

(17.8)

Test results show that the coefficient of variation for microirrigation emitters typically range from 3% to 20% (Solomon, 1979). Microsprinklers, microsprays, and minisprinklers usually have low manufacturer’s flow variations of less than 3%, although there can be large differences in the uniformity of water application patterns. Emitters with CVM values greater than 0.20 are not acceptable for microirrigation system design. 17.4.5.2 Temperature effects on emitter flow. The water temperature in a microirrigation line will be affected by the temperature of the air and soil surrounding the line. Exposed laterals and water will also be heated by solar radiation. Water temperature in a lateral line showed a 12°C to 17°C increase in bright sun (Gilad et al., 1968; Parchomchuk, 1976).

650

Chapter 17 Microirrigation Systems

The temperature effect on a microirrigation system can be considered in two parts: the effect on emitter flow and the effect on lateral-line hydraulics. The first one depends on the design and shape of the emitter and the second one depends on the friction situation in the line caused by the increase or decrease of viscosity of water due to temperature changes. The hydraulics of a lateral line determines the pressure distribution and water movement in the system which, in turn, affects the temperature variation in the microirrigation system. Emitter flows will be affected by the water temperature at the base of the emitters along the lateral lines. The effect of temperature on emitter flow depends on the type of emitter (Keller and Karmeli, 1974; Parchomchuk, 1976; Moser, 1979; Zur and Tal, 1981; Wu and Phene, 1984; Rodriguez-Sinobas et al., 1999). Temperature effects can be considered as insignificant when using both orifice and labyrinth-type turbulent-flow emitters (Wu and Phene, 1984; Rodriguez-Sinobas et al., 1999). Long-path laminar-flow emitters show an increasing emitter flow with increasing temperature. Vortex-type emitters show a decreasing emitter flow with increasing temperature (Rodriguez-Sinobas et al., 1999). The relationship between the emitter flow and temperature change can be expressed as a linear function. The temperature profile along a lateral line can be shown as a power function (Solomon, 1984) or a straight line when the temperature of the last point was considered to be caused by end effect and neglected (Wu and Phene, 1984). A theoretical evaluation of friction drop along a lateral line with a linear temperature gradient showed that the shape of the energy gradient line is not affected (Peng et al., 1986). 17.4.6 Emitter Variation The emitter exponent, x, also affects the relationship between emitter flow variation, qvar, and pressure variation, hvar, of a microirrigation system. This can be derived from Equation 17.1 (Wu et al., 1979) as: qvar = 1 − (1 − hvar ) x

(17.9)

where qvar can be simply expressed as the range of variation (Wu and Gitlin, 1974): qvar =

qmax − qmin qmax

(17.10)

where qmax is the maximum emitter flow and qmin is the minimum emitter flow. The pressure variation, hvar, is derived in the same fashion as emitter flow variation, but should be within ±10% of the average emitter pressure. Equation 17.9 shows that emitter flow variation is zero when x = 0, regardless of the pressure variation in the system. When x = 1, the emitter flow variation will have the same variation as pressure variation. This indicates that when the pressure variation, hvar, is 20%, the emitter flow variation for laminar flow emitters will also be 20%. But, for turbulent flow emitters, x equals 0.5, and a pressure variation of 20% in a microirrigation system will produce only about 10% emitter flow variation. 17.4.6.1 Combined variation. In the field, emitter flow rate variations are due to the combination of hydraulic variation and manufacturer’s variation. The relationship between them was first determined statistically (Bralts et al., 1981; Bralts et al., 1987a) and then verified by computer simulation (Wu et al., 1985). The total emitter flow variation caused by both hydraulic and manufacturer’s variation can be expressed by:

Design and Operation of Farm Irrigation Systems

651

(CV ) 2HM = (CV ) 2H + (CV ) 2M

(17.11)

where (CV)HM is the coefficient of variation of emitter flows caused by the combined effects of both hydraulic and manufacturer’s variation; (CV)H and (CV)M are the coefficient of variation of emitter flows caused by hydraulic design and manufacturer’s variation, respectively. 17.4.6.2 Grouping effects. When a number of consecutive emitters are grouped and considered as a unit (e.g., several emitters to irrigate a single tree), the uniformity of water application per tree will be improved (Wu et al., 1988b). The improvement depends on the magnitude of emitter flow variation caused by the hydraulic and manufacturer’s variations. For the case in which the emitter flow variation is caused by hydraulics only and the manufacturer’s variation is zero, there will be no grouping effect; that is, [(CV ) H ] g = (CV ) H

(17.12)

where [(CV)H]g is the coefficient of variation of emitter flow by hydraulics after grouping and (CV)H is the coefficient of variation of emitter flow from only hydraulics. For the case in which emitter flow variation is caused by manufacturer’s variation only and hydraulic variation is zero, the emitter flows will follow a normal distribution and the grouping effect will be shown by the relation

[(CV ) M ] g =

(CV ) M N

(17.13)

where [(CV)M]g = coefficient of variation of emitter flow by manufacturer’s variation after grouping (CV)M = coefficient of variation of emitter flow by manufacturer’s variation only N = number of emitters grouped together. When the emitter flow is affected by both hydraulic and manufacturer’s variations, the grouping effect can be expressed by the regression equation (Wu et al., 1989), [(CV ) HM ] g =

A N

+ 1.2487 B − 5.3935 B 2 + 7.6749 B 3 + 2.3113 AB

(17.14)

(R2 = 0.99) where [(CV)HM]g = coefficient of variation of emitter flow caused by both hydraulic and manufacturer’s variations after grouping A = (CV)M B = (CV)HM – (CV)M N = number of emitters grouped together. Equation 17.14 can be used for up to 17 emitters per group. 17.4.7 Effects of Plugging on Design A major problem encountered in drip irrigation is the plugging or clogging of emitters. Emitter plugging can adversely affect the rate of water application and the uniformity of water distribution. The combined effect of hydraulics, manufacturer’s variation, and plugging was evaluated statistically (Bralts et al., 1981) and verified through computer simulation (Wu et al., 1988a). The coefficient of variation of emitter flow caused by hydraulics, manufacturer’s variation, and plugging can be expressed as:

652

Chapter 17 Microirrigation Systems

(CV ) 2HM P (17.15) + 1− P 1− P where (CV)HMP is the total emitter flow variation affected by all three factors; hydraulics, manufacturer’s variation, and plugging (complete plugging only) and P is the fraction of emitters completely plugged. For the case where (CV)HM is zero, the coefficient of variation caused by plugging alone can be expressed simply as a function of P: (CV ) 2HMP =

(CV ) P =

P 1− P

(17.16)

where (CV)P is the coefficient of variation of emitter flow caused by plugging alone. Equation 17.16 shows that plugging can affect the uniformity tremendously. For example, 10% plugging will produce (CV)P of 33% for emitter flow while the ranges of (CV)H and (CV)M are 0.03 to 0.07 and 0.03 to 0.20, respectively, for the same impact. Similar to the situation of grouping emitters, an evaluation of contiguous plugging (Wu et al., 1991) showed that effect of four or more emitters plugged together is less than 1% for 10% and 20% plugging, respectively. Therefore, if four or more emitters irrigate a tree, the chances that a tree would receive no water because of plugging are greatly reduced.

17.5 MICROIRRIGATION DESIGN 17.5.1 Performance-Based Criteria Design calculations can proceed once the emission device, the required average emitter flow rate, emitter spacing, emitter variation, allowable pressure losses, and other criteria are determined. Microirrigation systems are designed based on the uniformity of water application with respect to crop needs. Two primary uniformity considerations used for performance based designs are the system emission uniformity and the spatial uniformity of the irrigation water in the crop root zone. However, the design procedures are basically the same regardless of the selected uniformity criteria. Emission uniformity (EU) describes how uniformly the overall system can distribute water from each emission device in the field and should be designed for at least 80% (90% with chemigation). The design criteria affecting the system emission uniformity include hydraulic design, manufacturer’s variation, temperature, plugging, and the number of emitters per plant. System emission uniformity is usually the most appropriate for design for microirrigation systems designed for widely spaced trees. Spatial uniformity is a measure of the distribution of the irrigation water in the crop root zone across the field. The primary design criteria affecting spatial uniformity include system uniformity, pattern of soil wetting, and emitter spacing. Spatial uniformity is more meaningful than system uniformity for irrigation of high-density plantings and is often used for designs where emitter wetting patterns overlap. Other uniformity metrics that may also be useful as design criterion include Christiansen’s uniformity coefficient, CU, and the coefficient of variation, CV. These uniformity measures are related as expressed by the following regression equations (Wu and Irudayaraj, 1987): CU = 1.0865 CV (r2 = 0.999) (17.17)

CV = – 0.0095 + 0.4288 qvar CU = 1.0085 – 0.3702 qvar

(r2 = 0.97) 2

(r = 0.97)

(17.18) (17.19)

Design and Operation of Farm Irrigation Systems

653

The high correlation between any pair of the uniformity measures expressed in Equations 17.17, 17.18, and 17.19 indicates all three uniformity measures can be used as criteria for hydraulic design. This justifies using the simple uniformity value, qvar, which is determined by only the maximum and minimum emitter flows for a lateral line or submain unit. 17.5.2 Emission Uniformity ASAE Engineering Practice EP405.1 (ASAE, 2005b) defines the emission uniformity (EU), also often referred to as distribution uniformity (DU) (Burt and Styles, 1994), of a microirrigation system for design purposes as: ⎡ ⎛ CVM EU = 100 ⎢1 − ⎜⎜1.27 × n ⎣⎢ ⎝

⎞ ⎛ Qm ⎟⎟ × ⎜⎜ ⎠ ⎝ Qa

⎞⎤ ⎟⎟⎥ ⎠⎦⎥

(17.20)

where CVM is the manufacturer’s coefficient of variation for point- or line-source emitters expressed as a percent. However, it is appropriate to use a combined representative CV including values presented above as well as other factors such as uneven drainage (during shutoff) and unequal spacing for initial design calculations because the actual EU will usually be less than the design EU. Qm is the minimum flow rate at the minimum pressure in the system, Qa is the average (or design) emitter flow rate at the average (or design) pressure, and n is the number of emitters per plant or the rooting diameter of plants divided by a given length of lateral line (often equals 1). The first factor represents the flow rate variation due to manufacturing (or combined) variation, and the second factor, Qm/Qa, expresses the variation resulting from system pressure changes. It should be noted that for evaluation purposes, EU = 100 (qLQ /qa) where qLQ is the average measured discharge in the lowest quarter of the measured field values while qa is the average of all measured values. The correlation coefficient (CV) is also related to the EU by the following relationship: CV = 0.77 × (1 – EU) (17.21) The general recommendation is that the selected combined flow variation and the flow variation ratio (Equation 17.20) should always result in a design EU above 80%. However, the actual selection of CV, CU, or DU depends on a number of factors including the cost of the system, the cost of water and related costs; the sensitivity of the crop (yield and quality) to stresses caused by nonuniform irrigation; the market value of the crop; and environmental concerns (e.g., leaching of agrichemicals to groundwater). Table 17.1 presents suggested ranges of uniformity values to use in design based on these factors. Table 17.1. Suggested range of design criteria for different uniformity expressions based on various economic, water supply, and environmental conditions. Design Considerations CV CU EU Abundant water and no environmental 20%–30% 75%–85% 60%–75% pollution problems Abundant water but environmental protection 10%–30% 80%–90% 75%–85% considerations are important Limited water resources but no environmental 15%–25% 80%–90% 70%–80% concerns Limited water resources combined with the 5%–15% 85%–95% 80%–95% need for environmental protection

654

Chapter 17 Microirrigation Systems

17.5.3 Hydraulic Design of Laterals The hydraulic design of a microirrigation subunit is based on the energy relations in the drip tubes, including the friction losses and energy changes due to slopes in the field (Bralts and Segerlind, 1985). Direct calculations of water pressures along a lateral line or in a subunit can be made by using an energy-gradient line approach (Feng and Wu, 1990; Wu and Gitlin, 1974). All emitter flows along a lateral line and in a manifold can be determined based on the corresponding water pressures along the lateral and manifold pipeline. When all the emitter flows are determined, the emitter flow variation, qvar, can be expressed by Equation 17.10. In general, the emitter flow variation, qvar, is used for hydraulic design. The design criterion for emitter flow variation, qvar, for drip irrigation design is recommended to be 10% to 20%, which is equivalent to a coefficient of variation from 0.033 to 0.076, respectively (ASAE Standard EP405.1, ASAE, 2005b). However, if justified, values from Table 17.1 may be used for design. Microsprinklers are often designed with orifice sizes over 1 mm in diameter to reduce plugging. These emitters can usually achieve system uniformity coefficients in the field above 90%. However, the individual water distribution patterns of these devices can be quite variable, which would not be evident from a uniformity coefficient based on flow rates. 17.5.4 Hydraulic Design for Subunits A microirrigation subunit is a fraction of the microirrigation system than is usually operated separately from other subunits but may be operated simultaneously with other subunits. For reasons of economy and water availability, microirrigation systems are often designed in four or more subunits. Thus, the irrigation pump, power supply, filtration system, and other water supply components can be smaller than if the entire production system was irrigated as one unit. However, it is sometimes desirable or necessary to operate the entire microirrigation system as one unit, such as when microsprinklers are used for frost/freeze protection (Evans et al., 1988; Evans, 1994), but this increases capital costs because many of the various system components must be considerably larger. A subunit consists of an irrigation manifold (or header) pipeline with laterals that are supplied water from the manifold. A valve (usually a solenoid valve) is used at the entrance to the manifold to control water applications to the subunit. A pressure gauge, flow meter, pressure regulator, and chemical injection port may also be located at the manifold entrance as needed. If there are several smaller blocks in a subunit, the design should prevent drainage from blocks at higher elevation causing excess applications in lower blocks by elevation control or the use of spring-loaded check valves. This also provides for more rapid pipe filling and better system uniformity because the piping system does not have to be recharged for each irrigation. Water is provided to the subunit by main or submain pipelines that are hydraulically much different from subunit pipelines. Subunits consist of multiple outlet pipelines with uniformly spaced outlets removing water along their lengths. Conversely, mainlines and submains have uniform flow along their lengths, leading to greater friction losses for the same pipe diameters and inflow rates. Subunit pipelines are designed to meet two criteria: high uniformity and low cost. However, these criteria often oppose each other because high uniformity is achieved

Design and Operation of Farm Irrigation Systems

655

by uniform pressures and minimal pressure losses that are accomplished by increasing pipe sizes, additional control valves, and other measures causing higher costs. For this reason, standards have been developed to define an acceptable degree of uniformity of water application (e.g., ASAE Standard EP405.1, ASAE, 2005b). These standards were written with the realization that it is expensive to achieve very high uniformities. Emission uniformities in the range of 70% to 95% are generally acceptable (Wu and Irudayaraj, 1992). Higher values may be appropriate on flat surfaces where it is less costly to achieve higher uniformities, while lower values are more acceptable on steeply sloping areas unless chemigation is used. Subunits must be designed considering head losses in both the manifold and lateral pipelines. Because some subunit head losses occur in both the manifold and the laterals, it is not appropriate to base the design on flow uniformity only in the laterals unless pressure is regulated at the entrance to each lateral. These differences may be especially significant when field slopes are large and the manifold is positioned upand-down slope. Subunits can be designed including both the laterals and manifolds by the following steps: 1. Select the design emission or spatial uniformity based on ASABE Standards or other sources. 2. Calculate the allowable flow rate variation within the subunit from the emission uniformity equation. 3. Calculate the allowable pressure variation within the subunit from the emitter hydraulic characteristics and the allowable flow rate variation. 4. Design the lateral using a fraction (e.g., 60% on first approximation) of the allowable pressure loss within the subunit (which is a proportion of the allowable whole system pressure losses). 5. Design the manifold using the remaining allowable head loss in the subunit that was not used in the lateral design. 6. Repeat steps 4 and 5 in a trial-and-error procedure that changes the split between the lateral and manifold pressures losses until a minimum-cost solution is obtained. Lateral calculations can be based on the energy gradient line (EGL) or revised energy gradient line (REGL) method (Wu, 1992a; Wu and Yue, 1993). Design charts were developed for lateral and submain designs for simple EGL (Wu and Gitlin, 1974) and REGL (Wu, 1992b) methods. The design criteria for the lateral line is calculated based on design emission uniformity criteria. Subunits have also been designed using a finite element approach (Bralts and Segerlind, 1985). Flow conditions in the laterals and manifold decreases steadily along their length, but can vary spatially depending on the layout (Howell and Hiler, 1974; Anyoji and Wu, 1987). A lateral line with hundreds of emitters makes a step-by-step (SBS) calculation for all sections between emitters very tedious. An energy-gradient line (EGL) approach was applied for determining pressure variation along the lateral (Wu and Gitlin, 1974). The concept of energy-gradient line offers a direct calculation of emitter flows along the lateral line because simple equations can be derived to determine all emitter flows along a lateral. There are several good commercial programs and spreadsheets available for computer-aided design of these systems. For example, computer programs have been developed for microirrigation design using a finite-element approach (Bralts and Seger-

656

Chapter 17 Microirrigation Systems

lind, 1985) and step-by-step calculations (Pitts et al., 1986; Meshkat and Warner, 1985). The energy-gradient line concept has also been applied in developing computer-aided design for microirrigation systems (Feng and Wu, 1990). 17.5.5 Main and Submain Pipeline Design Main and submain pipelines must deliver the necessary amount of water and energy (pressure) to the entrance of each subunit to meet emission uniformity criteria. They must be properly pressure-rated to withstand operating plus surge pressures. This generally requires that pipeline velocities be limited to values that limit surge pressures to acceptable levels (e.g., 1.5 m/s or less). Thus, main pipelines are designed based on economics using pipe and fittings that meet the required pressure ratings. A cost analysis of materials and energy use is required to determine the lowest-cost pipeline for the required flow rate and estimated hours of operation per analysis period. These analyses are usually made on an annual basis to amortize initial capital and installation costs and compare them with estimated annual operating costs.

17.6 DESIGNING THE SYSTEM CONTROL HEAD The microirrigation system control head is defined as all of the pumps, valving, filters, injectors, controllers, monitoring equipment, and other facilities required to deliver water at sufficient pressures and appropriate quantity and quality to the irrigation system. The irrigation system control head must be located with convenient access for maintenance and operation. Figure 17.1 presents a schematic of the various components and their placement in a typical microirrigation system control head. Site preparation should ensure drainage of excess storm water from the control head area as well as providing reliable access under adverse climatic conditions. The

Figure 17.1 Schematic representation of the all the components typically required in the system control head for a microirrigation system.

Design and Operation of Farm Irrigation Systems

657

control head should be installed on a level concrete pad of sufficient strength and size to mount all pumps, filters, flow meters, electrical control panels, valves, injectors, and other equipment. This provides a stable foundation to which equipment can be bolted to reduce vibration, avoid structural stresses, and facilitate maintenance. Thrust blocks may also be required at inlet and outlet pipes. Suitable supports must be provided beneath heavy components such as flow meters, control valves, and filters. Protect the installation from accidental mechanical damage by agricultural implements, vehicles, and tractors. The control head, ponds and settling basins, and chemical tanks should be fenced to keep children, unauthorized people, or animals from damaging the components or being exposed to dangerous chemicals. Drainage and/or spill-containment facilities should be provided around any chemical or fertilizer supply tanks to prevent direct contamination of any surface water from spills. Likewise, the wellhead should be protected to prevent contamination of the subsurface waters from bacterial and/or chemical sources. Filter backwash or other potentially contaminated water should be disposed of by land spreading (e.g., dust suppression) and not allowed to flow into drainageways, especially if it contains injected chemicals. A suggested checklist for considerations during structural design of the system control head is: 1. Design the height of the control head installation for convenient dismantling and assembly of the various components for cleaning, repair, and replacement, while minimizing the potential for debris or other contaminants to enter the system. Components should generally have a minimum height of about 0.4 m above the concrete surface to provide adequate working space. 2. Maintain appropriate distances between various components to ensure reliable function of meters and gauges and facilitate operation, maintenance, and cleaning of filters, and dismantling and replacement of defective parts. Make sure that the direction-of-flow arrows on components such as flow meters and check valves are in the direction of flow. 3. Ensure that the components can be isolated by valves for repair and maintenance work and that sufficient unions, Victaulic couplings, and/or flanges are installed to facilitate dismantling and repair of components. Locate all valves for easy access, opening, maintenance, and removal. Avoid directly joining dissimilar metals without a dielectric union to prevent electrolysis and corrosion of fittings. The exposed length of threaded steel pipe nipples after assembly should enable convenient access for a pipe wench. 4. Select resistant materials for all pipes and components that may come in contact with concentrated chemicals, including fertilizers. Special coatings or linings may sometimes be required to protect hydraulic components from direct chemical effects. 5. Pressure gauges or pressure-measurement taps should be provided immediately upstream and downstream of all major components that modify pressure (e.g., pressure regulators, filtration devices, fertilizer injectors, pressure sustaining valves, etc). 6. Electrical and hydraulic interlocks are required for injection equipment to prevent backflows from contaminating water supplies and to prevent chemical injection when the main water-supply pumps are not operating. Likewise, chemi-

658

Chapter 17 Microirrigation Systems

cals (except biocides such as chlorine) should not be injected during filter backwashing and flushing activities. Installation of backflow-prevention system interlocks, injection-line check valves, and other safety devices must be in accordance with local standards or regulations. 7. Solenoid valves should have manual-override or hydraulic-bypass capabilities. They should take 1 to 5 seconds to open and close to avoid waterhammer problems. 8. Pump-control valves that slowly bring the system online without water hammer and protect the pumps on shutdown are recommended, especially for turbine pump installations. Pumps should have low-pressure switches to prevent damage to the pump in case of water loss. 9. Protect workers from electrical hazards by installing and maintaining proper shielding, interlocks, and by providing adequate grounding of all electrical equipment. Ground-fault circuits should be provided for all injection pumps. 10. Pump motors or engines should have a cover to shield them from direct sunlight. This will increase the life of the components and reduce overheating. 11. Electric pump motors should have ammeters installed at the panel and the readings recorded as part of normal regular maintenance record-keeping programs.

17.7 INSTALLATION One of the most important considerations for installation of a microirrigation system is worker safety. Adequate room and suitable topography for equipment to operate safely without endangering the operator and other workers must be ensured. Trenches and other excavations deeper than 1 m must be sloped or special protective measures provided to protect workers from side-wall failures. The contractor should implement a quality control program during installation to ensure that all connections are made correctly and avoid entry of soil and debris into pipes and tubing. All mains and submains should be thoroughly flushed before hooking up tubing. Microirrigation systems should be thoroughly flushed immediately after installation is completed as well as after any new construction or repairs. The contractor should ensure that valves, pumps, and filters are properly installed and adjusted. The contractor should also test the system for proper pressure and flow distributions and ensure that there are no leaks. The emission uniformity of the new system should be evaluated to determine if the new system meets design specifications. Additional guidelines for the installation and post-installation evaluation of new microirrigation systems are presented in ASAE Engineering Practice EP405.1, Design and Installation of Microirrigation Systems (ASAE, 2005b). The irrigator should become familiar with the controls and characteristics of the new system. A suitable water management program should be implemented and new cultural and harvesting practices adopted as necessary.

17.8 MAINTENANCE Implementation of a diligent and rigorous maintenance program is central to the long-term success of microirrigation. A good maintenance program involves implementing good record keeping, an appropriate chemical water-treatment program and regular flushing to keep pipelines clean. Many maintenance problems can be circumvented by consistent records of flow meters and pressure gauges. This process can be greatly facilitated by remote commu-

Design and Operation of Farm Irrigation Systems

659

nication technologies tied to local computers. Frequent visual inspections are also required to ensure that all system components are functioning properly. Regular field inspections will help find emitters that are plugged, identify improperly working flush valves, and locate pipes and tubing damaged by coyotes, small rodents, insects, and other causes, including farm equipment. 17.8.1 Field Evaluation of Microirrigation Operation and Uniformity Water distribution uniformity measurements should be made on newly installed microirrigation systems to confirm that the system has been properly designed and installed as well as to provide a basis for later comparisons. Measurements of the uniformity of water distribution should be made before each crop season and compared to the new system evaluations. If this is too labor intensive or is impractical (as may be the case for subsurface systems), then, as a minimum, the irrigator should compare actual system flows and pressures at the inlet and distal ends of the system with the initial evaluations. Additional tests may be required for evaluation and adjustment of maintenance and operational procedures during the growing season, particularly where emitter plugging problems are severe. Decreases in distribution uniformity over time are a cause for concern. Although regular visual inspections will locate emitters that are completely or almost completely plugged, they will not identify small changes in emitter flow rates from partial plugging. Frequent examinations of flow-meter records and periodic field measurements of emitter flow rate and pressure variations will help ascertain changes in system performance. Early identification of problems should indicate the need for special chemical water treatments to clean partially plugged emitters before the problem becomes more serious. Subsequent comparisons, where partial emitter plugging may be present due to chemical precipitation, algae, or other causes, may be made using the Christiansen uniformity coefficient (CU) or other statistical measures of uniformity discussed in Chapter 5 and in Pitts et al. (1996). Emitter damage and wear will also affect flow rates as the emitters age. If the uniformity is low, additional samples should be taken to improve the statistical confidence. If the additional samples of emitter flow rate indicate that the distribution uniformity is still low, pressure distribution tests should be conducted to assist in identifying the cause(s). Crimped or leaking pipelines and laterals, improperly adjusted pressure regulators, improperly sized pipelines or fittings, and valves that fail to operate properly are all factors that can result in high hydraulic pressure variations. Conversely, if the hydraulic variation is low, then the poor water distribution uniformity is likely a problem with plugging or incorrect emitter selection. The water application by the drip irrigation system as well as the water infiltration in the field can usually be considered as a normal distribution as long as the coefficient of variation of emitter flow or spatial uniformity is less than 30% (Wu, 1988). The cumulative frequency distribution of a normal distribution can be approximated by a straight line. The linear distribution of irrigation application will produce both underand overirrigated areas, which can be quantitatively determined by simple mathematical equations (Karmeli, et al., 1978; Seginer, 1978; Sammis and Wu, 1985; Wu, 1988). ASAE Engineering Practice EP-458, Field Evaluation of Microirrigation Systems (ASAE, 2005d), defines general procedures for field emitter evaluations. EP-458 assumes a normal distribution of emitter flow rates measured in the field. However, even

660

Chapter 17 Microirrigation Systems

with partial plugging of emitters, the uniformity will usually be overestimated because plugging and pressures are not normally distributed throughout the system. In general, criteria for emitter flow variation are: 5% or less, excellent; 5% to 10%, very good; 10% to 15%, fair; 15% to 20%, poor; and greater than 20%, unacceptable. Hydraulic coefficient of variation criteria are: 10% or less, excellent; 10% to 20%, very good; 20% to 30%, fair; 30% to 40%, poor; and greater than 40%, unacceptable. Bralts et al. (1987a; 1987b) discussed statistical considerations in the determination of distribution uniformities. Camp et al. (1997) evaluated several different distribution uniformity evaluation techniques and discussed the limitations and advantages of each. Because many of the economic and environmental impacts are functions of climate, topography, and crop production systems, guidelines for acceptable uniformity often exist for specific locations. Smajstrla et al. (1997) presented specific steps for field evaluations in humid regions (also assuming statistically normal emitter-flow distributions). They presented tabular and graphical procedures that simplify data analysis for both flow rate and pressure variation to identify nonuniformity problems, to determine the required number of emitters to test, and to determine the cause of any nonuniformity observed. The graphical procedures require that a minimum of 18 emitters be randomly sampled in each subunit evaluated. 17.8.2 System Maintenance Manufacturer’s recommendations for maintenance should be followed for all components. Each component should be routinely inspected and tested to ensure that it functions properly. Consistency in all aspects of the maintenance program is the key to successful microirrigation. Keeping detailed records of irrigation schedules, chlorination, chemical treatments, chemigation, and maintenance activities is critical to document maintenance problems, properly schedule required maintenance, conduct financial analysis, and plan for future improvements. Inspection of buried pipelines and equipment is difficult. Therefore, indirect monitoring and evaluation by routinely charting flow-meter readings and pressures can be used to check the performance of subunits or the entire microirrigation system. Changes in system performance will indicate maintenance needs, even when such changes occur slowly. Monitoring pressures and flows identifies leaks or emitterplugging problems and documents how fast the problems are progressing. For example, gradually decreasing flow rate and increasing pressure may indicate gradual emitter plugging, while rapidly increasing flow rate and decreasing pressure can indicate leaks or broken pipelines. Instrumentation such as tensiometers or other soil-water sensors can also help call attention to field distribution problems. Periodic calibration or replacement of flow meters and pressure gauges will also be required. Mainlines, submains, and laterals should be flushed to remove sediments at least once each month or as needed during the season, depending on water conditions. Mainlines should be flushed first, then submains, manifolds, and finally laterals. Systems can be manually or automatically flushed. The whole system should be flushed at seasonal startup, at the end of the season, and whenever repairs are made. Flushed materials should be inspected for signs of chemical precipitations, algal buildups, or root intrusion. Monitoring the frequency of primary filter backwashing and pressure drops can direct attention to developing filtration problems. Filtration media should be replaced as needed. All filters should be manually inspected and cleaned on a regular basis. Ponds,

Design and Operation of Farm Irrigation Systems

661

canals, and settling basins require periodic mechanical or chemical cleaning to control sediment accumulations, aquatic weeds, and algae. Weed control with microirrigation can be a challenge because both wet and dry soil conditions exist over short distances. Widely different weed species, requiring different herbicides for control, will inhabit small areas, but most weeds will be in the fringe areas between the wet and dry soil zones. Fortunately, herbicides labeled for direct application through emitters tend to work well. However, high soilwater conditions can cause rapid leaching or degradation of many herbicides. Weeds are often successfully controlled using plastic mulches or multiple spray applications of glyphosate or other herbicide depending on the crop and location. It is also important to keep the control area free from weeds, brush, vines, or other materials that might block access or hinder maintenance activities. As mentioned earlier, coyotes, rodents, and other animals may damage tubing. This generally occurs when they are looking for water. Daily monitoring of system flow rates and visual inspections will help reduce resulting water distribution problems. If these are chronic problems, bitter oils can be periodically injected and strategic placement of water dishes around the field may be beneficial. Other, less regular, maintenance activities include flushing injection equipment with clean water after each use for safety and to avoid corrosion. Pumps, filters, valves, gauges, injectors, tanks, pipelines, and other hydraulic components must be protected from freezing in winter by removal or draining in cold climates. Insects may also enter air vents and cause them to leak. Electrical panels need to be kept free of moisture and dust.

17.9 MANAGEMENT In general, microirrigation requires higher levels of management than other irrigation methods because decisions must be made daily or more frequently. Specific management decisions will depend on crop, site, soil, and environmental conditions. Schwankl et al. (1995) discussed water management of microirrigated tree and vine crops while Hanson et al. (1994) presented a similar discussion for row crops. The questions concerning microirrigation management generally center on when to irrigate, how much to apply, how to accurately evaluate the water status of the plant, and integration of other cultural activities with irrigation needs. These decisions are facilitated by adoption of a sound irrigation scheduling program, which may be supported with automation and monitoring instrumentation. Chemical treatment of water, filter cleaning, routine flushing of pipelines and laterals, and a good overall maintenance program are also fundamental to good management. 17.9.1 Management in Arid Areas One of the most important microirrigation management considerations for arid areas is that active rooting volumes are small because water is often applied to 30% or less of the total potential rooting area. This can physically limit water and nutrient uptake causing stress during high ET-demand periods. Thus, management must focus on optimizing the use of a limited wetted soil volume for both water and nutrients. Increased sizes of wetted areas by microsprinklers instead of drippers may be required on sandy soils to improve soil water and nutrient availability. Reduced wetted volumes compared to other irrigation methods can affect management decisions regarding soil salinity and leaching, applied water quality (e.g., salts), plant nutrition, soil pH, and micronutrient availability. Because of the restricted root

662

Chapter 17 Microirrigation Systems

zone, however, fertigation programs work well on a high-frequency basis where nutrients can be applied as needed in small amounts with rapid uptake and minimal leaching, although foliar applications of micronutrients may be required in some cases. Water table contributions to plant ET may be a significant factor in irrigation water requirements for microirrigated crops depending on seasonal variations in aquifer depth. The benefits of maintaining slightly reduced root zone soil water levels for storage of precipitation are limited because soil water levels are already substantially reduced outside the wetted root zone areas, thus providing abundant storage for any precipitation. 17.9.2 Management in Humid Areas The crop-rooting volume is not limited to the irrigated zone in humid areas. Water supplies are plentiful and frequent rains allow root development and associated water and nutrient uptake to occur outside of the irrigated zone. Thus, it is important that the irrigated rooting volume be large enough to minimize stress because roots are not concentrated near emitters. It is generally recommended that a microirrigation system be designed to irrigate at least 50% of the crop root area in these situations. When drippers are used, low-flow emitters are relatively closely spaced with close lateral spacing to ensure optimum crop yields. There are tremendous advantages to microirrigation in humid areas. The cost of water applications is normally low because of the small amounts of water applied over a season and the low pressure requirements of these systems, but crop yield and quality increases can be substantial by avoidance of short-term drought effects. Irrigations can also be applied without wetting the plant foliage and maximize the time that the foliage remains dry between rainfall events which greatly reduces the incidence of foliar plant diseases that require the use of fungicides or other agrichemicals. Fertigation is highly effective in humid areas; however, leaching of nutrients is often a significant problem due to both heavy rainfall and overirrigation. Thus, microirrigation systems must be properly managed to avoid leaching and the associated problems of contamination of groundwater or surface water systems. This requires that both water and chemicals be applied in small doses so that the leachable quantity of chemicals is limited in anticipation of large rainstorms. As contrasted to microirrigation management in arid areas, excess water applications are rarely needed for salinity control in humid regions. Exceptions occur where very poor-quality irrigation water is used, when very salt-sensitive crops are grown, or during extended drought periods. The relatively frequent occurrence of large rainstorms normally provides adequate leaching and soil salinity management in humid areas. Boman and Parsons (1998) discuss the selection and design of microsprinkler systems for tree crops in humid regions. As in arid areas, water table contributions to crop water requirements can be significant. The amount will be site specific and must be considered in scheduling irrigation applications. These shallow aquifers are also easily contaminated, and irrigation schedules must be developed that avoid leaching. Both surface and subsurface drainage systems are often required in humid areas, especially on heavy soils or light soils in flat areas with restrictive layers that perch water tables near the soil surface. These drainage needs are reduced but not eliminated by microirrigation because of numerous high rainfall events.

Design and Operation of Farm Irrigation Systems

663

There is little benefit from delaying or reducing irrigation applications in anticipation of rainfall, thus full irrigations are normally applied for most crops to optimize yield and quality. The primary exception to this practice is microspray or microsprinkler irrigation of tree crops where the irrigated volume of soil is normally large enough to significantly increase effective rainfall by delaying or reducing the amount of irrigation. 17.9.3 Controlled Root Volumes Efficient root concentration within a limited wetted soil volume is readily achievable with microirrigation in arid areas. Benefits of maintaining concentrated root systems under an emitter may be: (1) improved water availability due to the reduced importance of soil hydraulic conductivity; (2) efficient application of water by minimizing losses due to evaporation and deep percolation; (3) efficient application of fertilizers and other water-soluble chemicals, particularly those which tend to be fixed by the soil particles (e.g., potassium and phosphorus); and (4) inducing physiological root restriction or drought effects on perennial plants to cause reduced vegetative-toreproductive growth ratios and better light penetration into the canopy. There are basically three variations of controlled root zone strategies, including regulated-deficit irrigation, controlled-deficit irrigation, and partial root zone drying. These are discussed below. 17.9.3.1 Regulated-deficit irrigation. One controlled root-volume technique is regulated-deficit irrigation (RDI), which is limited to relatively arid areas. This technique deliberately imposes specific plant water stresses during specific growth stages (usually early in the season) using daily irrigations but only replacing 10% to 30% of the plant’s daily water use. The wetted soil volume contracts from the sides and bottom of the root zone. At the end of the stress period (as indicated by various physiological markers), water application amounts are increased (e.g., up to 85% to 100% daily actual evapotranspiration), but soil water profiles are not refilled and the size of the small wetted soil volume remains constant. Vegetative growth must not be reinitiated by excess soil water conditions. RDI requires that adequate allocations of late-season water be available to “finish the crop” and that the system be designed to apply at least peak crop water use on a daily basis throughout the entire growing season. Automated microirrigation is highly desirable. To date, RDI has only been investigated on perennial crops. Research in Australia on peaches (Chalmers et al., 1981) and pears (Mitchell et al., 1984), Washington on apples (Proebsting et al., 1977; Middleton et al., 1981; Peretz et al., 1984; Evans et al., 1993, Ebel et al., 1995; Drake and Evans, 1997) and grapes (Evans et al., 1990; Wample, 1996, 1997) have produced beneficial responses. Additional work in California, Israel, Australia, Chile, and other arid locations on several crops has also shown that carefully managing the severity and duration of a uniform, constant level of water stress on fruit trees, wine grapes, and some other perennial crops can be advantageous. RDI has been found to control vegetative growth, increase fruiting, advance fruit maturity, increase precocity, and increase soluble solids in fruits. Annual water diversions can be reduced by 20% or more. The key to successful RDI is rigid control of soil water volumes to control vegetative growth. It is made possible by the practical ability to achieve high-frequency irrigation regimes and the capacity to carefully restrict soil water by controlling the application amount and the size of the wetted volume of soil available to the roots.

664

Chapter 17 Microirrigation Systems

17.9.3.2 Controlled-deficit irrigation. Controlled-deficit irrigation (CDI) generally refers to irrigation strategies that apply less than actual water use late in the growing season. For example, CDI is often used as a water conservation technique with perennial crops in arid areas, such as peaches, plums, or cherries, which are harvested in early to mid-summer Similarly, carefully managed CDI may also be used to induce a plant physiological response, such as inducing winter hardiness of perennial crops that are not physiologically adapted for growth in arid areas with cold winters. Irrigations are maintained at a reduced level after harvest for the remainder of the season; however, drought stress is not allowed to reach severe levels that would affect next year’s crop. The saved water is then used for other crops. 17.9.3.3 Partial root zone drying. Partial root zone drying is a simple drip irrigation technique on perennial tree and vine crops that utilizes drip lines located near the middle of the alley between plant rows (Leib et al., 2006). The systems are used in a manner similar to RDI in terms of timing; however, only one drip is irrigated at a time allowing the soil volume covered by the line on the other side of the plant to dry out. The next irrigation will apply water through the second drip line allowing the other side to dry to low soil water levels. Physiological responses are similar to those reported for RDI. This practice is commonly used on European (Vitis vinifera) wine grapes in south-central Washington, west-central Idaho, western Colorado, and northcentral Oregon.

17.10 SCHEDULING MICROIRRIGATION The basic philosophy of microirrigation is to replace water in the root zone in small increments as it is used by a plant at intervals ranging from several times a day to once every two to three days, rather than refilling a much larger soil water reservoir after several days or weeks. Thus, the major concern for scheduling microirrigation systems is how much to apply during an irrigation because the irrigation interval is often fixed by other factors. The estimated crop water use (ET), combined with the percent of the area that is irrigated, will determine the total volume of water to be applied (Clark, 1992). The maximum interval between irrigations is primarily controlled by soil hydraulic characteristics, soil profile layering, and tubing placement. The depth of soil, saturated hydraulic conductivities, and soil water-holding capacities will control the volume applied in a single irrigation to avoid runoff or excessive deep percolation. It is sometimes not possible to achieve optimum irrigation schedules because of irrigation system limitations. These may include inflexibility in controls and instrumentation, inadequate system hydraulic capacities (including fill times and system drainage), and the quantity and quality of available water throughout the season. Management considerations such as the quality and quantity of available labor can affect the ability to implement scheduled irrigations. Likewise, timing, amount, and label requirements for chemigation may influence irrigation timing and depth of application that can affect prior as well as subsequent irrigation schedules. Excess applications may have to be periodically scheduled to leach salts. Irrigation schedules may also have to be adjusted because of other cultural or harvesting considerations. Once the above factors are considered, irrigations can be scheduled whenever an estimated allowable depletion level has occurred, or to replace estimated or measured ET each day. Alternatively, irrigations can be automatically initiated and stopped whenever soil water matric potentials at selected points in the wetted soil volume

Design and Operation of Farm Irrigation Systems

665

reach predetermined levels as measured by soil water sensors. An economically optimal microirrigation schedule (Wu, 1995) can also be developed based on irrigation system uniformity and costs associated with supplying water, value of the yield, and the costs related to groundwater contamination due to seepage. 17.10.1 ET Estimation Irrigation scheduling must be a dynamic process because ET will vary spatially and temporally during the growing season. ET depends on the plant, soil, local environmental conditions, the percentage of the root zone irrigated, planting density, rooting characteristics, and canopy size. Pest and disease problems also reduce crop water use, and are usually variably distributed across a field, creating opportunities for localized excess applications and leaching. Likewise, nutrient availability and the plant’s uptake can strongly influence canopy growth affecting the total and the spatial distribution of water use. The calculation of ET is covered in Chapter 8. Rainfall will lessen the irrigation water requirements by the effective rainfall amount (Kopec et al., 1984). The contributions to ET from shallow water tables will likewise reduce irrigation requirements. The most reliable ET estimates are based on field irrigation experiments conducted with a wide range of irrigation treatments (e.g., Doorenbos and Kassam, 1979), lysimetry, or determined by calibrated ET models using weather variables (Jensen et al., 1990). ET under nonstressed conditions, such as those commonly found under microirrigation, may be higher than previously determined values which were developed under more traditional forms of irrigation. Conversely, total ET from widely spaced tree and vine crops may be reduced because soil evaporation and transpiration from grass under the crop canopy are less. Any deviations in estimating ET different from actual can result in reduced yields, wasted water, undesirable plant physiological responses, or a combination of these and other factors. Consequently, these uncertainties require that soil water or plant water potentials be monitored under all microirrigation methods for proper irrigation scheduling. The irrigated area, in general, is taken as the total surface area for row crops and other high-density plantings, considering that eventually most of the area is shaded when the crop matures. However, for low-density or very young plantings with small root zones, water applications and schedules should use projected canopy area or other measures of the affected cropped area. 17.10.2 Irrigation Frequency More than one water application per day may be required because either the total actual daily evapotranspiration cannot be stored in the limited wetted root zone volume (e.g., small vegetable crops) or single large water applications at a point may cause excessive deep percolation losses and leaching. Conversely, on heavier soils with high water-holding capacities or poor drainage, optimal irrigations might be only every second or third day. Some crops (e.g., cantaloupe, cotton, and many perennial crops) may perform better with less frequent irrigations (e.g., every 2 to 4 days) especially on heavy soils, whereas more water-sensitive crops (e.g., tomatoes, watermelons, lettuce) may need at least daily irrigations for best yields and quality. If leaching of fertilizers or other chemicals is a major concern, then sensorcontrolled, ultra-high frequency (e.g., 8 to 10 times/day) pulsed irrigation systems may be an option. Shallow-rooted crops often benefit from light, high-frequency irriga-

666

Chapter 17 Microirrigation Systems

tions. High-frequency microirrigation techniques have been shown to increase lateral wetting and reduce water and nutrient stresses, especially when used with fertigation, but the small wetted soil volume will reduce the plant’s ability to endure a drought stress of even short duration. However, growers need to be aware of management and soil water problems caused by frequent filling and localized draining of lines under high-frequency strategies. On the other hand, less-frequent irrigations may be beneficial to manage so that humidity levels in dense crop canopies are lowered to reduce incidence of fungal diseases (e.g., Botrytis on grapes). Irrigation deficits may be intentionally imposed at times to attain certain desirable crop quality or other plant physiological responses. 17.10.3 Monitoring Soil Water Soil water sensors make point or small-volume measurements in a field to monitor soil status and to control irrigations. All soil water monitoring devices should be placed at appropriate depths and locations to ensure that irrigation scheduling will be appropriate to optimize yields, minimize water usage, and minimize leaching to the groundwater. However, microirrigated soil water distributions are highly variable and there are major questions on determining suitable sensor locations and the correct interpretation of the readings. Preferential flow of soil water is often a major, but largely unquantifiable, factor in soil water distributions. Consequently, microirrigation scheduling is often “calibrated” to particular sensor placements with respect to a wateremission point that is correlated with plant water potential measurements or other independent variables. Calibration is typically required to optimize both water and nutrient utilization (Smajstrla and Locascio, 1996). The number of required sensors can be minimized by choosing representative plants and soil types across a field. Optimal sensor location will also be influenced by irrigation interval because a soil water gradient will develop from the emitter to the perimeter of the wetted volume during irrigation. This gradient decreases after irrigation due to water redistribution and the wetted soil volume approaches relatively uniform water content. Thus, sensors to control daily or more frequent irrigations are generally located within 10 to 15 cm of the emitters but may be located further away for less-frequent irrigations. Electronic soil matric potential sensors are often appropriate for these applications. Sensors that determine when to irrigate are normally placed in the upper one-fourth to one-half of the root zone within the most active areas of water and nutrient uptake. Sensors located in the lower portion of the root zone can be used to control the amount of water applied and avoid excessive applications. 17.10.4 Scheduling Criteria for Design Microirrigation systems are commonly designed for a 90% uniformity coefficient, especially when agrichemicals are to be injected through the system. Even with a 10% emitter plugging the uniformity will still be greater than 70%. Consequently, a field “scheduling” efficiency of 80%, which includes effects of emitter and hydraulic nonuniformities, is often used. The efficiency relationships between gross and net applications are discussed in Chapters 5 and 21 and will not be expanded here.

17.11 PLUGGING OF MICROIRRIGATION SYSTEMS Partial or total plugging of emitters is a chronic problem and the most serious constraint to the long-term operation of any microirrigation system. Inadequate consideration of the physical, biological, and chemical characteristics of the water supply will

Design and Operation of Farm Irrigation Systems

667

result in serious plugging problems. The most critical design factors affecting plugging are emitter design, filtration, and the chemical water treatment system. System operation and maintenance, including inadequate flushing of pipelines, will have major effects on microirrigation plugging problems. Improper installation, such as placing tapes with emission points facing down, may also contribute to plugging problems due to sediment accumulations along the bottom of the tape. 17.11.1 Causes of Plugging Plugging of microirrigation systems may occur from single or multiple factors. Physical factors such as suspended colloidal clays, silts and other materials passing through filters, broken pipes, root intrusion, and aspiration of soil particles into the emitter orifices are common physical causes of plugging. Chemical factors such as precipitation of carbonates and iron oxides, and precipitates from chemical injections are also significant causes of emitter plugging. Likewise, organic and biological factors such as oils, algae, aquatic weeds, bacterial slimes, fungi, as well as spiders, insects, worms, fish, frogs, snails, clams, and their eggs or larva can be major contributors. Low system pressures and flow rates will exacerbate plugging problems. 17.11.2 Sediment Routine flushing of pipelines is required to prevent emitter plugging from the gradual accumulation of particles which are too small to be filtered, but which settle out or flocculate at the distal ends of pipelines. Flushing velocities should be about 0.6 m s-1 to ensure transport and discharge particulate matter from the pipelines. This requires flow rates at the ends of the lines of about 0.12 L s-1 for 15-mm tubing and 0.22 L s-1 for 22-mm tubing. Flushing frequency should be at least once a month, but will vary through the season depending on the rate debris and particulates accumulate. Applying surfactants or dispersing agents such as sodium hexametaphosphate through the microirrigation system may reduce some plugging problems by preventing the flocculation of silts and colloidal clays, allowing them to easily pass through the emitters or be flushed from pipelines. Automated flush valves are sometimes used at the ends of the laterals to help flush fine particulates at the start of each irrigation; however, periodic manual flushing is still required. Use of these valves is generally not recommended because they tend to leak and waste water, requiring extra maintenance in addition to the added purchase cost. Use of these valves with chemigation may also be problematic due to leaks and the potential for chemicals to accumulate in these locations. 17.11.3 Algae and Bacterial Slimes Chlorine injection is the most common and least expensive method to prevent clogging by biological growth (algae, colonial protozoa, sulfur bacteria, and other mucous organisms). Bacteria that precipitate iron, sulfur, and manganese can also be effectively and economically controlled by chlorine treatments. Copper sulfate, chlorine, and organosulfur compounds are used to control algae and/or bacterial slimes in drip systems as well as in ponds or canals. The degree of control will vary with light and water temperature conditions. Some chemicals such as quaternary ammonium are effective when algal growth is slow to moderate, but will fail under conditions of rapid growth. Organic growers may be limited to copper sulfate at 100 to 200 mg L-1 concentrations for controlling algae, depending on local regulations.

668

Chapter 17 Microirrigation Systems

Soil bacteria can sometimes be aspirated back into an emitter and produce slimes that glue small particles together and plug the orifices. These have been fairly successfully, but expensively, removed by strong solutions of hydrofluoric acid (1000 mg L-1) combined with a surfactant (Wuertz, 1992). 17.11.4 Chemical Plugging Chemical precipitates can cause plugging of irrigation systems (Nakayama and Bucks, 1986; Hills et al., 1989, Burt and Styles, 1994; Burt et al., 1995). Precipitates of iron oxides (reddish-brown color), iron sulfides (black color), calcium carbonates (white color), and manganese oxides (black color) in irrigation water can clog emitters. Changes in water pH, temperature, pressure, dissolved oxygen levels, and injections of chlorine and other chemicals (e.g., some fertilizers) can induce chemical precipitations. Groundwater supply systems tend to have greater problems with chemical precipitates than surface water, and the insoluble precipitates that form may not dissolve even after treatments such as lowering the water pH or injection of special solvents . Concentrations of 0.15 to 0.22 mg L-1 of iron (>2 mg L-1) in irrigation water may be problematic when water pH exceeds 5. Manganese and iron oxides greater than 2 mg L-1 will need treatment if water pH is 4 or greater. Chlorine injection will cause oxidation and precipitation of iron and manganese (plus kill any iron and other precipitating bacteria that are present). The general recommendations are to inject 1 mg L-1 of free chlorine per 0.7 mg L-1 soluble iron or 1.33 mg L-1 free chlorine per mg L-1 of soluble manganese prior to the filtration system. Calcium and magnesium problems are best addressed by the injection of acids to maintain a water pH between 6.0 and 6.6. Temporary storage of water in ponds or other open containers with agitation is a recommended pretreatment for elevated levels of iron (>4 mg L-1) and manganese to facilitate more oxidation and precipitation before entering the irrigation system. Plugging by other chemical precipitates can often be reduced by acid treatment to lower pH and prevent precipitates from forming, avoiding the injection of insoluble fertilizers or incompatible fertilizer/chemical mixes, and regular flushing of lines. It is possible to reduce the risk of precipitation problems by carefully injecting various incompatible chemicals at different locations in a mainline so that they are sufficiently diluted and mixed before the next chemical is injected. For example, injection points for acid and chlorine should be a minimum of 1 m apart.

17.12 SUBSURFACE DRIP IRRIGATION Subsurface drip irrigation (SDI) uses buried lateral pipelines and emitters to apply water directly in the plant root zone. Laterals are placed deep enough to avoid damage by normal tillage operations, but sufficiently shallow so that water is redistributed in the active crop root zone by capillarity. SDI systems must be compatible with the total farming and cultural systems being used. Current commercial and grower interest levels indicate that future use of SDI systems will continue to increase. SDI requires the highest level of management of all microirrigation systems to avoid remedial maintenance. A poorly designed SDI system is much less forgiving than an improperly designed surface drip system. Deficiencies and water distribution problems are difficult and expensive to remedy. Lamm and Camp (2007) present an excellent, detailed review of SDI.

Design and Operation of Farm Irrigation Systems

669

These systems require safeguards and special operational procedures to prevent plugging and facilitate maintenance, but they also have numerous advantages. Jorgenson and Norum (1992) have presented an overview of the theory behind SDI as well as varied grower experiences and applications. Camp (1998) has prepared an excellent summary and analysis of published subsurface drip irrigation research results. Phene et al. (1987a; 1992) and Phene (1995) listed four operational characteristics of SDI relative to surface drip irrigation installations with respect to water conservation and salinity. These were: ƒ The top of the soil surface remains dry, limiting surface evaporation to the rate of vapor diffusion transport and preventing salt accumulations on the surface. ƒ The use of a very high irrigation frequency (several times per day) that matches actual crop water use will result in a constant wetted soil volume and a net upward hydraulic gradient, which minimizes leaching. ƒ Supplying water and nutrients directly to the root zone allows root uptake to be more efficient if irrigation and fertilization schedules are appropriate. ƒ Soil crusts, which may impede infiltration and cause ponding and runoff, are bypassed so that surface infiltration variability becomes insignificant. Also, Camp et al. (1987) and Grimes et al. (1990) found more uniform soil water distribution under subsurface than surface drip systems. Under proper management, properly designed and managed SDI irrigation systems offer several other advantages to growers because of their potential for: ƒ maintaining access to fields with tillage, planting, spray and harvest equipment that is not restricted by irrigation; ƒ obtaining better weed suppression with minimal chemicals because there is less seed germination with dry soil surfaces; ƒ efficiently and safely applying labeled plant-systemic pesticides and soil fumigants for improved disease and pest control; ƒ reducing surface wetting often reduces fungal disease incidence (e.g., molds, mildews) by maintaining dryer plant surfaces and lower air humidity within the plant canopy; ƒ reducing pesticide exposures for workers when chemicals are applied below the soil surface; ƒ implementing minimum tillage, permanent beds, and multiple cropping systems (Bucks et al., 1981), although much of the necessary equipment modifications and farming techniques remain to be developed; and ƒ minimizing flow-rate sensitivity to temperature fluctuations because emitters are buffered by the soil. Phene et al. (1987a; 1992) and Phene (1995) also listed several disadvantages, including: ƒ initial system cost may be high; ƒ potential for rodent damage; ƒ salt may accumulate between drip lines and soil surface; ƒ low upward water movement in coarse-textured soils; ƒ high potential for emitter plugging; and ƒ insufficient technical knowledge, dissemination, and hands-on experience by growers and researchers. Specific problems that have been observed include plugging by root intrusion (Tollef-

670

Chapter 17 Microirrigation Systems

son, 1988; Bui, 1990); pinching of hoses due to compaction or squeezing by large roots (Bui, 1990); and rodent and insect damage (Bui, 1990). In addition, fertility management becomes more critical with SDI because roots tend to grow deeper than with surface drip systems and some surface applied nutrients may not be sufficiently available (Phene, 1995). A broad range of yield increases have been observed under SDI when compared to surface, sprinkler, and even surface drip irrigation systems from small to up to more than a 100% difference. SDI research has been reported on crops including cotton (Plaut et al., 1985; Tollefson, 1988; Constable and Hodson, 1990; Hutmacher et al., 1995), field corn (Lamm et al., 1995), sweet corn (Bar-Yosef et al., 1989), tomatoes (Davis et al., 1985; Phene et al., 1987a; Grattan et al., 1988; Bogle et al., 1989), cantaloupe (Phene et al., 1987b), potatoes (Bisconer, 1987), asparagus (Sterret et al., 1990), alfalfa (Oron et al., 1989; Bui and Osgood, 1990), cabbage and zucchini (Rubeiz et al., 1989). Most yield increases have been attributed to better fertilization, better water management, improved water distribution uniformities, and improved disease and pest control. Grattan et al. (1988) cited better weed control as the major factor in their observed yield increases. Moore and Fitschen (1990) reported that the conversion of 5900 hectares of sugarcane in Hawaii from furrow to SDI over a period of 12 yr resulted in an average 27% net yield increase. 17.12.1 Design of SDI Systems Designs should follow the same general requirements as for all microirrigation systems. However, extra attention to filtration, water treatment, pressure regulation, proper location of check and air-vacuum relief valves, and flushing are crucial to the success of SDI systems. As with surface microirrigation systems, injected chemicals and fertilizers must be compatible with all other injected chemicals, and control of water pH is critical. Both tapes and tubing have been used successfully for SDI. Solomon (1992) discussed trade-offs and selection criteria of emission devices for SDI. A major factor in the life of these systems is tape wall thickness with thicker-walled tapes generally lasting longer. However, even thicker tapes require special considerations and must be buried deep enough to avoid being disturbed by tillage or harvesting equipment, but shallow enough to prevent permanent collapsing of the tape by soil weight or cultural operations. SDI tapes are often used with high-value, shallow-rooted (e.g., strawberries) and annual or biannual crops (e.g., various vegetables, melons, sugar cane, cotton). They are typically placed at shallower depths in semipermanent minitill beds on land with little slope or short runs because tapes are not pressure-compensating. As a general rule, there should never be any farm equipment wheel traffic or other activities directly over the tape that would compact the soil and flatten the tape, permanently destroying its utility. Tubing is used more often than tape for SDI on perennial crops, and may be placed at deeper depths than for annual crops. Pressure-compensating emitters on tubing allow SDI systems to be used on diverse topography with steep slopes. Tubing is more rigid and thus more resistant to pinching and compaction, but the same general operational considerations for tapes should also be followed. Individual SDI laterals are often connected with manifolds at both the top and distal end of the tubing. The bottom manifold typically provides greater flexibility and saves time with flushing operations. The additional expense also creates a hydraulically

Design and Operation of Farm Irrigation Systems

671

looped system that can supply water from both ends of a lateral. This can be beneficial over the short term if an individual lateral is pinched by roots or soil compaction caused by farm equipment. Provisions for adequate air-vacuum relief on both the head and flushing manifolds is critical to reduce emitter plugging due to soil particles being pulled into the orifices by vacuum conditions at shutdown. Wetting the soil surface, or “surfacing,” negates many of the benefits of SDI and is usually undesirable, especially on permanent crops. It can increase weed growth, move salts to the surface, increase soil evaporation, and increase fungal disease incidence. Reducing the negative effects of surfacing may require plastic film mulches and more aggressive water and soil management practices. On the other hand, surfacing is sometimes appropriate to ensure adequate wetting across the full width of a bed for germination, fumigation, and/or proper growth of shallow-rooted crops (e.g., onions). Surfacing is primarily the result of water application rates exceeding the saturated hydraulic conductivity. However, it may also occur if the tubing is placed too shallow and the surface is wetted by capillarity, or if water from several emitters runs along the lateral and collects at a point and the low gravity head forces water to the soil surface. In addition, surfacing may be the end result of chemical processes that effectively reduce saturated hydraulic conductivity, including high-bicarbonate waters that precipitate calcium and plug soil pores, application of irrigation water with low electrical conductivity on saline or sodic soils, or application of water with high sodium levels (see Chapter 7). Some growers in California have reportedly tried using narrow shanks ahead of the buried tubing installation machine to till deeper than tubing placement as a way to discourage surfacing, with mixed results. Nevertheless, exceeding the effective saturated hydraulic conductivity with SDI systems causes water to be discharged against higher-than-atmospheric pressures, forcing the water to the soil surface. Because manufacturers design drippers to apply water at atmospheric pressure, this back pressure reduces emitter flow rates as much as 50% (Sadler et al., 1995; Shani et al., 1996; Warrick and Shani, 1996). Thus, surfacing negatively affects uniformity, the ability to accurately schedule water application amounts, and presents significant design as well as management implications. The surfacing process often creates light-textured “chimneys” where the fine soil particles are flushed from around the emitter resulting in a direct, low-pressure pathway to the soil surface that tends to make the problem worse. These chimneys will sometimes disappear after a year or two as the soil structure adjusts or pathways along the tubes disperse, but they usually require tillage for remediation. Short, frequent pulsing of water applications throughout the day may help reduce surfacing problems. Low-EC waters may require injection of finely powdered gypsum (e.g., >2.5 meq L-1) or other amendments into the water to improve infiltration on sodic soils. To avoid surfacing problems, emitters on SDI tend to have smaller flow rates and closer spacing than surface drip systems, thus creating a higher potential for plugging. Consequently, filters should be designed to remove particles as small as 150 to 200 microns. Check valves, vacuum breakers, and air vents must be installed to prevent backsiphoning and aspiration of soil particles into the emitter orifices at system shutdown. Designs must ensure that each lateral can be flushed properly. 17.12.1.1 Depth considerations. The ideal depth of buried SDI laterals depends on discharge rate, emitter spacing, soil type, root distributions of crops in a rotation, water

672

Chapter 17 Microirrigation Systems

movement for seed germination (also affected by seed depth and soil salinity), chemigation programs, and the desirability of surface wetting. The tubing must be buried deep enough so that it is not damaged by equipment, but shallow enough to effectively move water into the crop root zone. Tubing and tape depths may be adjusted on bed systems by use of equipment to remove or replace soil. It is generally recommended that the tubing be placed as shallow as possible without surface wetting because most of the roots, air exchange, and biological activity is located in the upper soil layers. As a general rule, tubing is placed at shallower depths on coarse-textured soils and slightly deeper with finer-textured soils. Most SDI systems are installed at 10 to 50 cm depths. SDI can be installed on shallow-rooted row crops such as strawberries or onions at 2- to 8-cm depths up to 40 to 50 cm deep for crops such as cotton, maize, potatoes, or sugar beets. Many vegetable crops use SDI laterals at 15- to 20-cm depths near the plants. Thin-walled tape products are rarely buried more than 15 cm below the surface in beds. There are trade-offs with tube placement depth. Philip (1991) showed for a steady state subsurface irrigation source that the deeper the source relative to the roots, the larger the deep percolation and the smaller the soil evaporation. Barth (1995) discussed placing V-shaped impermeable foil barriers under the SDI lateral to control downward movement of water and reported good results on sandy soils. Deeper installations reduce surface weed germination and allow more tillage operations. However, if laterals are too deep, much of the water is applied below the crop root zone, reducing biotic activity (critical to nutrient uptake) in the shallow soil layers. In addition, deep installations may limit crop germination potential and limit use of some surfaceapplied chemicals. It may be possible to inject air through deeper lines to encourage more root development and soil biota as well as aerate waterlogged soils, but this practice has not been thoroughly investigated. Shallow depths may result in excessive surface soil salinity when using saline waters and may also be more subject to damage from burrowing rodents and insects. Deeper tubing placements will require more tractor power for installation. Shallow tube placement may be necessary to supply water near the soil surface on plantings with limited root zones. For some soils it is possible to wet a seed bed from a depth of 30 to 50 cm. However, on coarse-textured soils, emitter lines may have to be within 5 to 20 cm of the surface. 17.12.1.2 SDI on annual row crops. Hanson et al. (1994) discussed installation, operation, and maintenance of SDI for row crops. Annual row crops can be grown on temporary or permanent beds with or without plastic mulches However, it is sometimes necessary to use sprinkler irrigation for activating herbicides or germination of small-seeded crops such as lettuce or onions, particularly under saline soil conditions. Emitters with high flow rate or tapes with closely spaced emission points (every 20 to 45 cm) that wet the soil surface are common on annual crops. Shallow tapes are generally expected to last 1 to 2 yr, although some installations of thicker-walled tapes on permanent beds may last 5 to 15 yr. These require special tillage and cultural practices (such as controlled traffic and permanent wheel tracks) to avoid damage. To ensure adequate wetting of the root zone of annual crops and to avoid dry areas due to emitter plugging, emitters should be spaced close enough together to produce an overlapped, line-source wetting pattern. Spacing between emitter lines or tapes are often 1 to 1.5 m but will vary depending on soil type, crop characteristics, and cultural practices.

Design and Operation of Farm Irrigation Systems

673

Use of semi-permanent SDI on annual or biannual row crops requires the establishment of consistent row spacing for all crops used in the rotation and replanting at the same locations. If these practices are not followed, the drip lateral position may vary from being centered under the bed and row to being located under the furrow. However, research results have demonstrated that yields were not reduced when the drip tubing was not exactly centered under the bed (Ayars et al., 1995) on cotton. But, the incidence of mechanical damage increased as the lateral line was moved from the center towards the edge of the bed. Emerging technologies such as auto-steer on tractors and harvesters will help ensure proper alignment between beds and drip lines year after year.. 17.12.1.3 SDI on perennial crops. SDI systems on perennial tree and vine crops should have a life expectancy of 7 to 20 yr with appropriate design and maintenance. Depths of 20 to 40 cm are common with some as deep as 1 m depending on crop root development patterns and soil type. Generally, low-flow emitters (