Geotechnical Design for Sublevel Open Stoping

Ernesto Villaescusa

Western Australian School of Mines

Geotechnical Design for Sublevel Open Stoping

Geotechnical Design for Sublevel Open Stoping Ernesto Villaescusa

Boca Raton London New York

CRC Press is an imprint of the Taylor & Francis Group, an informa business

CRC Press Taylor & Francis Group 6000 Broken Sound Parkway NW, Suite 300 Boca Raton, FL 33487-2742 © 2014 by Taylor & Francis Group, LLC CRC Press is an imprint of Taylor & Francis Group, an Informa business No claim to original U.S. Government works Version Date: 20130923 International Standard Book Number-13: 978-1-4822-1189-4 (eBook - PDF) This book contains information obtained from authentic and highly regarded sources. Reasonable efforts have been made to publish reliable data and information, but the author and publisher cannot assume responsibility for the validity of all materials or the consequences of their use. The authors and publishers have attempted to trace the copyright holders of all material reproduced in this publication and apologize to copyright holders if permission to publish in this form has not been obtained. If any copyright material has not been acknowledged please write and let us know so we may rectify in any future reprint. Except as permitted under U.S. Copyright Law, no part of this book may be reprinted, reproduced, transmitted, or utilized in any form by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying, microfilming, and recording, or in any information storage or retrieval system, without written permission from the publishers. For permission to photocopy or use material electronically from this work, please access www.copyright. com (http://www.copyright.com/) or contact the Copyright Clearance Center, Inc. (CCC), 222 Rosewood Drive, Danvers, MA 01923, 978-750-8400. CCC is a not-for-profit organization that provides licenses and registration for a variety of users. For organizations that have been granted a photocopy license by the CCC, a separate system of payment has been arranged. Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. Visit the Taylor & Francis Web site at http://www.taylorandfrancis.com and the CRC Press Web site at http://www.crcpress.com

Contents Foreword............................................................................................................... xiii Preface................................................................................................................... xvii Acknowledgments............................................................................................... xix Author.................................................................................................................... xxi 1. Introduction......................................................................................................1 1.1 Mining Method Selection..................................................................... 1 1.2 Self-Supported Mining Methods......................................................... 1 1.3 Sublevel Open Stoping..........................................................................3 1.4 Factors Controlling Stope Wall Behavior...........................................7 1.4.1 Excavation Geometry............................................................... 7 1.4.2 Rock Mass Strength..................................................................9 1.4.3 Induced Stresses...................................................................... 12 1.4.4 Ground Support...................................................................... 13 1.4.5 Blast Damage........................................................................... 15 1.4.6 Drill Drive Layout................................................................... 16 1.5 Scope and Contents of This Book...................................................... 17 2. Sublevel Stoping Geometry........................................................................ 19 2.1 Introduction.......................................................................................... 19 2.2 Stoping Geometries............................................................................. 19 2.2.1 Cutoff Slot................................................................................ 19 2.2.2 Production Rings....................................................................22 2.2.3 Diaphragm Rings....................................................................22 2.2.4 Trough Undercut..................................................................... 23 2.2.5 Drawpoints.............................................................................. 26 2.3 Multiple-Lift Open Stoping................................................................ 26 2.3.1 Tabular Orebodies.................................................................. 28 2.3.2 Massive Orebodies................................................................. 29 2.4 Single-Lift Stoping............................................................................... 31 2.4.1 Conventional Vertical Crater Retreat Stoping....................34 2.4.2 Modified Vertical Retreat Stoping........................................ 36 2.5 Shallow Dipping Tabular Orebodies................................................ 37 2.6 Bench Stoping....................................................................................... 38 3. Planning and Design.................................................................................... 47 3.1 Introduction.......................................................................................... 47 3.2 Geological and Geotechnical Characterization............................... 49 3.3 Stress Analysis in Stope Design......................................................... 49 3.4 Design of Stoping Blocks.................................................................... 52 v

vi

Contents

3.4.1 3.4.2

3.5

Orebody Delineation.............................................................. 53 Global Extraction Sequences................................................. 53 3.4.2.1 Massive Orebodies..................................................54 3.4.2.2 Steeply Dipping Orebodies.................................... 58 3.4.3 Numerical Modeling.............................................................. 73 3.4.4 Regional Pillars....................................................................... 75 3.4.5 Block Development................................................................. 78 3.4.5.1 Shaft Stability........................................................... 78 3.4.5.2 Ramp Access............................................................ 81 3.4.5.3 Crown Pillar............................................................. 82 3.4.5.4 Sublevel Interval......................................................84 3.4.5.5 Access Crosscuts.....................................................84 3.4.5.6 Raises and Orepasses............................................. 86 3.4.5.7 Fill Infrastructure.................................................... 86 3.4.6 Stope Production Scheduling................................................ 89 3.4.6.1 Long-Term Production Scheduling.......................90 3.4.6.2 Medium-Term Activity Schedules........................ 90 3.4.6.3 Short-Term Activity Schedules.............................. 91 3.4.7 Ventilation................................................................................ 92 3.4.8 Global Economic Assessment............................................... 93 Detailed Stope Design......................................................................... 93 3.5.1 Geological Information.......................................................... 97 3.5.2 Development............................................................................ 98 3.5.3 Geotechnical Assessment.................................................... 100 3.5.4 Stope Design Philosophy..................................................... 102 3.5.4.1 Production Rings................................................... 102 3.5.4.2 Diaphragm Rings.................................................. 103 3.5.4.3 Cutoff Slot Design................................................. 104 3.5.4.4 Drawpoint Design................................................. 104 3.5.5 Stope Design Note................................................................ 106 3.5.6 Stope Firing Sequences........................................................ 107 3.5.7 Production Monitoring........................................................ 109 3.5.8 Ventilation.............................................................................. 110 3.5.9 Financial Analysis................................................................ 111

4. Rock Mass Characterization..................................................................... 113 4.1 Introduction........................................................................................ 113 4.2 Characterization from Exploration Core........................................ 115 4.2.1 Drilling Layout Design........................................................ 118 4.2.2 Underground Drilling.......................................................... 118 4.2.3 Core Transfer to Surface....................................................... 118 4.2.4 Drill Core Logging............................................................... 119 4.2.5 Geological Database............................................................. 120 4.2.6 Interpretation of the Orebody and Main Geological Features.............................................................. 120

Contents

4.3

4.4

4.5

4.6

4.7 4.8

vii

4.2.7 Orebody Meshing in Three Dimensions........................... 121 4.2.8 Problems with Data Analysis.............................................. 121 Analysis of Logging Data................................................................. 122 4.3.1 Discontinuity Linear Frequency......................................... 122 4.3.2 Rock Quality Designation................................................... 125 4.3.3 Rock Mass Classifications from Core Logging................. 131 4.3.4 Advantages, Disadvantages, and Biases in Core Logging.................................................................... 141 Geotechnical Mapping of Underground Exposures.................... 142 4.4.1 Cell Mapping......................................................................... 144 4.4.2 Line Mapping........................................................................ 145 4.4.3 Strip Mapping....................................................................... 146 4.4.4 Description of Mapping Parameters.................................. 147 4.4.5 Mapping Biases..................................................................... 151 4.4.6 Geological Strength Index................................................... 152 Analysis of Mapping Data................................................................ 153 4.5.1 Discontinuity Orientation................................................... 153 4.5.2 Number of Discontinuity Sets............................................ 155 4.5.3 Discontinuity Spacing.......................................................... 157 4.5.4 Discontinuity Trace Length................................................. 159 4.5.5 Rock Mass Classification Models....................................... 164 Intact Rock Strength.......................................................................... 165 4.6.1 Uniaxial Compressive Strength.......................................... 168 4.6.2 Point Load Strength.............................................................. 173 4.6.3 Confined Compressive Strength......................................... 175 Mechanical Properties of Rock Masses.......................................... 178 4.7.1 Hoek–Brown Empirical Strength Criterion...................... 179 4.7.2 Rock Mass Deformation Modulus...................................... 182 Rock Stress.......................................................................................... 183 4.8.1 Stress Tensor.......................................................................... 184 4.8.2 Stress Measurements Using Oriented Core...................... 185

5. Span and Pillar Design............................................................................... 191 5.1 Background......................................................................................... 191 5.2 Empirical Span Determination Using Rock Mass Classification Methods...................................................................... 191 5.2.1 Span Determination Using Bieniawski’s RMR System.......................................................................... 192 5.2.2 Span Determination Using the Tunnel Quality Index (Q) System.................................................... 197 5.3 Stability Graph Method..................................................................... 197 5.3.1 Updated Determination of the Stability Graph Parameters................................................................. 200 5.3.1.1 Factor A................................................................... 201 5.3.1.2 Factor B................................................................... 203

viii

Contents

5.4 5.5

5.3.1.3 Factor C................................................................... 206 5.3.1.4 Hydraulic Radius.................................................. 207 5.3.2 Prediction of Stope Stability................................................ 209 5.3.3 Use of the Stability Graph as a Design Tool...................... 213 5.3.4 Design Validation................................................................. 219 Numerical Modeling of Stope Wall Stability.................................222 5.4.1 Linear Elastic Numerical Modeling...................................223 5.4.2 Nonlinear Numerical Modeling.........................................225 Pillar Stability Analysis..................................................................... 231 5.5.1 Basic Concepts....................................................................... 231 5.5.2 Average Pillar Stress Using the Equivalent Area Approach...................................................................... 232 5.5.3 Empirical Rib Pillar Stability Chart................................... 233 5.5.4 Confinement Pillar Stability Chart....................................234 5.5.5 Numerical Modeling for Pillar Design.............................. 240

6. Drilling and Blasting.................................................................................. 245 6.1 Introduction........................................................................................ 245 6.2 Longhole Drilling............................................................................... 245 6.2.1 Top-Hammer Drilling.......................................................... 247 6.2.2 In-the-Hole Drilling.............................................................. 247 6.2.3 Drilling Equipment Selection............................................. 248 6.2.4 Drilling Deviation................................................................. 249 6.2.4.1 Collar Positioning.................................................. 250 6.2.4.2 Drillhole Alignment............................................. 251 6.2.4.3 In-the-Hole Deviation........................................... 252 6.3 Blast Design Parameters................................................................... 258 6.3.1 Drilling Orientation.............................................................. 260 6.3.2 Blasthole Diameter................................................................ 262 6.3.3 Blasthole Length................................................................... 264 6.3.4 Burden.................................................................................... 265 6.3.5 Spacing................................................................................... 267 6.3.6 Stemming and Uncharged Length..................................... 268 6.4 Ring Design........................................................................................ 269 6.4.1 General Procedure................................................................ 270 6.4.2 Parallel Patterns.................................................................... 273 6.4.3 Radial Patterns...................................................................... 274 6.4.4 Vertical Crater Retreat Blasting.......................................... 277 6.5 Explosive Selection............................................................................ 280 6.5.1 Packaged versus Bulk Explosives....................................... 281 6.5.2 Ammonium Nitrate-Based Explosives.............................. 281 6.5.3 ANFO..................................................................................... 282 6.5.4 Watergels or Slurries............................................................ 282 6.5.5 Emulsions............................................................................... 283 6.5.6 Special ANFO and Emulsion Blends................................. 285

Contents

ix

6.6

Explosive Placement.......................................................................... 285 6.6.1 Powder Factor........................................................................ 287 6.6.2 Energy Distribution.............................................................. 288 6.7 Initiation Systems............................................................................... 289 6.7.1 Pyrotechnic Delay Element Detonators............................. 289 6.7.2 Available Timing and Sources of Timing Error for Pyrotechnic Delay Elements............................................... 290 6.7.3 Electronic Delay Element Detonators................................ 292 6.7.4 Priming................................................................................... 294 6.7.5 Sequencing and Timing....................................................... 295 6.8 Raise and Cutoff Slot Blasting.......................................................... 298 6.8.1 Longhole Winzes.................................................................. 298 6.8.2 Cutoff Slots............................................................................. 302 6.9 Trough Undercut Blasting................................................................ 307 6.10 Rock Diaphragm Blasting.................................................................308 6.11 Mass Blasting......................................................................................309 6.11.1 Control of Ground Vibration............................................... 312 7. Rock Reinforcement and Support............................................................ 315 7.1 Introduction........................................................................................ 315 7.2 Terminology........................................................................................ 317 7.2.1 Continuous Mechanical Coupled....................................... 318 7.2.2 Continuous Friction Coupled.............................................. 318 7.2.3 Discrete Mechanical and Friction Coupled...................... 319 7.2.4 Load Transfer Concept......................................................... 319 7.2.5 Embedment Length Concept.............................................. 320 7.2.6 Reinforcement Performance Indicators............................. 321 7.3 Ground Support Design.................................................................... 322 7.3.1 Location of Failure due to Overstressing.......................... 324 7.3.2 Depth of Failure: Stress or Strain Controlled................... 324 7.3.3 Depth of Failure: Structurally Controlled......................... 326 7.3.4 Ground Reaction Curve Concept....................................... 328 7.3.5 Ground Support for Massive Rock and Low Stress......... 329 7.3.6 Ground Support for Massive Rock and Moderate Stress..................................................................... 329 7.3.7 Ground Support for Massive Rock and High Stress....... 330 7.3.8 Ground Support for Layered Rock and Low Stress......... 332 7.3.9 Ground Support for Layered Rock and Moderate Stress..................................................................... 333 7.3.10 Ground Support for Layered Rock and High Stress.......334 7.3.11 Ground Support for Jointed Rock and Low Stress..........334 7.3.12 Ground Support for Jointed Rock and Moderate Stress.................................................................... 336 7.3.13 Ground Support for Jointed Rock and High Stress......... 337 7.3.14 Design by Precedent Rules.................................................. 338

x

Contents

7.4

7.5

7.6 7.7 7.8

7.3.15 Design by Rock Mass Classification...................................340 7.3.16 Reinforcement Layout..........................................................343 7.3.17 Energy Release......................................................................343 7.3.18 Rock Mass Demand..............................................................344 Rock Bolting of Open Stope Development Drives........................345 7.4.1 Continuous Mechanical Coupled Rock Bolts...................346 7.4.1.1 Cement-Encapsulated Threaded Bar..................346 7.4.1.2 Resin-Encapsulated Threaded Bar..................... 347 7.4.2 Continuous Friction Coupled Rock Bolts.......................... 352 7.4.2.1 Split-Tube Friction Rock Stabilizers.................... 352 7.4.3 Discrete Mechanical or Friction Coupled Rock Bolts......354 7.4.3.1 Expansion Shell Rock Bolts.................................. 355 7.4.4 Rock Bolts with Yielding Mechanisms............................... 357 Cable Bolting of Open Stope Walls................................................. 360 7.5.1 Cable Bolt Reinforcement Mechanisms............................. 363 7.5.2 Cable Bolt Types.................................................................... 366 7.5.2.1 Plain Strand Cable Bolts....................................... 366 7.5.2.2 Modified Strand Cable Bolts................................ 366 7.5.2.3 Debonded Plain Strand Cable Bolts................... 368 7.5.2.4 Cable Bolt Plates.................................................... 369 Cable Bolt Corrosion.......................................................................... 370 7.6.1 Corrosivity of Cable Bolt Strands....................................... 370 7.6.2 Corrosivity of Cable Bolt Anchors...................................... 374 Cement Grouting of Cable Bolts...................................................... 378 7.7.1 Collar to Toe Grouting......................................................... 378 7.7.2 Toe to Collar Grouting......................................................... 380 Support Systems................................................................................. 383 7.8.1 Plates....................................................................................... 383 7.8.2 Straps......................................................................................384 7.8.3 Mesh........................................................................................ 385 7.8.3.1 Mesh Testing.......................................................... 386 7.8.3.2 Mesh Force and Displacement............................ 389 7.8.4 Thin Spray on Liners............................................................ 395 7.8.5 Shotcrete Layers.................................................................... 395 7.8.5.1 Shotcrete Support Mechanisms.......................... 396 7.8.5.2 Shotcrete Reaction to Transverse Loading........ 397 7.8.5.3 Shotcrete Reaction in Tension.............................. 398 7.8.5.4 Shotcrete Reaction in Compression.................... 398 7.8.5.5 Shotcrete Toughness............................................. 399

8. Mine Fill........................................................................................................ 405 8.1 Introduction........................................................................................ 405 8.2 Unconsolidated Rock Fill.................................................................. 406 8.2.1 Rock Fill for Bench Stope Support...................................... 409 8.3 Cemented Rock Fill............................................................................ 412

Contents

8.4 8.5 8.6

8.7

xi

8.3.1 Cemented Aggregate Fill..................................................... 413 Hydraulic Fill...................................................................................... 418 Cemented Paste Fill...........................................................................422 Open Stope Fill Operations Systems............................................... 426 8.6.1 Material Preparation............................................................. 427 8.6.1.1 Chemistry and Mineralogy................................. 428 8.6.1.2 Particle Size Distribution..................................... 428 8.6.1.3 Binders.................................................................... 428 8.6.1.4 Admixtures............................................................ 431 8.6.1.5 Mixing Water......................................................... 431 8.6.1.6 Mix Design............................................................. 432 8.6.2 Stope Preparation..................................................................434 8.6.2.1 Design Criteria for Fill Barricades......................434 8.6.2.2 CHF Barricades......................................................434 8.6.2.3 CRF Barricades...................................................... 437 8.6.2.4 CPF Barricades....................................................... 438 8.6.3 Material Delivery.................................................................. 439 8.6.3.1 Rock Fill Passes......................................................440 8.6.3.2 Slurry Fill Passes...................................................440 8.6.4 Fill Placement........................................................................ 441 8.6.4.1 CHF Placement...................................................... 441 8.6.4.2 CRF Placement....................................................... 441 8.6.4.3 CPF Placement.......................................................442 Fill Monitoring and Quality Control..............................................443 8.7.1 Fill Supply..............................................................................443 8.7.2 Fill Plant.................................................................................443 8.7.3 Fill Reticulation.....................................................................444 8.7.4 Fill Placement........................................................................444 8.7.5 Barricade Performance.........................................................445

9. Dilution Control.......................................................................................... 447 9.1 Introduction........................................................................................ 447 9.2 Types of Dilution................................................................................ 449 9.2.1 Internal Dilution................................................................... 449 9.2.2 External Dilution................................................................... 450 9.2.3 Geological Dilution............................................................... 451 9.2.4 Ore Loss................................................................................. 451 9.3 Economic Impact of Dilution............................................................ 452 9.4 Parameters Influencing Dilution..................................................... 453 9.4.1 Dilution at the Orebody Delineation Stage....................... 455 9.4.2 Dilution at the Design and Sequencing Stages................. 456 9.4.3 Dilution at the Stope Development Stages........................ 458 9.4.4 Dilution at the Production Drilling and Blasting Stages............................................................... 459 9.4.5 Dilution at the Production Stages....................................... 460

xii

Contents

9.5 9.6 9.7

9.4.6 Dilution Issues for Mine Management..............................463 Cavity Monitoring System................................................................464 Dilution Control Plan........................................................................ 466 9.6.1 Stope Performance Review.................................................. 468 Scale-Independent Measures of Stope Performance..................... 472 9.7.1 Conventional Measures....................................................... 474 9.7.2 Circularity Measures............................................................ 476 9.7.3 Extensivity Measures........................................................... 476 9.7.4 Hemisphericity Measures.................................................... 477 9.7.5 Cannington Mine Example................................................. 478

References............................................................................................................ 481

Foreword Underground metalliferous mining in Australia began in the mid-1840s at the copper and silver–lead mines in and around Kapunda and Burra in South Australia. Mining in the Victorian goldfields following the discovery of gold and the Gold Rush of 1851 was initially alluvial but soon evolved into the underground mining of deep leads and then quartz veins. By 1895, the 180 Mine at Bendigo was, at 970 m deep, the deepest mine in the world. The rich silver–lead–zinc orebodies of Broken Hill were discovered in 1883 and gold in Western Australia in 1892. By that time, Australia’s mining industry had already seen a number of boom-and-bust cycles. However, new discoveries have continued to be made and new mines developed up to the present day, with mining remaining a mainstay of Australia’s export economy, particularly in recent decades. In the 1950s, the dry fill formerly used was replaced by hydraulically placed fill in a number of Australian underground metalliferous mines. Mechanized cut-and-fill methods of mining were introduced for the lead orebodies at Mount Isa in 1964 and were soon adopted by other mines. During the 1960s, mining in Australia and elsewhere benefited greatly from the advances that were then taking place in the emerging science of rock mechanics. By the 1970s, cut-and-fill was one of the major underground metalliferous mining methods used in Australia, and in Canada and Scandinavia as well, but demand for higher productivity led to a transition to a range of sublevel and longhole open stoping methods, usually with backfill, until these became the most widely used methods in Australia. Although mass mining methods using sublevel and block caving have been used increasingly since the 1990s for mining some types of orebody, sublevel open stoping remains the primary method used for the underground mining of base and precious metals in Australia. The mining literature of recent decades includes conference proceedings and specialist monographs on cut-and-fill and caving methods of mining and on the mining of tabular orebodies such as the deep level gold-bearing reefs of South Africa. Because of its continuing importance in most of the world’s major metalliferous mining countries, including, but not limited to, Australia, Canada, and the Scandinavian and South American countries, it is entirely appropriate that a book should now appear synthesizing 40 years’ accumulated international experience with modern sublevel open stoping methods. As will be argued in the following text, the author of this book, Professor Ernesto Villaescusa, is supremely well qualified to undertake this important task. I first met Ernesto Villaescusa in early 1988 shortly after I had moved to the University of Queensland, Brisbane, Australia, from Imperial xiii

xiv

Foreword

College, London. Ernesto was introduced to me by Professor Alban Lynch, AO, the distinguished foundation director of the University’s world famous Julius Kruttschnitt Mineral Research Centre (JKMRC). Ernesto had just joined the Centre as a research scholar in its then Mining Research Group. He was interested in doing his PhD research in an area of mining rock mechanics and was looking for a supervisor. Previously, Professor Lynch had kindly invited me to become associated with the JKMRC and to carry out my then necessarily limited research-related activities through the Centre. I have to admit that, initially, I was not at all enthusiastic about taking on a PhD student when I was trying to establish myself in a senior position in a new university. However, Ernesto’s enthusiasm, persistence, and determination, and Professor Lynch’s more gentle powers of persuasion, jointly won the day, and I became Ernesto’s PhD supervisor for the next three years. That was the beginning of a friendship and close professional relationship that has continued now for 25 years. After completing an excellent PhD thesis in 1991, Dr. Villaescusa joined Mount Isa Mines as a rock mechanics engineer. In 1994–1995, he spent some time at the Noranda Technology Centre in Canada, before returning to Mount Isa Mines in 1995 as principal rock mechanics engineer. Then in 1997, at a very young age for a full professor in an Australian university, he was appointed professor of mining geomechanics at the Western Australian School of Mines (WASM), Kalgoorlie, a position that he continues to hold in what since 2004 has been the industry-sponsored industry chair in mining rock mechanics. At WASM, Professor Villaescusa has built up a leading applied mining rock mechanics research group, taught mining engineering at undergraduate and postgraduate levels, carried out and/or supervised a wide range of industry-sponsored mining rock mechanics research projects, and acted as a consultant to the industry in Western Australia, elsewhere in Australia, and in South America, mainly in the general area of underground metalliferous mining. Because of his directly relevant industry, applied research, teaching, and consulting experience and his extensive list of publications in the area, Professor Ernesto Villaescusa is eminently well-qualified to write this book, Geotechnical Design for Sublevel Open Stoping. In particular, he has wide practical experience of sublevel open stoping and its variants at a large number of mines that use these and other mining methods in Australia, Canada, Chile, New Zealand, and in his native Mexico. He also has the great advantage of having gained research training and practical mining experience in the basic mining science of rock mechanics. I have enjoyed the unusual privilege of having been asked by Professor Villaescusa to offer comment and advice on the contents of his book as it has developed through the various stages of its preparation. Although, as the title suggests, the book has a geotechnical engineering orientation, it also contains considerable practical detail on open stoping layouts, design, and operations and includes chapters on drilling and blasting, rock support and

Foreword

xv

reinforcement, mine fill technology, and dilution control. Some of this material draws heavily on results obtained, and understandings developed, in industrially sponsored research projects carried out by Professor Villaescusa, his colleagues, and his students at WASM. I believe that this book will serve multiple purposes. It will serve as a specialist textbook for mining courses at the advanced undergraduate and postgraduate levels. It will also provide an authoritative, practically oriented reference work for those involved in the industry, both in mining operations and as consulting engineers, particularly for those in the early stages of their careers and those seeking to develop new understandings and skills. I congratulate Professor Villaescusa on this outstanding achievement and unhesitatingly recommend the book to those having an interest in the industrially important sublevel open stoping methods of underground mining. Edwin T. Brown, AC Senior Consultant, Golder Associates Pty Ltd, Brisbane, Queensland, Australia Emeritus Professor, University of Queensland, Brisbane, Queensland, Australia President, International Society for Rock Mechanics, 1983–1987

Preface Sublevel open stoping is one of the most widely used mining methods in underground metalliferous mining. This method allows for low cost, high recovery, and productivity while providing operational safety to personnel and equipment. The success of the method relies on the stability of stope walls and crowns, as well as any fill masses exposed. Although it is not a selective method, the stope boundaries can be designed so that dilution and ore loss can be minimized. Over the last 30 years or so, increased understanding of the factors controlling stope spans and stability have been developed. In addition, improvements in drilling equipment, ventilation, cablebolt reinforcement, fill mass strength, and routine implementation of stope void monitoring systems have led to significant improvements in sublevel open stoping. In the future, the method is likely to be used under more difficult geotechnical conditions, and therefore, a better understanding of all technical and operating factors influencing its success is required. This book was written primarily for fourth year undergraduate students, graduate students, and junior practitioners not yet entirely familiar with the mining method. The book is divided into nine chapters that closely follow the approach used by most mining houses in implementing the method worldwide. After the basic nomenclature is introduced, the method is reviewed from orebody delineation, planning and design through key operations such as drilling and blasting, ground support of access drives and stope walls, as well as stope void filling. The book also includes a dilution control chapter given that documentation of stope performance is critical to improve the design to optimize the method. The material presented draws heavily on my experience at Mount Isa Mines as well as from technical reviews of many mine sites worldwide. The book also relies upon results of industry-sponsored research undertaken at the Western Australian School of Mines (WASM) over the last 16 years or so. Without the results of my postgraduate students, the book would not have been possible. I wish to record my deep gratitude to Professor Ted Brown, who has provided me with inspiration and advice throughout the entire process of book writing including technical content, layout, and numerous comments for improvement. My gratitude also goes to Professor Will F. Bawden who early in the process provided me with unpublished material and comments to chapters. At WASM, I benefited from the friendship and technical support of the principal research fellows Dr. Alan Thompson and Chris Windsor as well as the administrative and financial support from WASM

xvii

xviii

Preface

directors including Professors Peter Lilly, Eric Grimsey, Paul Dunn, and Steve Hall. I also wish to thank the CRC Mining directors, Professors Mike Hood and Paul Lever, for their financial support. I wish to thank Mount Isa Mines for their permission to use previously unpublished material. Similarly, I wish to thank other organizations and authors who freely gave me permission to reproduce published material. Ernesto Villaescusa Western Australian School of Mines

Acknowledgments This book was written with university undergraduate students in mind. It draws heavily on the knowledge and practical experience gained during my years of employment at Mount Isa Mines from 1991 to 1997, my course notes and interaction with students while teaching underground rock mechanics at the Western Australian School of Mines (WASM) from 1997 to 2007, as well as on outcomes from my WASM research students from 1997 to 2013. I wish to acknowledge the following important contributions to this book: • Professor Ted Brown, AC, who over the years has provided me with many ideas and made invaluable suggestions about the book content and layout and has also carefully reviewed every chapter. His friendship and technical advice started while doing my PhD studies and continues to this day. • Dr. Alan Thompson, who has been a great friend, for his technical support and encouragement, which made writing the book a lot easier. I also acknowledge his deep intellect and his contributions to Chapter 7. • Chris Windsor, who over the years has provided many technical suggestions as to how to improve our research work at WASM. His friendship has always been of great support. I would like to acknowledge his contributions to Chapters 4 and 7. • Professor Will F. Bawden, who provided comments to some early draft chapters and substantially wrote Sections 5.1 and 5.2. He also gave me permission to use his contributions to write Sections 5.2.1 and 5.5. • Dr. Peter Cepuritis, who, as part of his PhD studies, undertook many of the calculations that are included in the book. • Dr. Kelly Fleetwood, who carefully reviewed and made suggestions to Chapter 6 and personally wrote Section 6.5. • Dr. John Player, for his decade-long innovative work in ground support at WASM, some of which is presented in Chapter 7. • Dr. Jianping Li, my first PhD student at WASM, who made many contributions to rock testing and in situ stress measurements, the results of which are reflected in Chapter 4. • Dr. Rhett Hassell, for his contributions to the corrosion work presented in Section 7.6.

xix

xx

Acknowledgments

• Nixon Saw, for his excellent work on fill testing presented in Chapter 8. • Ellen Morton, for her work on mesh and shotcrete testing and for her contributions to Chapter 7. I am also grateful to a number of friends and colleagues that I have worked with at a number of mine sites over the last 20 years or so. Their practical approach and ideas have helped me write this book. In particular, I would like to thank • Leigh Neindorf, Mark Adams, and Mike Sandy, then of Mount Isa Mines • Dave Finn, then of WMC and later Placer Dome • Peter Teasdale, then of WMC • Cam Schubert, then of Mount Isa Mines and later McArthur River Mining I also wish to thank all those who supported me throughout this undertaking, especially • Luis Machuca and Moises Cordova, for their great friendship and constant, unwavering support • Mike Westerman, for his permission to use previously unpublished material from Mount Isa Mines; the support of Mount Isa Mines is also gratefully acknowledged • The WASM rock mechanics technical staff, including Brett Scott, Lance Fraser, and Pat Hogan • The administrative staff and a number of WASM undergraduate and graduate students, including Ben Auld, Tom Parrott, Cesar Pardo, Andrea Roth, Catherine Winder, Ayako Kusui, and Andres Brzovic, among many others • The sponsors of the WASM Rock Mechanics Chair who funded my position at WASM, which include Goldfields, Barrick, Barminco, Newcrest, and Curtin University of Technology; the financial support of CRC Mining is also gratefully acknowledged • The authors and publishers who have given permission for the reproduction of previously published figures and tables • Finally, but most importantly, to Carolyn and Tiana, for their love, patience, tolerance, and understanding of my dedication to exploration, mining, and rock mechanics

Author Professor Ernesto Villaescusa received his BEng in mining engineering (first class honors) from Universidad de Sonora, Mexico in 1984; his MSc in mining engineering from Colorado School of Mines, Golden, Colorado in 1987; and his PhD in mining engineering from the University of Queensland, Brisbane, Queensland, Australia in 1991. He has over 25 years of applied research experience having worked with a large number of mining houses such as MIM Holdings, Noranda, WMC Resources, Peñoles, Minera Autlan, CODELCO, BHP Billiton, Placer Dome Asia Pacific, and Normandy to develop guidelines for effective underground mining, leading to a safe, economical extraction of ore. He has undertaken applied research in all aspects of mining methods for a range of rock mass and geotechnical conditions ranging from shallow depth cut-and-fill mines, room and pillar, to deep sublevel open stopes and block cave mines (the picture below shows him inspecting stope hangingwalls at Mount Isa Mines). Over the last 16 years, he has worked at the Western Australian School of Mines (WASM) as a professor of mining geomechanics, where he has secured over 21 million dollars of ­industry-funded mining research income. He has supervised over 30 master’s and 10 PhD student theses and has written over 100 technical papers. In 2004, he was appointed to an industry chair in mining rock mechanics at WASM. The chair is currently sponsored by Barrick, Goldfields, Barminco, Newcrest, and Curtin University.

xxi

1 Introduction

1.1  Mining Method Selection The design and selection of a mining method requires a systematic approach, with the dip, size, and shape of an orebody; the strengths of the ore and the host rock mass; as well as economics being some of the fundamental parameters influencing the planning and design process (Hamrin, 1982; Brady and Brown, 2004). Distinctions can be made between orebodies having significant width, height, and length and those that are small in one dimension and are either steeply or shallowly dipping. For example, orebodies with significant vertical dimensions can be accessed through drifts developed at successive depths. Gravity is used to advantage in ore-breaking and ore-handling operations, as the broken material can be directed to the conveniently located draw (collection) points. When an orebody is thin, requiring full entry for personnel and equipment, a critical consideration, as the mining face is advanced, is protection from rock falls (Figure 1.1). In most cases, when an orebody is large in all dimensions, access is via small drifts that are located outside the main production zones. The selected mining method will exclude other options on a safety, productivity, recovery, and dilution control basis. Brady and Brown (2004) have discussed the general relation between the geotechnical properties of an orebody, the host rock mass, and the most appropriate mining method.

1.2  Self-Supported Mining Methods The stability of the rock mass greatly influences the choice of mining method. Stable rock masses allow extensive exposures of the backs (roofs) and walls of underground openings (Figure 1.2). Self-supported openings are those in which the overlying load is redistributed through the rock mass and carried by the side walls and pillars. The ore can be removed from an underground

1

2

Geotechnical Design for Sublevel Open Stoping

FIGURE 1.1 Stabilized stope access drift prior to sublevel open stoping extraction.

FIGURE 1.2 Very stable and large stope back in a good-quality rock mass.

opening without the use of materials for back and wall support. F ­ or safety, ground support may still be required at individual locations or at regular intervals. Examples of self-supporting mining methods (Brady and Brown, 2004) include open stope mining (the subject of this book) and room and pillar mining, which will not be discussed further here.

Introduction

3

Sublevel stoping is designed for the progressive extraction of specified ore blocks between pillars of surrounding material. The objective is to mine as much of a deposit as possible in the initial open stopes with low risk of ground movement and without jeopardizing the recovery of adjacent pillar ore. Therefore, open stoping represents an integrated and staged system of total ore recovery. Primary open stoping is usually followed by secondary and sometimes tertiary extraction phases to recover pillar ore. The stope walls must be self-supporting to ensure that the excavation is stable to allow primary stope mining without dilution. The ore should also be strong to ensure stable secondary and tertiary pillars. Pillar recovery requires the use of consolidated fill material that is placed into the primary stope voids to allow stable secondary and sometimes tertiary stope extraction. Although sublevel open stoping is essentially a self-supported mining method, in this sense it can overlap with artificially supported methods as identified by Brady and Brown (2004).

1.3  Sublevel Open Stoping Sublevel open stoping methods are used to extract large massive or tabular, often steeply dipping, competent orebodies surrounded by competent host rocks, which in general have few constraints regarding the shape, size, and continuity of the mineralization. The success of the method relies on the stability of large (mainly unreinforced) stope walls and crowns, as well as the stability of any fill masses exposed. In good quality rock masses, open stopes can be relatively large excavations (Figure 1.3), in which ring drilling and blasting is the main method of rock breakage. Ore dilution consisting of lowgrade, waste rock or minefill materials may occur at the stope boundaries. In addition, ore loss due to insufficient breakage can also occur within the stope boundaries. Two basic stope configurations are possible: longitudinal and transverse. In both configurations, the ore is mined from sublevels by some form of benching and flows by gravity to a drawpoint. Longitudinal sublevel stoping is used for comparatively narrow, usually less than 15 m, steeply dipping orebodies with the stoping running parallel to the strike of the orebody. For thick orebodies, the stopes are oriented perpendicular (transverse) to the strike of the deposit with pillars left between the primary stopes. Full recovery of stope and pillars requires the use of consolidated fill (Brady and Brown, 2004). The method is widely applied worldwide and offers several advantages, including low cost and efficient nonentry production operations. It utilizes highly mechanized, mobile drilling and loading production equipment to achieve high production rates with a minimum level of personnel. Furthermore, the production operations of ring drilling, blasting, and drawpoint mucking are concentrated into few locations. The disadvantages

4

Geotechnical Design for Sublevel Open Stoping

Mount Isa Mines Lead, silver, and zinc stopes

350 m

300 m Typical

Tallest

250 m

200 m

150 m

100 m

50 m

0m FIGURE 1.3 Large-scale stoping operations at Mount Isa Mines. (Courtesy of Mount Isa Mines, Mount Isa, Queensland, Australia.)

include a requirement for a significant level of development infrastructure before production starts, thus incurring a high initial capital investment. However, a large part of the development occurs within the orebody. In addition, the stopes must be designed with regular boundaries, and internal waste pockets cannot be separated within the broken ore. Similarly, delineated ore cannot be recovered beyond a designed stope boundary. Technical developments regarding the understanding of rock mass and fill behavior, in conjunction with dilution measuring techniques, improved blasting, equipment, ventilation, and ground support practices, currently allow for the successful application of this method in increasingly complex geological and mining situations, even at great depth. In particular, an increased understanding of the method is required to facilitate improved stope access configurations and optimized extraction sequences, leading to full orebody recovery while achieving dilution control. The complexity of the method and

Introduction

5

the current depth of the orebodies being extracted worldwide suggest that adequate planning and control of the operations are critical to the successful implementation of optimum stope sizes and sequences of extraction. The method is commonly known throughout the world as open stoping, sublevel stoping, and longhole or blasthole stoping. The following are the essential common elements of sublevel stoping (Mathews, 1978; Bridges, 1983): • The stopes are open and extracted without substantial wall collapse or caving. • The stopes extend from sublevel to sublevel, with operations carried out only at these sublevels. • The blasted rock moves by gravity alone to the stope drawpoints. • The method uses long blastholes for rock breakage, achieving good fragmentation (Figure 1.4). • The blastholes are located within planes called rings. • The holes can be drilled downward or upward. • The initial expansion slot is located on the side, center, or bottom of each stope. • The method is nonentry, and personnel do not have access to the open portion of a stope (Figures 1.5 and 1.6).

FIGURE 1.4 Typical rock fragmentation from sublevel open stoping blasting.

6

Geotechnical Design for Sublevel Open Stoping

FIGURE 1.5 A view inside an open stope. Hangingwall Drill drives

Endwall

Production drill rings Access crosscut

Footwall access drives Extraction level

Trough undercut Drawpoints

Tipple

FIGURE 1.6 Three-dimensional view of a multiple lift, transverse sublevel open stope. (From Villaescusa, E., A review of sublevel stoping, in G. Chitombo, ed., Proceedings of the MassMin 2000, Brisbane, Queensland, Australia, October 29 to November 2, 2000, pp. 577–590, AusIMM, Melbourne, Victoria, Australia. With permission.)

Introduction

7

FIGURE 1.7 A large-scale longitudinal bench stope.

Within this context, the extraction of narrow, lenticular orebodies by longitudinal bench stoping (Villaescusa et al., 1994) is also included among the sublevel stoping geometries and is considered in detail in this book (Figure 1.7). Over the last 20 years or so, technology has been developed to improve the safety and economics of ore extraction by sublevel stoping and benching. Experience indicates that geological discontinuities, stresses, blast damage, excavation geometry, and ground support are the main factors controlling stope wall behavior and stability. These factors will be introduced and discussed briefly in the following subsections. They will be discussed in more detail in the subsequent chapters of this book.

1.4  Factors Controlling Stope Wall Behavior 1.4.1  Excavation Geometry In sublevel stoping, drilling and blasting is undertaken from drilling drives located on sublevels strategically placed over the height of a stope.

8

Geotechnical Design for Sublevel Open Stoping

Stope height

Transition zone S

T

U

S = Stable T = Transitional U= Unstable

Unstable shapes Critical dimension Stope length/width

Unstable shapes Critical dimension Stope height U Transition zone

T S

Stope length or width FIGURE 1.8 Stable shapes for sublevel stoping.

Because of the limited cablebolt reinforcement that can be provided at the exposed stope walls, the excavations must be designed to be inherently stable. In this regard, experience has shown that, in general, it is possible to achieve stope wall stability with minimal dilution by creating openings having high vertical and short horizontal dimensions. An example would be a stable, vertical raisebore that is extended laterally, until it becomes unstable. Stability is also achieved by forming openings having long horizontal and short vertical dimensions. An example would be a long, stable tunnel, whose height is increased until it becomes unstable. Square-shaped stopes are the most ineffective in terms of potentially stable volumes (Figure 1.8). The shape of the conceptual transition curve in Figure 1.8 is hyperbolic and indicates that for multiple lift sublevel open stopes (excavations with walls that have high vertical and short horizontal dimensions) the critical spans are either the exposed horizontal lengths or the stope widths. Length and width, that is, dimensions in plan view, are the critical stope dimensions as they also control the dimensions of the stope crowns. Bench stopes are excavations where the longest dimension is the strike length and the critical spans are usually the exposed heights, as the orebody width is usually narrow. Figure 1.9 shows an example of hangingwall performance for single- and double-lift stopes extracted in a similar geotechnical domain.

9

Introduction

60

Stope up-dip span (m)

50

40

30 Depth of failure (m) 0–1 m 1–2 m

20

2–3 m

10

0

3–4 m >4 m 0

10

20 30 40 Stope strike length (m)

50

60

FIGURE 1.9 Stope performance—steeply dipping tabular rock mass, Mount Marion Mine.

The case study data show that for the single-lift stopes, stope performance is not controlled by geometry, as the depth of failure is not correlated with stope dimensions. However, as the stope height is increased, the depth of failure increases with an increase in stope strike length. An immediate conclusion is that a reduction in stope size may not necessarily result in better stope performance. Another case study is shown in Figure 1.10, in which stope performance is clearly related to stope geometry. 1.4.2  Rock Mass Strength It is generally accepted that the behavior of the stope walls is largely controlled by the strength of the rock mass surrounding the stope. This rock mass strength depends upon the geometrical nature and strength of the geological discontinuities as well as the physical properties of the intact rock bridges. Single or combinations of major discontinuities (usually continuous on the scale of a stoping block) such as faults, shears, and dykes usually have very low shear strengths and, if oriented unfavorably, provide failure surfaces when exposed by the stope walls (Figure 1.11). Such geological discontinuities largely control overbreak and stability around exposed stope walls. This is particularly the case for those discontinuities having platy and water-susceptible mineral infill such as talc, chlorite, and sericite.

10

Geotechnical Design for Sublevel Open Stoping

100 5.6

Stope up-dip span (m)

80

5.3 3.4

60

3.4

40 2.5

20

0

0

20

40

2.8 2.5 3.6

5.2

2.8

Stope depth of failure (m) 60

80

100

Stope strike length (m) FIGURE 1.10 Stope performance—shallowly dipping tabular rock mass, Davyhurst Mine. (From Parker, B. 2004. Geotechnical study of shallow dipping orebodies—Lights of Israel Underground Gold Mine. BEng thesis, Mining Engineering Department, WA School of Mines, Curtin University of Technology, Perth, Western Australia, Australia.)

FIGURE 1.11 Stope hangingwall stability controlled by large-scale faulting.

11

M

80

°

°

52

55

M

5500 N N5 4

4500 N

5000 N

Introduction

44

J 54

T 45

S4

8

80

°

O

50

52

53

80

T

°

75°

1100 Cu orebody

70

°

°

75 °

FIGURE 1.12 Plan view of major structures affecting sublevel stoping—1100 Orebody, Mount Isa Mines. (From Alexander, E.G. and Fabjanczyk, M.W., Extraction design using open stopes for pillar recovery in the 1100 ore body at Mount Isa, in D.R. Stewart, ed., Design & Operation of Caving & Sublevel Stoping Mines, SME of AIME, New York, 1981, pp. 437–458.)

In some cases, instability can be linked to activities in concurrent voids along the strikes or dips of major geological features such as fault zones (Logan et al., 1993). Ideally, the location of large-scale geological discontinuities is well defined and most open stoping mines have a threedimensional model of the local fault/shear network (Figure 1.12). These features can also be seismically active, further increasing falloff at the excavation boundaries, especially in narrow orebodies. When large-scale structures are exposed, stope wall overbreak is usually very difficult to control, regardless of the blasting practices used, and can only be minimized by stope sequencing. Stope wall behavior is also a function of the number, size, frequency, and orientation of the minor-scale geological discontinuities. Such discontinuity networks usually control the nature and amount of overbreak at the stope boundaries. Rock mass characterization techniques can be used to estimate the shapes and sizes of blocks likely to be exposed at the final stope walls. The geometrical discontinuity set characteristics (size, frequency, orientation, persistence, surface strength, etc.) relative to the stope walls largely control the amount of dilution experienced at those walls (Figure 1.13). Individual joints have a limited size and they may either terminate in intact rock, forming an intact rock bridge, or against another structure within a discontinuity network. These intact rock bridges are significantly stronger than the naturally occurring discontinuities and provide a higher resistance to failure within a rock mass.

12

Geotechnical Design for Sublevel Open Stoping

FIGURE 1.13 Example of stope large-scale footwall and hangingwall falloff. (From Villaescusa, E., A review of sublevel stoping, in G. Chitombo, ed., Proceedings of the MassMin 2000, Brisbane, Queensland, Australia, October 29 to November 2, 2000, pp. 577–590, AusIMM, Melbourne, Victoria, Australia. With permission.)

1.4.3  Induced Stresses Extraction within a stoping block can generate large concentrations of stress around the excavation boundaries. If the local (induced) stresses increase beyond the strength of a rock mass, then changes in the rock mass quality around a stope will occur, and localized failures are likely to be experienced either following discontinuity surfaces or directly through intact rock. Where movement through discontinuities occurs, stresses are relieved. This may in turn lead to overbreak, dilution, or large failures (Figure 1.14). Rock mass quality changes around the boundaries of a stope result from a combination of stress redistributions, near field blast damage, and the effects of the excavation itself. In cases where stope wall failures do not occur due to stress concentration, vibration and gases from nearby blasting may damage the intact rock bridges, which define and interlock the in situ rock blocks, causing overbreak or dilution at the stope boundaries. Furthermore, the dynamic behavior of an unsupported wall is directly related to the amount of intact rock available within the rock mass. The less intact rock available, the more cracking, slabbing, and visible stope wall displacement will result from the blasting process.

13

Introduction

Roc k fa

ll

FIGURE 1.14 Stress-related bench stope brow failure following ring blasting.

In addition, stope wall failures due to stress changes of a tensional nature can also be experienced (Bywater et al., 1983). Stope extraction in a destressed orebody may lead to normal stresses of very low magnitude across some of the exposed walls. Buckling-type failures may occur, depending upon the frequency of discontinuities parallel to a stope wall, the size and frequency of any cross discontinuities, and the size and shape of the exposed spans (Figure 1.15). 1.4.4  Ground Support Reinforcement by cablebolting provided at selected locations, usually constrained by the distance between drilling sublevels, can be used to reduce the deformations experienced at the stope boundaries (crowns, walls, and rib pillars). Stope walls are pre-reinforced prior to any stope firings and, in most cases, cablebolts are installed from rings drilled within the stope access drives. Thus, stope wall reinforcement tends to be localized in continuous bands that are separated by the distance between the sublevel intervals. The function of such an arrangement is to divide the stope walls into a number of stable stope wall spans as well as arresting up-dip hangingwall failures (Figure 1.16). Support from fill can also be used to minimize the deformations experienced by the stope walls while providing a restraint to any adjacent rock masses. In general, cemented fill is needed to recover ore from secondary stopes where stable fill exposures are required to minimize dilution. Cemented fill is essential in chequerboard extraction patterns within

14

Geotechnical Design for Sublevel Open Stoping

FIGURE 1.15 A large-scale, structurally controlled, stope hangingwall failure. (From Villaescusa, E., A review of sublevel stoping, in G. Chitombo, ed., Proceedings of the MassMin 2000, Brisbane, Queensland, Australia, October 29 to November 2, 2000, pp. 577–590, AusIMM, Melbourne, Victoria, Australia. With permission.)

Cablebolt reinforcement Hangingwall failure

Cablebolt reinforcement Hangingwall failure

FIGURE 1.16 A large stope hangingwall failure arrested by a row of cablebolts installed prior to stope firings. (From Villaescusa, E., A review of sublevel stoping, in G. Chitombo, ed., Proceedings of the MassMin 2000, Brisbane, Queensland, Australia, October 29 to November 2, 2000, pp. 577–590, AusIMM, Melbourne, Victoria, Australia. With permission.)

15

Introduction

Filling Bench limit

Production blasting

Critical strike length Mucking

Ore

Fill support Previous bench filled

FIGURE 1.17 Continuous extraction and filling operations in bench stoping.

massive orebodies (Bloss, 1992), while uncemented fill is normally used in conjunction with bench stoping operations (Villaescusa and Kuganathan, 1998). An example of a bench stoping extraction strategy linked to fill is shown in Figure 1.17. Here, the exposed wall length is usually limited to a critical value, defined by the distance between the fill and the advancing bench brow (Villaescusa et al., 1994). 1.4.5  Blast Damage Blast damage to a blasted rock mass refers to any strength deterioration of the remaining rock due to the presence of blast-induced cracks and to the opening, shearing, and extension of a preexisting or newly generated planes of weakness (Figure 1.18). It is generally accepted that the damage is caused by expanding gases through the geological discontinuities and the vibrations experienced from the blasting process. However, it is not easy to establish the approximate contribution to damage caused by the expanding gases, as it is difficult to measure their path within a rock mass discontinuity network. Nevertheless, significant backbreak may be regularly observed when the explosive gases are well confined within a volume of rock, and in some cases the gases can travel well beyond the location of the explosive charges. Damage by the shock energy from an explosive charge close to a blast can be related to the level of vibrations measured around the blasted volume. Repetitive blastings also impose a dynamic loading to the exposed stope walls away from a blasted volume, and may trigger structurally controlled falloff and ultimately overbreak. Conventional blast monitoring and simple geophysical techniques can be used to measure the effects of blasting in the near field. Vibrations and frequency levels from the shock wave can be measured reasonably accurately (Fleetwood, 2010). These data can be related to damage provided the contribution (to the total damage) from the shock energy can be estimated. Vibration and frequency levels at the mid-spans of

16

Burden

Open stope void

Hangingwall

Geotechnical Design for Sublevel Open Stoping

Stope brow

Blasthole

FIGURE 1.18 Structurally controlled damage around a hole in an open stope brow. (From Villaescusa, E., A review of sublevel stoping, in G. Chitombo, ed., Proceedings of the MassMin 2000, Brisbane, Queensland, Australia, October 29 to November 2, 2000, pp. 577–590, AusIMM, Melbourne, Victoria, Australia. With permission.)

instrumented stope walls can be used to characterize the dynamic response to blasting at the stope boundaries (Villaescusa and Neindorf, 2000). 1.4.6  Drill Drive Layout Additional factors such as poorly located or preexisting drives, which undercut the stope walls, also contribute to dilution or falloff at the stope boundaries. In general, the number and location of drilling drifts in open stoping are usually functions of the width of the orebody. In wide orebodies, hangingwall and footwall drill drives are used to provide cablebolt reinforcement and to minimize the impact of blasting at the stope boundaries (Figure 1.19). In such cases, drilling and blasting can be carried out in a plane parallel to the final stope walls or to any exposed backfill masses. Suitable values of standoff distance for the perimeter holes parallel to a stope boundary can be determined depending upon the rock type and the hole size being used (Villaescusa et al., 1994). Excessive wall damage, dilution, and ore loss may be experienced in cases where stoping requires drilling holes at an angle to a planned fill exposure or a stope boundary. Furthermore, hole deviation at the toes may create an uneven stope surface, thereby preventing effective rilling of the broken

Introduction

17

FIGURE 1.19 Twin drill drives allowing drilling parallel to the stope boundaries. (Courtesy of Mount Isa Mines, Mount Isa, Queensland, Australia.)

material to the stope drawpoints. In addition, hole deviation may cause excessive confinement at the hole toes, thus causing breakage beyond the orebody boundaries.

1.5  Scope and Contents of This Book Sublevel open stoping—including variants, such as bench stoping—is one of the most widely used mining methods in underground mining. Improvements in technology over the last 30 years or so have seen increases in sublevel spacing due to advances in the drilling of longer and accurate production holes, as well as advances in explosive types, charges, and initiation systems. Improvements in slot rising either through vertical crater retreat, inverse drop raise or raise boring have also been experienced. Increases in sublevel spacing have meant larger unsupported stope walls that must stand without collapsing. Consequently, an understanding of rock mass characterization is required to minimize dilution and increase recovery. Methodologies to design optimum open spans, pillars, rock reinforcement, and fill are required. Furthermore, in the same period, a greater understanding has developed regarding the sequencing of stoping blocks to minimize in situ stress concentrations. In the future, sublevel stoping is likely to be practiced at ever-­increasing depths (Thomson and Villaescusa, 2011) and a better understanding of all the variables required to optimize the method is required.

18

Geotechnical Design for Sublevel Open Stoping

This book will cover the topic in nine chapters, as follows:

1. Introduction 2. Sublevel Stoping Geometry 3. Planning and Design 4. Rock Mass Characterization 5. Span and Pillar Design 6. Drilling and Blasting 7. Rock Reinforcement and Support 8. Mine Fill 9. Dilution Control

The chapter topics are presented according to the conventional sublevel stoping process used by most mining houses, in which a sublevel stoping geometry is chosen for a particular mining method, equipment availability, and work force experience. Planning of access infrastructure and overall extraction sequences takes into account rock mass characterization information, which is first collected from the orebody delineation process. Detailed planning of stope span and pillars is followed by access development, where production drilling and blasting take place. Ground support becomes an important aspect to provide safe personnel and equipment access to a limited number of areas where open stoping activities take place. Following extraction, a number of strategies are available to fill resulting open stope voids, in which a reconciliation of dilution control and ore loss is critical to achieve the most economical extraction of ore. The book has been written primarily for fourth-year undergraduate students who are not yet familiar with the mining method. The book presents the state of the art and also results from the applied research at the Western Australian School of Mines (WASM) and, hence, the book could also be used for postgraduate student research. Furthermore, some mining practitioners and junior consulting engineers may find the book useful.

2 Sublevel Stoping Geometry

2.1 Introduction In sublevel stoping, ore is broken by drilling and blasting. Stope access is achieved by mining drilling and extraction drives, which can be accessed either transversally or longitudinally with respect to the orebody strike. The first stage is to create a slot between the vertical horizons defining the planned stope. This is achieved by enlarging a suitably located raise or longhole winze (LHW). The slot is created as an expansion void into which the remainder of the stope is formed by the sequential blasting of production holes. In most cases, the production holes are drilled in rings parallel to the orebody dip between the drilling drives. Mining proceeds through the sequential firing of production rings into the advancing void with the broken ore being recovered from a specific extraction horizon (Figure 2.1). The following section describes the stoping geometries required to achieve production from sublevel open stoping.

2.2  Stoping Geometries 2.2.1  Cutoff Slot Sublevel open stopes are created by the sequential blasting of production rings into an initial expansion slot, called the cutoff slot. This initial opening is used to create sufficient void for the remaining portion of the stope to break into (Figure 2.2). The cutoff slot is usually located on a side or in the center of a stope either transversally (across) or longitudinally with respect to the strike of the orebody. An important point relates to whether the cutoff blasting will expose a critical stope wall, such as a hangingwall or a fill mass, very early in a stope blasting sequence. The cutoff slot raises are blasted upward from sublevel to sublevel in order to expose the full stope height. At each level, the expansion slots are formed by sequentially blasting parallel holes into an LHW or a raise-bored hole. The slot must be expanded 19

20

Geotechnical Design for Sublevel Open Stoping

FIGURE 2.1 Remote-controlled production mucking in sublevel open stoping. (Courtesy of Mount Isa Mines, Mount Isa, Queensland, Australia.)

Cutoff slot Drill drives

tion duc Pro l rings dril

Cutoff slot

h Troug cut under Dr

Dr

aw

aw

poi

nts

po

int

s

FIGURE 2.2 A three-dimensional view of a cutoff slot. (Courtesy of WMC Resources, Kalgoorlie, Western Australia, Australia.)

21

Sublevel Stoping Geometry

to the full width of the plane defined by the production holes that will be subsequently blasted into this initial opening. High powder factors are normally used during slot blasting in order to ensure breakage and thus have a free face and a void available into which the remainder of the stope is to be blasted. The choice of slot location depends upon rock mass conditions, stope access, and the extraction sequence chosen. In a steeply dipping orebody, where the critical stope boundary is usually an inclined hangingwall, transversally oriented slots are used to ensure sequential hangingwall exposure by the production rings. In large, massive orebodies, the choice of slot orientation is also controlled by factors such as fill exposures, stress regime, and preestablished access (Bloss and Morland, 1995). In general, a slot must be designed so that failure within the main or production rings is minimized. In highly stressed pillars, a slot can be oriented normally to the major principal stress to shadow the main production holes. This is likely to minimize hole squeezing or dislocation due to stress-related damage. In cases where a stope access can be redesigned, the slot should be placed normal to any large-scale geological features likely to fail and damage the main ring geometries (Figure 2.3). Damage to fill masses from cutoff slot blasting can be minimized by placing a cleaner ring between a cutoff and a fill boundary (Figure 2.4). The rock mass adjacent to a fill mass is usually preconditioned by stress redistributions and is likely to fail following a cleaner ring blasting. In order to minimize hangingwall failures, cutoff slots are oriented transversally to the orebody strike. This allows the hangingwall plane to be sequentially exposed within a predetermined stable range. In secondary stope extractions, where longitudinal cutoff slots may be located parallel °

P4

1

5°

Fa

Potential falloff within rings

ul

t

°

P4

1

5°

6

65 Fa

ul

t Cutoff slot

/W

/W

1F

1F

P4

P4

t-off slotslot Cutoff

6

65

a spl y

ay

spl

(a)

(b)

FIGURE 2.3 Exposure of weak geological features by a cutoff slot. (a) Poor (preliminary) design and (b) improved (actual) design. (Courtesy of Mount Isa Mines, Mount Isa, Queensland, Australia.)

22

Geotechnical Design for Sublevel Open Stoping

Cleaner Production ring rings Blasting sequence 1. Holes near raise

Extracted (filled)

2. Mid cutoff

Cutoff slot

3. Complete cutoff + cleaner ring Plan view

FIGURE 2.4 Cleaner ring geometry to minimize fill damage from blasting. (From Villaescusa, E., A review of sublevel stoping, in G. Chitombo, ed., Proceedings of the MassMin 2000, Brisbane, Queensland, Australia, October 29 to November 2, 2000, pp. 577–590, AusIMM, Melbourne, Victoria, Australia. With permission.)

(and adjacent) to a stope hangingwall, the expansion slot exposes a full hangingwall plane early in a blasting sequence. This usually limits the size of exposures that can be safely excavated, as this critical wall of the stope may fail when subjected to repetitive dynamic loading by the rest of the stope firings as shown conceptually in Figure 2.5. In addition, when the stopes are accessed centrally, drill design requires that the holes toe into any adjacent fill masses, thereby increasing the likelihood of fill dilution. 2.2.2  Production Rings A design stope shape is achieved by sequentially blasting rings of blastholes into the opening created by the initial expansion or cutoff slot. Stopes are usually sequentially sliced up, from sublevel to sublevel, firing rings toward the open cutoff slot. The production rings are sequentially blasted, attempting to minimize undercutting of the internal solid portion of a stope. An approximately straight face is kept along the entire stope height by firing a similar number of rings at each sublevel. The firing sequence advances upward as shown in Figure 2.6. Maintaining a straight retreating face minimizes the creation of large brows or corners, which can be highly stressed or intersect large-scale structures, thereby contributing to stope falloff. This, in turn, can severely affect productivity during the subsequent production mucking operations. 2.2.3  Diaphragm Rings Diaphragm rings consist of rings drilled parallel to a fill exposure. The purposes of a diaphragm ring are to prevent fill failure from a known weak cemented fill mass, to contain uncemented fill in adjacent stopes, and to prevent fill failure from exposures of greater dimension than is considered

23

Sublevel Stoping Geometry

Filled

Filled Slot

Stope hangingwall plane Repetitive production ring blasting FIGURE 2.5 Dynamic loading of a fully exposed hangingwall plane. (From Villaescusa, E., A review of sublevel stoping, in G. Chitombo, ed., Proceedings of the MassMin 2000, Brisbane, Queensland, Australia, October 29 to November 2, 2000, pp. 577–590, AusIMM, Melbourne, Victoria, Australia. With permission.)

stable. Experience has shown that although parts of a diaphragm against fill do fall off, this rarely results in excessive fill dilution, as the fill mass remains comparatively undisturbed, compared to when blasting takes place next to the fill (Figure 2.7). A diaphragm is not capable of load-bearing capacity and so is likely to deform considerably. However, when a large portion of the diaphragm remains intact, this enables clean stope extraction until the diaphragm is either fired or the stope is completed. 2.2.4  Trough Undercut The lower portion of a stope is shaped using trough undercut (TUC) rings in order to facilitate the draw of fragmented ore to and from the stope drawpoints. A TUC ring consists of parallel upholes, drilled inclined toward the cutoff slot. Usually the toes of the TUC ring interlock with the toes of the main ring downholes from the sublevel above (see Figure 2.8). Drilling and blasting of the TUCs is usually carried out using relatively small diameter holes (70–89 mm) compared to production holes. An improved explosive distribution likely to minimize rock mass damage around the stope drawpoints is achieved by using such small diameter holes. A disadvantage is

24

Geotechnical Design for Sublevel Open Stoping

Cutoff slot

Cutoff slot

Drill and blast access

Drill and blast access

Stope undercut fall-off potential

Broken ore Mucking

Mucking

Mucking

FIGURE 2.6 A longitudinal section view showing two production blasting strategies. (From Villaescusa, E., A review of sublevel stoping, in G. Chitombo, ed., Proceedings of the MassMin 2000, Brisbane, Queensland, Australia, October 29 to November 2, 2000, pp. 577–590, AusIMM, Melbourne, Victoria, Australia. With permission.)

140 mm blasthole

Fill mass

·

· · · · · · · · · ·

Rock diaphragm

Remainder of stope · · extracted · · · · · · · · ·

2m from fill

Fillmass

3 m burden on diaphragm ring FIGURE 2.7 Idealized sketch and photo showing a stope diaphragm ring. (From Villaescusa, E., A review of sublevel stoping, in G. Chitombo, ed., Proceedings of the MassMin 2000, Brisbane, Queensland, Australia, October 29 to November 2, 2000, pp. 577–590, AusIMM, Melbourne, Victoria, Australia. With permission.)

25

Sublevel Stoping Geometry

14

13

10

7

4

3

14

12

9

6

2

11

8

5

1

Longitudinal view FIGURE 2.8 Firing sequence of a TUC with production rings in an open stope. (From Villaescusa, E., A review of sublevel stoping, in G. Chitombo, ed., Proceedings of the MassMin 2000, Brisbane, Queensland, Australia, October 29 to November 2, 2000, pp. 577–590, AusIMM, Melbourne, Victoria, Australia. With permission.)

the limited drilling length achieved, and the inability to match the burden drilled for the production ring holes immediately above. Because the TUC rings are drilled with a different burden to the production rings, the lower portion of a stope is usually blasted ahead of the main rings. This leads to a moderate undercutting of the main rings, which can lead to falloff, especially in cases where large geological discontinuities are present or in regions of high stress redistribution.

26

Geotechnical Design for Sublevel Open Stoping

Drawpoints

S

P Orebody

S hangingw

P

all

FIGURE 2.9 A fixed, transverse drawpoint geometry in sublevel stoping. P, primary stope; S, secondary stope. (From Villaescusa, E., A review of sublevel stoping, in G. Chitombo, ed., Proceedings of the MassMin 2000, Brisbane, Queensland, Australia, October 29 to November 2, 2000, pp. 577–590, AusIMM, Melbourne, Victoria, Australia. With permission.)

2.2.5 Drawpoints Production mucking can be carried out longitudinally or transversely across the strike of an orebody. Transverse mucking requires the introduction of fixed and specialized drawpoint geometries that may be located outside an orebody boundary (Figure 2.9). The factors considered during drawpoint design include size of equipment, tramming distance from access drives, and gradient and orientation with respect to a stope boundary. The drawpoint dimensions must be sufficient to suit the equipment, but kept as small as possible to minimize instability. Drawpoint access should be straight and restricted to 15–20 m from a stope access drive to the stope brow. This will ensure that auxiliary ventilation will not be required while mucking, and also that the rear of the mucking unit is inside the drawpoint. Drawpoint spacing is determined by ground conditions and stope geometry. In most cases, the minimum spacing used is 10–15 m between center lines.

2.3  Multiple-Lift Open Stoping Multiple-lift stopes extend vertically over a number of sublevel intervals, in some cases exceeding hundreds of meters in vertical extension. The method requires sequential blasting of the production rings into an initial

27

Sublevel Stoping Geometry

vertical opening formed by a cutoff slot. Ore breakage is achieved by rings of parallel or fanned blastholes, depending upon the type of drilling access used. TUCs are developed at the base of the stopes in order to direct the broken ore into the drawpoints for extraction. Cablebolt reinforcement of hangingwall and stope crowns can be provided from suitably located drilling drives. The number of drawpoints is usually a function of the stope size, but in most cases at least two drawpoints are designed. Because the drawpoint location is fixed, permanent reinforcement can be achieved at minimum cost per unit of ore extracted. Access to the stope on each of the other sublevel locations is required for drilling, blasting, and filling purposes (Figure 2.10). Usually, a single crosscut access is required on each sublevel, significantly decreasing development in waste. In general, multiple-lift stopes minimize back cablebolting within the intermediate sublevels because a permanent back (full area) is only exposed at the actual crown of a stope. Cablebolting coverage at a stope crown is a function of the degree of development within the top sublevel. In addition, the requirements for permanent reinforcement within any intermediate

1. Development

2. Cablebolt drilling

3. Production drilling

4. Production blasting

5. Production mucking FIGURE 2.10 Sequence of mining activities within a multiple-lift sublevel open stope at the Kanowna Belle Mine.

28

Geotechnical Design for Sublevel Open Stoping

sublevel are minimized by the fact that all the back exposures within the drill drives are consumed by the stoping process itself. Conventional multiple sublevel stoping requires the sequential exposure of high vertical, short horizontal stope walls likely to remain stable and provide undiluted ore. The strike lengths exposed during the initial stope extraction are unlikely to exceed the critical stable stope spans. As the excavations are enlarged and several rings are sequentially blasted into the void formed by the cutoff and the initial production rings, confining stresses are reduced, excess strain energy is induced, and displacement of the stope walls is experienced. Depending on the structural nature of the exposed walls, the rock may tend to displace following a sheetlike behavior, in which a group of layers move together (in bedded rock), or the movement may be isolated to individual blocks that partially rotate and slide against each other. 2.3.1  Tabular Orebodies The layout for multiple-lift sublevel stopes in tabular orebodies is usually associated with the use of long blastholes drilled from drives parallel to the orebody strike. Depending upon the orebody width, these drill drives may be either of full orebody width or located at the boundaries of the orebodies. In such orebodies, the stope boundaries are usually well defined by the orebody itself. Crown, hangingwall, footwall, endwalls, and a drawpoint can be defined for each stope. The stability of stope crowns and hangingwalls is usually the most critical factor in the stope design and related extraction sequences. A conventional design usually consists of multiple drilling sublevels with a single mucking horizon at the bottom of the stope as shown in Figure 2.11. One of the advantages of this design is that drilling and blasting can be done in a plane parallel to the final stope walls. Hangingwall and footwall drill drives are used to minimize the impact of blasting at the stope boundaries, greatly decreasing the likelihood of dilution due to blast damage. In addition, the method reduces stope development in waste, given that, except for the mucking horizon, a single stope drilling access is actually required at each sublevel location. In cases where sublevel stoping is used to extract large but tabular orebodies having a moderately dipping hangingwall, the extraction can be divided into a number of primary, secondary, and sometimes tertiary stopes, which can be extracted in a checkerboard sequence. In order to optimize stope stability, the stope walls are designed vertically, except for the hangingwall as shown in Figure 2.12. Drilling drives parallel to the hangingwall can be used to provide cablebolt reinforcement, and facilitate drilling and blasting parallel to the hangingwall planes. Stope crown stability can be optimized with the implementation of a floating sublevel to optimize cablebolt reinforcement. The use of conventional drawpoint geometries increases productivity.

29

Sublevel Stoping Geometry

Production H/W drive Cleaner ring rings Extracted (filled)

Trough undercut

Cablebolted area

Cablebolted area Drawpoint 1

Cutoff slot

Cutoff slot

Hangingwall

Drawpoint 2

Cross cut access

F/W drive F/W access drive

F/W access drive

(a)

Extracted (filled)

(b) Crown reinforcement

Cutoff slot

H/W

rein

forc

eme n

t

Crown reinforcement

Production rings

Cutoff slot

F/W drive

Trough undercut

(c)

(d)

FIGURE 2.11 Sublevel stoping in a steeply dipping tabular orebody. (a) Plan view—mucking horizon, (b) plan view—intermediate level, (c) cross section view—production rings, and (d) long section view. (From Villaescusa, E., A review of sublevel stoping, in G. Chitombo, ed., Proceedings of the MassMin 2000, Brisbane, Queensland, Australia, October 29 to November 2, 2000, pp. 577–590, AusIMM, Melbourne, Victoria, Australia. With permission.)

2.3.2  Massive Orebodies Open stoping in large, massive orebodies consists of a mining sequence that requires several stages of stoping in conjunction with the application of delayed fill methods to enable pillar recovery. Usually, a number of stopes are designed between the orebody boundaries. In such cases, stoping comprises a number of stages that includes primary, secondary, and tertiary stopes that are usually extracted using a checkerboard sequence (Alexander and Fabjanczyk, 1981). The number of fill exposures ranges from none (in a primary stope) up to three exposures in the late stages of stoping (Grant and DeKruijff, 2000).

30

Geotechnical Design for Sublevel Open Stoping

Ha

ng

ing

wa

ll r e

inf o

rce

me

nt

Floating sublevel for crown reinforcement

Footwall access drive

Drawpoint

FIGURE 2.12 Stope design for a large tabular orebody. (Courtesy of Mount Isa Mines, Mount Isa, Queensland, Australia.)

Large vertical dimensions can be designed with the height of the stopes usually constrained by the orebody thickness or by the stability of any exposed fill masses required for secondary and tertiary stope extraction. Stope dimensions in a plan view are usually constrained by stope crown instability. The broken ore is extracted in the bottom part of the stope (Figure 2.13). In cases where the ground conditions are favorable, stope dimensions can be very large in plan, with full orebody height extraction being achieved in a single stope (Bloss, 1996). Drilling and blasting is carried out from a series of sublevel locations ranging from 40 to 60 m apart. Blastholes are mainly drilled downward, with some short upholes drilled within the TUCs and sometimes at the stope crown when a top access is not available. Following pillar extraction (secondary and tertiary stopes), a number of fill exposures are created depending upon the location of the stope in the mining sequence. Early on in the life of a massive orebody, primary stopes usually account for a significant part of the production. As orebody extraction increases, the shift to pillar mining as the primary method of extraction

31

Sublevel Stoping Geometry

7

1

Extraction sequence 5

4

2 8

3 1

7

6

(a)

(b)

FIGURE 2.13 Multiple-lift stoping in a massive orebody. (a) Plan view and (b) three-dimensional view. (From Villaescusa, E., A review of sublevel stoping, in G. Chitombo, ed., Proceedings of the MassMin 2000, Brisbane, Queensland, Australia, October 29 to November 2, 2000, pp. 577–590, AusIMM, Melbourne, Victoria, Australia. With permission.)

becomes evident. In such cases, the stability of the fill exposures is of primary importance in achieving the target production figures (Bloss and Morland, 1995). In cases where the upper orebody boundary does not coincide with the predetermined location of the upper sublevel interval, drilling into or through the orebody crown may be required. If the top of the orebody is above the highest sublevel interval location, upholes may be drilled into the stope crown in order to define a designed stope shape. In cases where the highest sublevel is located above the orebody boundary, downholes may be drilled through the orebody crown, with the lowest portion of the holes blasted to define a stope shape. In both cases, the stope crown remains unsupported, and a preferred alternative is to develop a “floating” sublevel through the top of the stopes to facilitate deep cablebolt reinforcement and drilling of holes parallel to the designed stope crown (Figure 2.14).

2.4  Single-Lift Stoping A single-lift design is the most basic arrangement for sublevel open stope extraction. The stope shape and size is constrained by two sublevels: the extraction or undercut horizon, and the drilling or overcut horizon.

32

Geotechnical Design for Sublevel Open Stoping

le rp Pu fa ul t

Gr f ay lt au

24A 2200 RL

S613 lt au df Re

25A 2150 RL

26D 26B FIGURE 2.14 Drilling and blasting strategies for a stope crown. (Courtesy of Mount Isa Mines, Mount Isa, Queensland, Australia.)

Access to the stopes is via crosscuts off a permanent access drive parallel to the orebody. Effectively, this method requires a “moving” drawpoint system, as the stoping extraction progresses upward. Following the filling of a stope void, a previous drilling horizon becomes the next extraction level (Figure 2.15). Development

Cablebolting/drilling

Blasting

Filling

Drawpoints

Drawpoints

Drawpoints

FIGURE 2.15 Three-dimensional view of single-lift sublevel stoping. (From Potvin, Y. et al., CIM Bull., 82(926), 53, 1989.)

33

Sublevel Stoping Geometry

West

East

Stoping block 3

Block 3–4 sill pillar

Cave zone

Stoping block 4

Cave zone

FIGURE 2.16 Longitudinal section view of the Williams Mine B zone. (From Bawden, W. F. et al., Lessons in control of mine costs from instrumented cable bolt support. In J. Girard, M. Leibman, C. Breeds, and T. Doe (eds.), Proceedings of the Fourth North American Rock Mechanics Symposium, Seattle, WA, 31 July–3 August, A.A. Balkema, Rotterdam, the Netherlands, 2000, pp. 633–642.)

In order to optimize mucking productivity, up to two access crosscuts per stope may be required at each sublevel interval. This actually increases the overall access development in waste to actual stoping ratio. The method requires very good control of the stope back and brow stability, especially in a highly stressed environment. Stress redistribution due to the stoping sequence itself can create significant back failures, especially if shallow dipping discontinuities are present within a rock mass. Figure 2.16 shows a typical extraction configuration using single-lift stopes at the Williams Mine in Canada, where several major rockfall occurrences within the sill pillar have been reported by Bawden et al. (2000). The rockfalls delayed the mining of approximately 1 million tons containing some 300,000 oz, seriously affecting production from the mine. Extended backs, pillars, and highly stressed brows are likely to be formed somewhere within the stoping sequence, and full cablebolting coverage of the stope backs is required to minimize the potential failures at each sublevel location. Full cablebolting coverage requires stripping the orebody access to the full stope width, thereby minimizing the sizes of stopes that can be developed safely. As a result, single-lift stopes tend to be relatively small openings compared with multiple-lift stopes.

34

Geotechnical Design for Sublevel Open Stoping

Pendant pillar

Filled

Filled

Pendant pillar

FIGURE 2.17 Idealized stoping sequence for single stopes on a 1-4-7 extraction sequence. (From Potvin, Y. et al., CIM Bull., 82(926), 53, 1989.)

Primary development requires the extension of the access crosscut to a proposed hangingwall location, where both the drill and the extraction sublevels are completely silled out to allow the installation of cablebolt reinforcement. In addition, the drilling of parallel blastholes is also facilitated with full stope undercut and overcut geometries. Drilling of parallel holes is the preferred way in vertical retreat stoping, which is linked to single-lift stoping. The method requires a significant amount of remote mucking due to the flat-bottom nature of the single-lift stope geometries, thereby increasing the overall mining cost compared to a conventional TUC drawpoint geometry. In wide orebodies, a number of stopes may be designed across the strike in a given area, and in all cases, adjacent primary stopes are extracted to a level above that of a secondary stope. This type of sequence creates what is called a pendant pillar. A pendant pillar is a solid piece of ground that has many degrees of freedom for movement, as most stopes around it have been extracted (Figure 2.17). Large pillar failures may be experienced in such stoping geometries (Milne and Gendron, 1990). 2.4.1  Conventional Vertical Crater Retreat Stoping Vertical crater retreat (VCR) is a single-lift stoping method where the stope’s shape is defined by a lower (undercut) and upper (overcut) horizon

35

Sublevel Stoping Geometry

Full overcut

Filled

FIGURE 2.18 VCR mining within a single-lift stope. (From Villaescusa, E., A review of sublevel stoping, in G. Chitombo, ed., Proceedings of the MassMin 2000, Brisbane, Queensland, Australia, October 29 to November 2, 2000, pp. 577–590, AusIMM, Melbourne, Victoria, Australia. With permission.)

(Trotter, 1991). Large-diameter holes are drilled in order to minimize deviation, and the holes are charged from the overcut and blasted by means of horizontal slices of ore progressing from the bottom level to the top level (Figure 2.18). The separation between the undercut and overcut is a function of stope wall stability, the nature of the orebody, and drilling accuracy. Following blasting, only a slight amount of broken ore is mucked, so that enough room is available for a subsequent blast to break into. This keeps the stope full of broken rock, thereby providing passive support to the exposed stope walls until blasting to the stope overcut is complete. Once blasting is completed and all the ore within the stope is mucked, the undercut accesses are closed off and the stope is filled. As mining progresses upward, the stope overcut becomes the next mucking horizon in the sequence. The method has a number of perceived advantages including the requirement for few large-diameter blastholes, likely to reduce the overall in-thestope drilling. Large holes enable a larger sublevel interval to be used, thus reducing the overall sublevel development cost. The cost of raising and slashing to create a slot is eliminated, and all the drilling and loading operations are carried out from the overcut, thereby increasing safety.

36

Geotechnical Design for Sublevel Open Stoping

The disadvantage of this method is the potential for blast damage from crater blasting at the stope boundaries (Platford et al., 1989). Small-diameter holes cannot be used due to hole closure caused by ground movement following the individual stope blasts (Hills and Gearing, 1993). In addition, this method may be susceptible to poor fragmentation (falloff) from the unsupported areas defined by blasting, especially if an uneven back is formed and high stresses are subsequently redistributed upward. Blast damage from cratering is even more detrimental when shallowly dipping geological discontinuities are present within a rock mass. 2.4.2  Modified Vertical Retreat Stoping A modified vertical retreat method uses a winze or a raise-bored hole, which is located near the middle of the stope, into which a radiating pattern of blastholes is sequentially fired in horizontal lifts. The raise is used to overcome the limited free face available in a conventional vertical retreat stope. In order to facilitate the initial blasting, the method requires close spacing of the holes near the raise (Figure 2.19). All the holes in a horizontal lift are fired, and the possibility of collar damage exists when the inner holes near the raise do not perform. In addition, hole damage (closure, requiring

Parallel rings of vertical blastholes to be fired in a radiating pattern from the raise

Plan view

Raise bored hole 1.1 m diameter

Position after one ring firing Section view FIGURE 2.19 Typical blast layout for a modified vertical retreat stope in the Porgera Mine. (From Hills, P.B. and Gearing, W.G., Gold ore mining by the Porgera Joint Venture at Porgera, Papua New Guinea, in J.T. Woodcock and J.K. Hamilton, eds., Processing Australasian Mining and Metallurgy, AusIMM, Melbourne, Victoria, Australia, 1993, Chapter 12: Gold, pp. 897–902.)

37

Sublevel Stoping Geometry

redrilling) within the last lift in the stope may be continuously experienced with this method (Hills and Gearing, 1993). On the other hand, the method is considered to be relatively safe because no vertical opening is made within the stope until the last firing.

2.5  Shallow Dipping Tabular Orebodies Tabular orebodies in which the dip angle does not allow the flow of broken ore utilizing gravity can be extracted using a type of sublevel stoping called uphole retreat panel stoping (Kaesehagen and Boffey, 1998). Typically, an orebody can be divided into panels, running parallel to the strike of the orebody and defined down-dip as shown in Figure 2.20. The stopes are extracted by developing a footwall extraction drive from which drilling, blasting, and mucking operations can be carried out. The stopes are accessed from a footwall drive, with a slot established at the far end of the panels, and the stopes are progressively blasted retreating back to the access end of a panel (Figure 2.21). Cablebolt reinforcement is provided from the hangingwall drives located within the primary stopes. In addition, permanent pillars can be left within the secondary stopes to provide additional hangingwall support. Flat lying orebodies can also be extracted by individual stopes in conjunction with cablebolting drives and mine fill operations. The stopes are extracted by developing a TUC horizon in waste to allow the flow of ore Hangingwall cablebolting Production up-holes

Panel slot (drilled downhole)

Decline

Panel access drive Longitudinal view p S

p

S

S

Section view FIGURE 2.20 Typical panel stoping layout. P, primary stope; S, secondary stope. (From Kaesehagen, M.R. and Boffey, R.H., Development of the Osborne Mine—with a focus on technical and operational aspects, Proceedings of the Seventh Underground Operators’ Conference, Townsville, Queensland, Australia, June 30 to July 3, 1998, pp. 29–37, AusIMM, Melbourne, Victoria, Australia. With permission.)

38

Geotechnical Design for Sublevel Open Stoping

FIGURE 2.21 An unsupported uphole panel stope following extraction.

to the stope drawpoints. Downhole drilling is undertaken from a series of hangingwall drives, from which cablebolt reinforcement is also provided (Figure 2.22). This method results in an increased lead time in stope preparation as well as additional costs, as noneconomical material is developed. The overall stope extraction retreats up-dip and toward the access end of the drilling drives. Experience indicates that only half of the back of a previously extracted stope (down-dip) can be filled effectively. The methodology consists of extracting stopes having either single or double drilling drives, depending upon their location with respect to the orebody abutment and with respect to each other in the extraction sequence. Alternating single and double drilling drives is likely to optimize hangingwall reinforcement as the extraction progresses up-dip.

2.6  Bench Stoping Bench stoping is used to extract steeply dipping and relatively narrow (up to 12–15 m wide) veins, lenses, lodes, or any stratiform deposit extending in

39

Sublevel Stoping Geometry

18

2

12 6 15

1 3

19 10

7

2 11

Plan view

20

17 8

13

Ca

16

4

9 5

11

ble

bo

ltin

g

7 3

14 1

Extraction sequence Stope boundary Filled stope

1

Section view

FIGURE 2.22 Overall extraction sequence and cross section showing cablebolt reinforcement. (From Villaescusa, E., Extraction sequences in sublevel stoping, Proceedings of the 12th International Symposium on Mine Planning & Equipment Selection, Kalgoorlie, Western Australia, Australia, April 23–25, 2003, pp. 9–18, AusIMM, Melbourne, Victoria, Australia. With permission.)

two dimensions (along strike and down-dip). The method involves the initial mining of both a drilling and an extraction drive for the entire length and width of the orebody (Figure 2.23). A slot is created between the two horizons at one end of the orebody by enlarging a cutoff raise (or LHW) located near the footwall of the orebody. The slot created is used as an expansion void into which the remainder of the bench stope is formed by the sequential blasting of production holes. In most cases, the production holes are drilled in rings parallel to the orebody dip between the two drives (Figure 2.24). Stoping proceeds through the sequential firing of downhole (or uphole) blasthole rings into the advancing void, and ore is then remotely mucked along the orebody from the extraction horizon (Figure 2.25). The likelihood of ore dilution increases if the ore is left within the stope floor for long periods of time and wall failures may cause ore loss or damage the mucking units. The success of bench stoping relies on the stability of the exposed unsupported spans, the ability to provide support with cablebolting and fill, tight control on drilling and blasting, as well as the application of remote mucking technology (Villaescusa et al., 1994). Downhole bench stoping geometries are linked to up-dip overall extraction sequences in conjunction with fill. Uphole bench stopes are often extracted without the use of fill, and retreating topdown in conjunction with permanent, nonrecoverable pillars. In most mining operations, the bench heights are fixed during the early stages of mine development, and the extraction strategy is the only variable that can be used to optimize the economics of bench stoping. In downhole benches, the extraction is followed by filling of the void with waste, hydraulic sand fill, or aggregate to the floor of the drilling drive, which becomes the new extraction drive on the next lift up-dip. A number of extraction strategies

Ore

Auxiliary ventilation

Bench brow closed

Rockfill

Retreat direction

Crosscut

(b)

Longitudinal access

Access FW drive

Crosscut

Sto

pe

reat Ret tion c dire kfill Roc

FIGURE 2.23 Details of bench stope extraction. (a) Longitudinal view and (b) 3D view. (Courtesy of Kanowna Belle Mines, Kalgoorlie, Western Australia, Australia.)

(a)

RAR to primary exhaust

40 Geotechnical Design for Sublevel Open Stoping

Sublevel Stoping Geometry

41

0.5 m

0.8 m

17D

6865 N

78°

83°

17B

0.5 m

72°

0.8 m

67°

17.5 m 17.5 m 17.6 m 17.9 m

2m

FIGURE 2.24 A typical cross-sectional view and the results of exceptional downhole bench stoping. (Courtesy of Mount Isa Mines, Mount Isa, Queensland, Australia.)

FIGURE 2.25 Longitudinal remote mucking of broken ore in bench stoping. (Courtesy of Mount Isa Mines, Mount Isa, Queensland, Australia.)

42

Geotechnical Design for Sublevel Open Stoping

FIGURE 2.26 Longitudinal ore extraction in conjunction with fill support. (Courtesy of Mount Isa Mines, Mount Isa, Queensland, Australia.)

have been considered for downhole benching (Villaescusa and Kuganathan, 1998). The most common involves using a continuous dry fill mass (waste rock having a rill angle between 38° and 42°) that follows an advancing bench stope brow at a fixed distance (not exceeding a critical unsupported strike length) along the entire bench length (Figure 2.26). Benches can also be extracted using hydraulic fill, with the stopes extended to a maximum stable unsupported strike length, followed by fill in conjunction with brick bulkheads. Filling is followed by pillar recovery and the process is repeated along the entire bench length (Figure 2.27). Although this strategy is primarily linked to hydraulic fill, the use of cemented fill would ensure that minimal fill dilution would be experienced following pillar recovery. Cemented fill can only be justified during extraction of very high-grade orebodies. Recent applications of cemented paste fill are replacing the use of hydraulic fill, thus minimizing the need for brick bulkheads. Another strategy is to leave (planned) permanent pillars between independent (unfilled) hangingwall spans along the entire bench length. Filling is done on bench completion using either dry or hydraulic fill (Figure 2.28). In this strategy, it is critical to establish the optimum distances between the pillars in order to minimize the number of pillars required, especially in high-grade orebodies. Pillar dimensions are a function of the ground conditions, the expected stress levels, and the optimum extraction of the adjacent LHWs. In weak rock masses, the stability of unfilled spans may be affected by blasting in adjacent spans along the strike of the orebody, as the individual spans may show time-dependent behavior with related

43

Sublevel Stoping Geometry

Temporary pillar (drilled) New slot

Recovered pillar

Production blasting

Hydraulic fill

Maximum strike length (void filled)

Mucking

Ore

Bench limit

Barricades

Extracted and filled FIGURE 2.27 Hydraulic fill and pillar recovery. (From Villaescusa, E. and Kuganathan, K., Backfill for bench stoping operations, in M.L. Bloss, ed., Minefill 98, Proceedings of the Sixth International Symposium on Mining with Backfill, Brisbane, Queensland, Australia, April 14–16, 1998, pp. 179– 184, AusIMM, Melbourne, Victoria, Australia. With permission.)

Permanent pillar New slot

Production blasting

Permanent pillar

Maximum unsupported strike length (void to be filled at bench completion) Mucking

Ore Extracted and filled

Bench limit

Permanent pillar Extracted and filled

FIGURE 2.28 Nonrecoverable permanent pillars in conjunction with fill. (From Villaescusa, E. and Kuganathan, K., Backfill for bench stoping operations, in M.L. Bloss, ed., Minefill 98, Proceedings of the Sixth International Symposium on Mining with Backfill, Brisbane, Queensland, Australia, April 14–16, 1998, pp. 179–184, AusIMM, Melbourne, Victoria, Australia. With permission.)

44

Geotechnical Design for Sublevel Open Stoping

FIGURE 2.29 Unsupported spans and permanent pillars in shallow dipping bench stoping.

deformation. Figure 2.29 shows a top-down bench extraction strategy that relies on a combination of unsupported spans and permanent pillars. Bench stopes can be also extracted using a continuous and tight filling technique called Avoca. Initially, the bench stope is extracted to a maximum stable length, followed by tight filling to the brow. Any subsequent blasting is then undertaken with no free face as shown in Figure 2.30. The success of

Filling

Filling

Production blasting (no free face) Mucking

Continuous AVOCA fill

Bench limit

Ore

Extracted and filled

FIGURE 2.30 Full Avoca bench extraction method. (From Villaescusa, E. and Kuganathan, K., Backfill for bench stoping operations, in M.L. Bloss, ed., Minefill 98, Proceedings of the Sixth International Symposium on Mining with Backfill, Brisbane, Queensland, Australia, April 14–16, 1998, pp. 179– 184, AusIMM, Melbourne, Victoria, Australia. With permission.)

45

Sublevel Stoping Geometry

Slot

this method is a function of the fill stability following blasting. This is controlled by the orebody width and height and the moisture and particle size distribution of the fill material used. The option of extracting a bench beyond its stable limits and then leaving a (unplanned) pillar to arrest a hangingwall failure has not been considered because it does not represent good design or operational practice. The extraction option shown in Figure 2.27 is related to extracting the bench using pillars that have been designed at the very early stages, and it is assumed that the spans between the pillars are stable and independent (from a deformational point of view) of each other. Uphole benches are often related to top-down sequences of extraction where the orebodies are partitioned into blocks separated by horizontal crown pillars. Individual uphole benches are defined within a block, and retreated to a central or end access crosscut. Typical uphole drilling heights range from 15 to 25 m, and the individual rings are inclined forward (70°) to promote a safe brow for the blasthole charging crews. The design of forward dumping rings also reduces muck throw, which in turn minimizes remote mucking. Hangingwall reinforcement is provided from the drilling drives. In addition, in good-quality rock masses, filling can be introduced following the extraction of an entire stoping block (Figure 2.31). Hydraulic fill Crown pillar extraction

ble

bo

lti

ng

Crown pillar

in

gw

all

ca

985 L

Ha ng

Hydraulic fill

Hydraulic fill

965 L

Rock fill

945 L

925 L

Crown pillar Pillar extraction

Open void

890 L

Ramp access

Slot drilling

Broken ore

870 L

850 L

Pillar

Slot

Uphole drilling

Pillar extraction Retreat to access Open void Broken ore

Uphole drilling

830 L

Crown pillar 785 L

Cross-sectional view

Central access

Level development

Longitudinal view

FIGURE 2.31 Schematic of uphole bench extraction sequences, Osborne Mine. (After Kaesehagen, M.R. and Boffey, R.H., Development of the Osborne Mine—with a focus on technical and operational aspects, Proceedings of the Seventh Underground Operators’ Conference, Townsville, Queensland, Australia, June 30 to July 3, 1998, pp. 29–37, AusIMM, Melbourne, Victoria, Australia. With permission.)

3 Planning and Design

3.1 Introduction Mine planning is an engineering process that encompasses all of the major technical functions undertaken in sublevel open stoping, with the key performance indicators being safety, dilution control, recovery, productivity, and mining cost. Mine planning provides the means for the safe, efficient, continuous, and economic recovery of ore while considering the life of mine issues and their implications for short-term planning and design. It also helps to maintain the long-term security of production, while ensuring satisfactory economic returns (Trout, 1997). Mine planning prepares and evaluates all future stope design and operating strategies. Parameters such as ore reserve estimation, overall sequences of extraction, dimensioning of regional pillars and sublevel intervals, design of ore haulage systems, as well as fill and ventilation systems are determined during the process (see Figure 3.1). Although it is beyond the scope of this book to review such topics in detail, geotechnical aspects of the process from orebody delineation to stope extraction are considered within this chapter. The approach suggested here requires interaction between geology, mine planning, rock mechanics, and operating personnel throughout the entire mine-planning process (Villaescusa, 1998). The overall rational methodology for the stope planning process is shown in Table 3.1. The orebody delineation and rock mass characterization stages constitute the basic inputs. The requirements consist of an early determination of rock mass properties on a block scale, followed by the selection of the mining method and an estimate of the likely loading conditions from the stoping sequences. The process requires both global and detailed design stages. Global design issues are relevant and applicable within entire areas of a mine, such as an extension of an existing orebody, while detailed design issues are applicable to the extraction of individual stopes. Finally, a monitoring and back analysis strategy that allows a documented closure of the design loop is required.

47

48

Geotechnical Design for Sublevel Open Stoping

Orebody Orebodydelineation delineation

Geology

Rockmass characterization Rockmass characterization

Geology and Geology & rock rock mechanics

Mining method selection

Mine planning

Access and&infrastructure Access infrastructure

Rock Rockmechanics Mechanics

Global Globalsequences sequences (stress (Stressanalysis) analysis)

Mine planning and rock mechanics

Global Globaleconomics Economics

Mine Mine planning planning

d e s i g n

Acceptable Acceptable design design Yes Infill Infilldelineation delineation drilling drilling

D e t a i l e d

Geology Geology

Drill blastdesign design Drilland & blast

Mine Mine planning planning

Rock Rockreinforcement reinforcement

No

G l o b a l

Mine Mine planning planning

Stope and pillarsizes sizes Stope & pillar

No

Input data

Rock Rockmechanics mechanics

Detailed Detailedeconomics economics

Mine Mine planning planning

Extraction monitoring

Operations, mine planning geology and rock mechanics

Acceptable Acceptable design design

Yes

Document Document results results

End

d e s i g n Closure of design loop

FIGURE 3.1 Flowchart of mine-planning process. (From Villaescusa, E., Geotechnical design for dilution control in underground mining, in R.K. Singhal, ed., Proceedings of the Seventh International Symposium on Mine Planning & Equipment Selection, Calgary, Alberta, Canada, October 5–9, 1998, pp. 141–149, Balkema, Rotterdam, the Netherlands.)

TABLE 3.1 Key Stages within a Stope Planning and Design Process Stope Design Process Stages Basic Input Orebody delineation Rock mass characterization Mining method selection

Control of Ground Behavior

Closure of the Design Loop

Stope block design Detailed stope design

Monitoring Back analysis Documentation

Planning and Design

49

3.2  Geological and Geotechnical Characterization The orebody delineation and rock mass characterization stages provide the input for the entire stope design process. In most cases, however, the main role of a mine geology department is limited to the definition and delineation of the ore zones within a deposit, the geological interpretation for further delineation and exploration strategies, and making ore reserve estimations. Rock mass characterization is rarely undertaken by mine geology as a routine process, as the significant demands of a robust orebody delineation leave no time for additional geotechnically related duties. Sometimes, a lack of proper training and awareness of the relevant geotechnical issues by the mine geologists also contributes to deficient data collection approaches. The suggested approach is to obtain representative, mine-wide, rock mass properties required during the subsequent global excavation design and stability analysis stages. This information is obtained from diamond drill holes consisting of mainly core logging of nonoriented holes followed by direct mapping of underground openings. Geophysical tools can also be used for orebody delineation and rock mass characterization, but such techniques have not been widely implemented to date. The confidence in the geological information must be sufficient to establish the nature and irregularities of the orebody, the nature and location of major controlling geological structures, the general rock mass characteristics, as well as allowing an economic evaluation to be carried out to determine whether a particular stoping block should be mined. This type of information requires that the sampling process extends beyond the orebody boundaries in order to determine the likelihood of failure from orebody hangingwalls, footwalls, or stope crowns. The first step in any rock mass characterization process is a three-­dimensional definition of rock-type contacts and alteration halos. In addition, large-scale geological discontinuities such as faults and shears likely to play a major role in the overall mechanical behavior of the entire deposit must be identified. The second step in a rock mass characterization program is to determine the rock mass behavior away from the main geological discontinuities by defining structural domains for design. This can be achieved by core logging and direct mapping of joint set characteristics such as number of joint sets, joint orientation, frequency, trace length, planarity, and surface strength (Brown, 1981).

3.3  Stress Analysis in Stope Design As illustrated in Figure 3.1 and discussed in more detail subsequently in this chapter and in Chapter 5, many of the steps in the overall stope

50

Geotechnical Design for Sublevel Open Stoping

block and detailed stope design processes require the use of some form of stress analysis. In modern mining practice, computational or numerical methods of stress and deformation analysis are used to evaluate the stresses and deformations induced around excavation boundaries and within the surrounding rock mass as a result of excavation. They find widespread use as aids to decision making in establishing overall open stoping layouts and extraction sequences, and in the detailed design and dimensioning of components of the overall mining structure, including items of infrastructure, accesses, stopes, and pillars. The numerical methods that are most commonly used in addressing mining rock mechanics problems may be classified as follows (Jing, 2003): Continuum methods • Finite difference method (FDM) • Finite element method (FEM) • Boundary element method (BEM) Discontinuum methods • Discrete or distinct element method (DEM) • Discrete fracture network (DFN) methods Hybrid continuum/discontinuum models • • • •

Hybrid FEM/BEM Hybrid BEM/DEM Hybrid FEM/DEM Other hybrid models

Figure 3.2 illustrates the two-dimensional discretization concepts used in the FDM, FEM, BEM, and DEM for a fractured or discontinuous rock mass. As shown in Figure 3.2, the modeling of faults in FDM, FEM, and BEM requires the introduction of special joint or displacement discontinuity elements. The discussion of the details of the numerical methods of stress and deformation analysis used in underground mining, including open stoping, is beyond the scope of this book. Useful general introductions are given by Brady and Brown (2004) and Jing (2003). The following overview of the main numerical methods is based on that given by Brady and Brown (2004). Computational methods of stress analysis may be divided into two categories: differential methods and integral methods. In differential methods, the problem domain is divided or discretized into a set of subdomains or elements as shown in Figure 3.2b. The solution procedure may be based

51

Planning and Design

Joints

Faults

Joint element (b)

(a) Region 1 Region 4

Block Region 2

Block

Region 3

(c)

Element of displacement discontinuity

(d)

Regularized discontinuity

FIGURE 3.2 Two-dimensional representation of the fractured rock mass shown in (a) by (b) FDM or FEM, (c) BEM, and (d) DEM. (Reprinted from Int. J. Rock Mech. Min. Sci., 40(3), Jing, L., A review of techniques, advances and outstanding issues in numerical modelling for rock mechanics and rock engineering, 283–353, Copyright 2003, with permission from Elsevier.)

on numerical approximations of the governing equations, that is, the differential equations of equilibrium, the strain–displacement relations, and the stress–strain equations, as in FDM. Alternatively, the procedure may exploit approximations to the connectivity of elements, and continuity of displacements and stresses between elements as in the FEM. The FEM can readily accommodate nonlinear and heterogeneous material properties, but the problem domain is defined arbitrarily, and discretization errors may occur throughout the domain (Brady and Brown, 2004). In integral methods, the problem is specified and solved in terms of surface values of the field variables of traction (surface stress components) and displacement. As only the problem boundary is defined and discretized as in Figure 3.2c, this BEM effectively provides a unit reduction in the dimensional order of the problem. This offers a significant advantage in terms of computational efficiency, particularly in the solution of three-dimensional problems. These methods are best suited to linear material behavior and homogeneous material properties. However, they model far-field boundary conditions correctly, restrict discretization errors to the problem boundary, and ensure fully continuous variations of stress and displacement throughout the medium (Brady and Brown, 2004).

52

Geotechnical Design for Sublevel Open Stoping

The DEM represents the fractured rock mass as an assembly of blocks interacting through deformable discontinuities having definable stiffnesses. The equations of motion of these blocks are solved through continuous detection and treatment of the contacts between blocks. The blocks can be rigid or made deformable using FDM or FEM discretizations. The method can model large displacements caused by the rigid body motion of individual blocks including block rotation, fracture opening, and complete block detachment (Jing, 2003). Detailed accounts of the fundamentals of discrete element methods and of their application in rock engineering are given by Jing and Stephansson (2007). Examples of the use of numerical modeling in extraction sequencing, assessing stope wall stability, pillar design, and ground support design are given in Sections 3.4, 5.4, 5.5, and 7.3, respectively.

3.4  Design of Stoping Blocks Stope block design issues are related to the global design and stability of large sections of a mine, such as a new adjacent orebody, extensions at depth, or in the abutment of an existing deposit (Chileshe and Kulkarni, 1995). Global design issues are represented schematically in Figure 3.1 and listed in detail in Table 3.2. The issues involved include global orebody delineation, mine access and infrastructure, dimensions of sublevel intervals, fill requirements, equipment, and ventilation considerations. Stress analysis of the global production schedules is critical to determine the loading conditions likely to result from any proposed mine-wide stoping sequences. TABLE 3.2 Stope Block Design Issues Exploration drilling requirements for orebody delineation for the designed area Area-wide rock mass characterization from borehole data and direct access Overall mining method selection Quantity and grade of ore required with respect to scheduled metal targets Access and infrastructure development requirements—ore-handling systems, workshops, etc. Production scheduling, details, and timing Induced stresses from scheduled sequences, including extraction directions Primary and secondary stope dimensions, including regional access pillars Fill system requirements Equipment requirements Ventilation Global economic assessment

Planning and Design

53

3.4.1  Orebody Delineation The geological analysis on a block scale requires information on orebody boundaries, grade, major geological structures, as well as the major rock types within and around the orebodies. A grade distribution and a geotechnical model on a block scale are constructed from the geological interpretation of the data, which is initially collected from widely spaced surface diamond drill holes. The preliminary design of a stoping block layout is based on confirmatory exploration drilling, with holes drilled at 60–80 m spacing. Additional geological information is required to provide the ore limits and grade information suitable for a detailed stope design. This information can be collected as underground access becomes available and stope delineation drilling at 20–40 m spacing can be carried out. In addition, geological and geotechnical mapping is carried out from the exposed rock mass around a stope block development. The geological and geotechnical models are used by a mine-planning engineer to develop a geometrical model of a stoping block in three dimensions. The major geological structures likely to influence overall block stability are determined and included in the analysis. The resulting three-dimensional model is then used to calculate tones and grade for the global design block (Thomas and Earl, 1999). Following mining method selection and an economic analysis for the block, the design of the development, ore- and wastehandling systems, services, and ventilation can be undertaken. 3.4.2  Global Extraction Sequences One of the limiting factors affecting the design of an underground excavation is the maximum excavation size that a rock mass can sustain without failure. This failure may take place either as a function of movement along planes of weakness or through a combination of failures through intact rock and on geological discontinuities. In most orebodies suitable for open stoping, the volume that may be excavated safely such that stope wall failures are avoided, is many times smaller than the orebody itself. Consequently, a series of individual stopes must be excavated to achieve full orebody extraction. One of the most important tools that a design and planning engineer has for controlling the overall behavior of a rock mass is the extraction sequence of the stopes contained within a given area of an orebody. Extraction sequences are fundamental to achieving production targets safely and economically throughout a stope life. In most stoping mines, various stages of development, production, and filling occur at any one time. The ore sources are likely to be scheduled from a number of locations and extraction methods. In general, a stoping sequence is driven by ore grade requirements, operational issues, and induced stress considerations (Potvin and Hudyma, 2000). A technically sound strategy is to avoid creating blocks of highly

54

Geotechnical Design for Sublevel Open Stoping

stressed rock within an orebody. This can be achieved by retreating stopes to an orebody abutment instead of creating pillars located within central orebody areas (Beck and Sandy, 2003). In general, the overall stope extraction sequence is influenced by the nature of the orebody in question. 3.4.2.1  Massive Orebodies Massive orebodies are extracted using multiple stopes (primary, secondary, and, when required, tertiary) in conjunction with mass blasting techniques and cemented fill. A number of sequencing options can be used including temporary or permanent rib or transverse pillars, strike slots with continuous or discontinuous advance, and chequerboard sequences. Each overall extraction sequence can be engineered to manage the induced stress redistributions on a global scale. Ideally, the initial stopes are extracted within a chosen area of an orebody and subsequent stopes are retreated systematically toward orebody abutments taking into account the stress redistributions, production tonnage requirements, and access constraints. One extraction option used in extremely good-quality rock masses is to mass blast secondary stopes into adjacent primary stopes to create very large, but stable, openings (Mikula and Lee, 2000). In order to increase recovery and achieve stability, the resulting voids can be filled using either consolidated or unconsolidated fill with the individual stopes separated by rib (longitudinal) and transverse pillars (Figures 3.3 and 3.4). The latter option leaves a high proportion of ore tied up in the rib and transverse pillars. Methods such as sublevel caving retreat have been used to achieve complete recovery of these pillars (Alexander and Fabjanczyk, 1981). The concept of a discontinuous strike slot for a 12-stope extraction sequence is shown in Figure 3.5. Assuming the major principal stress to be normal to the long axis of an orebody, the primary, secondary, and tertiary stopes are designed with an overall stress management philosophy consisting of stress shadowing and orebody abutment retreat. Once the strike slot has been completed (stopes 1–4), all the remaining stopes are effectively stress shadowed. Stress shadowing occurs when two or more excavations are aligned along a major principal stress trajectory. Stresses redistribute, and some areas may be stress-relieved as the rock lies in the shadow cast by the excavations. In addition, stress may be intensified in other areas, depending upon the distance between the excavations (Figure 3.6). In very high stress environments, sequences using transverse pillars or discontinuous transverse/strike slots may concentrate stress even in the early stages of extraction. At the Creighton Mine in Canada, a series of central stopes were extracted adjacent to each other to form a continuous slot within an initial mining block in order to create a stress shadow for the remaining stopes (Figure 3.7). In order to form a continuous strike slot, the fill from the initial stope must be cured before extraction of the immediately adjacent stopes can proceed. Production from the first three stopes is slowed by the

55

Planning and Design

Open cut Open cut

100 m 5L

Block A Cut and fill (filled)

A Cut and fill

7

Open stope B3

9L 300 m Crown pillar

C 2

C3

C1

D1

13L

F4

B1

B2

B1

3

1

2

6

C4

F3

Block B detailed extraction sequence

E1

Blocks A,B,C, D, E, and F extracted and filled with unconsolidated fill

F1

F2

5

7

21L

G4

G3

G4

G1

G2

4

B3

D2 E2

E3

15L 500 m

B2

5

9

G5

G3 4

6

3

G2 1

2

8 H

Orebody outline

Rib pillar

Block G detailed extraction sequence

5500 N

5000 N

4500 N

FIGURE 3.3 Temporary rib pillars within a top-down sequence at Mount Charlotte Mine. (From Ulla, Z., Applicability of the Mathews stability graph for evaluating stability of open stopes at the Mount Charlotte Mine, MEngSc thesis, Curtin University of Technology, Kalgoorlie, Western Australia, Australia, 1997.)

1500 E

Rib pillar

S4

8f

(a)

Transverse pillar

au

2000 E

lt

(b)

FIGURE 3.4 (a) Rib and transverse pillars at Mount Isa Mines and (b) transverse pillar at the Darlot Gold Mine. (From Alexander, E.G. and Fabjanczyk, M.W., Extraction design using open stopes for pillar recovery in the 1100 ore body at Mount Isa, in D.R. Stewart, ed., Design & Operation of Caving & Sublevel Stoping Mines, SME of AIME, New York, 1981, pp. 437–458.)

56

Geotechnical Design for Sublevel Open Stoping

10

3

σ1

7

2SLOS 1SLOS 2SLOS

σ1

8

1

6

4

12

2

2SLOS

P

5

1SLOS

9

1SLOS 2SLOS 2SLOS

σ1

2SLOS

P

11

2SLOS

10

3

8

σ1

2SLOS

σ1 σ1 σ 1

7

2SLOS 1SLOS 2SLOS

6

12

σ1

11

7

2SLOS

6

σ1

σ1

2SLOS

8

1SLOS

9

10

2SLOS

5

4

1SLOS 2SLOS2SLOS 2SLOS

σ1

5

4

1SLOS

9

1SLOS 2SLOS2SLOS

12

2SLOS

2SLOS

σ1 σ

1

11

2SLOS

σ1

FIGURE 3.5 Plan view of discontinuous transverse slot extraction sequence for a massive orebody. (From Villaescusa, E., Extraction sequences in sublevel stoping, Proceedings of the 12th International Symposium on Mine Planning & Equipment Selection, Kalgoorlie, Western Australia, Australia, April 23–25, 2003, pp. 9–18, AusIMM, Melbourne, Victoria, Australia. With permission.)

Low stress Stress concentration area lines

Extracted stopes

Low stress area

Highly stressed area

Low stress area

Low stress area

Undisturbed stress field

FIGURE 3.6 Plan view showing stress shadowing across a series of stopes. (From Villaescusa, E., Extraction sequences in sublevel stoping, Proceedings of the 12th International Symposium on Mine Planning & Equipment Selection, Kalgoorlie, Western Australia, Australia, April 23–25, 2003, pp. 9–18, AusIMM, Melbourne, Victoria, Australia. With permission.)

requirements to not expose the initial fill mass simultaneously on both sides. This means that the third stope within the strike slot must wait until the fill mass in the second stope has cured. Another alternative is to adopt a chequerboard pattern of stope extraction. The process starts with primary stopes filled with consolidated fill followed by secondary and tertiary stope extraction of stope pillars having multiple fill mass exposures (Grant and DeKruijff, 2000). The stoping front can either move longitudinally or adopt a continuous retreat strategy depending upon

57

Planning and Design

12 9

9

10

6

10

7

3

7

4

1

5

8

2

8

11

6

11

9

9 12

FIGURE 3.7 Plan view showing a continuous, pillarless, stoping sequence. (After Min. Sci. Technol., 13, Trotter, D.A., Vertical crater retreat mining in the Sudbury Basin, 131–143, Copyright 1991, with permission from Elsevier.)

the level of in situ stress and the production tonnage requirements. Figure 3.8 shows the massive 1100 orebody at Mount Isa Mines, in which a north to south global extraction sequence has continuously stepped out to access primary stoping blocks. The extraction was designed with large, 40 × 300–400 m east–west transverse pillars for access, ventilation, and services (Grant and DeKruijff, 2000).

1100 Orebody 1500 m

Hangingwall lens Lower footwall lens

Footwall lens

Northern 1100 orebody

1900 Orebody Man and supply shaft

Stopes to be mined Stopes filled or empty

Mount ISA mines Copper mine

FIGURE 3.8 Plan view of the 1100 orebody showing extracted stopes. (From Grant, D. and DeKruijff, S., Mount Isa Mines—1100 orebody, 35 years on, in G. Chitombo, ed., Proceedings of the MassMin 2000, Brisbane, Queensland, Australia, October 29 to November 2, 2000, pp. 591–600, AusIMM, Melbourne, Victoria, Australia. With permission.)

58

Geotechnical Design for Sublevel Open Stoping

The advantages of a properly designed chequerboard extraction sequence include stable primary stopes which must be tight-filled in time to provide support to the remaining stopes and crown pillar (Alexander and Fabjanczyk, 1981). A disadvantage is the large amount of ore tied up within the remaining tertiary stope pillars, where localized stope design can be complex and a function of existing development and the number of fill exposures as mine life progresses. A chequerboard sequence is dependent upon successful mass blasting practices and the development of stable fill masses that provide support to adjacent rock masses with minimal dilution during multiple-fill exposures (Bloss, 1992). 3.4.2.2  Steeply Dipping Orebodies In the case of steeply dipping and relatively narrow orebodies, the most common orebody access is through crosscuts off access drives that are connected to ramps located in the footwall of the orebodies. The crosscuts intersect the orebodies from footwall to hangingwall and ore drives are developed from the crosscuts along the strike of the intersected orebodies. In cases where bench stoping is used as the preferred mining method, extraction can be retreated toward the access crosscuts using either a top-down or a bottomup extraction sequence. A top-down bench stope extraction sequence usually requires permanent rib pillars to minimize dilution between individual stopes along strike. In addition, a series of sill pillars may be required to control overall stability and dilution and to isolate any unconsolidated fill that may be introduced into the upper stopes as extraction progresses downward (Figure 3.9). A bottom-up sequence requires fill in order to provide a working floor as the extraction proceeds upward. The need for crown pillars is minimized by the use of rib pillars along the strike of the orebody and the beneficial impact of the fill masses (Figure 3.10). Flexibility and productivity can be greatly enhanced with the introduction of two access crosscuts as shown in Figure 3.11. Although more costly, this configuration increases tonnage and allows for a better stress redistribution as the initial stopes can be located in the center of the mining block with subsequent retreat toward the abutments. In cases where multiple-lift sublevel stoping is used to extract narrow tabular orebodies, a series of primary and secondary stopes can be designed along the strike of the orebody. The pillar stopes are designed large enough to enable safe recovery between primary stopes. Figure 3.12 shows the stoping sequence for the Kanowna Belle orebody, Stoping Block A. The extraction sequence is based upon a primary stope extraction and filling with consolidated fill, before the secondary pillars are extracted. In other cases, stope extraction in conjunction with unconsolidated fill and separated by permanent pillars can be used to extract low-grade orebodies (Figure 3.13).

59

Planning and Design

Drill drives parallel to orebody 1

3

2 Permanent rib pillars

4

Central access crosscuts

5

2

3 Stope void unfilled

Permanent sill pillar 8

7

8

10

9

4

5

Permanent sill pillar 6

1

6

7

10

9

FIGURE 3.9 Longitudinal view of a top-down extraction sequence, permanent pillars, and retreat to a central crosscut, no fill.

Drill drives 9

8

6

Rib pillars optional 4

3

1

5

2

8

Filled stope

Central access crosscuts

7

9

10

10

6

7

5

3

4

2

1

FIGURE 3.10 Longitudinal view of a bottom-up extraction sequence, retreating to a central crosscut; pillars are optional.

Figure 3.14 shows an example of primary and secondary stope extraction from Mount Isa Mines, where some of the secondary stopes are mass blasted into the void created by the primary stopes in what is called a “triplet” stope extraction (Bywater et al., 1983). Figure 3.15 shows a primary and secondary stope extraction sequence for a shallowly plunging orebody at Kambalda, Western Australia. The cost of cement binder in the fill is minimized by filling the secondary stopes with unconsolidated waste rock.

60

Geotechnical Design for Sublevel Open Stoping

Retreat to abutment access

Retreat to abutment access

6

5

4

5

6

3

2

1

2

3

Optional pillars

Uncemented fill

FIGURE 3.11 Longitudinal view of a bottom-up extraction sequence with double access.

The advantage of primary and secondary stoping sequences lies in the initial high flexibility and productivity and low cost during primary stoping. The overall cost is minimized by the use of unconsolidated fill within the secondary stopes. A disadvantage is that stress redistributions may cause rock mass damage late in the extraction sequence. Figure 3.16 shows an example in which the induced stresses increase as the stope extraction progresses within an abutment area. The results show normal stress in excess of 70 MPa in the crown pillar region of the seven orebody L692–L698 stopes at Mount Isa Mines. The induced stresses were predicted using the program NFOLD, and confirmed with in situ stress measurements. Previous studies indicated that 70 MPa compressive stress was considered to be a critical value within the seven orebodies. Field observations identified severe spalling and cracking of the L690 pillar on the 13th level (Bywater et al., 1983). The effects of stress can be minimized by avoiding the undercutting of individual stopes and by mass blasting those highly stressed regions within a stoping block. Multiple-lift primary and secondary stopes have been used very successfully to achieve complete extraction with minimal dilution within the steeply dipping lead orebodies at Mount Isa Mines (Goddard, 1981; Bywater et al., 1983; Beck et al., 1997) and also at the Kanowna Belle Mine (Magee, 2005; Cepuritis et al., 2007). 3.4.2.2.1  Pillarless, Center-Out Sequences Pillarless, center-out mining sequences have been proposed to eliminate the need for secondary stopes (Morrison, 1996). The perceived advantage from such sequences is the slow rate of convergence of the host rocks as stoping

190 m Sub

13

Filled

9 2

7

8

1

14

Fill in progress

18

10

3

5

11 15 4

Current source

17

6

16

Decline

Scheduled

12

Vent raise

FIGURE 3.12 Longitudinal view of Kanowna Belle Mine—Stoping Block A. (From Bywater, S. and Fuller, P.G., Cable support of lead open stope hangingwalls at Mount Isa Mines Limited, in O. Stephansson, ed., Rock Bolting: Theory and Application in Mining and Underground Construction, Proceedings of the International Symposium on Rock Bolting, Abisko, Sweden, 1983, pp. 539–555, Balkema, Rotterdam, the Netherlands.)

200 m RL

160 m Sub

100 m RL 110 m Sub 135 m Sub

Vent raise

Planning and Design 61

62

10/7/01

3/7/01

23/6/01

19/6/01

13/6/01

6/6/01

14/5/01

9/5/01

11/5/01

Geotechnical Design for Sublevel Open Stoping

10/7/01

3/7/01

23/6/01

1

10/7/01

3/7/01

3 23/6/01

21/6/01 19/6/01 21/6/01

Permanent pillar

6/6/01 6/6/01

20/5/01 20/5/01

17/5/01

3/5/01

14/5/01 14/5/01

2

11/5/01 17/5/01

4/5/01

3/5/01

1

Permanent pillar

2/5/01

Overall sequence

FIGURE 3.13 Permanent pillars left between primary stopes (filled with unconsolidated fill)—Mount Marion Mine. (From Villaescusa, E., Extraction sequences in sublevel stoping, Proceedings of the 12th International Symposium on Mine Planning & Equipment Selection, Kalgoorlie, Western Australia, Australia, April 23–25, 2003, pp. 9–18, AusIMM, Melbourne, Victoria, Australia. With permission.)

proceeds from the center toward the orebody abutments (Figure 3.17). It is argued that the slow rate of convergence is likely to minimize the magnitude of the local seismic events. In addition, the small single-lift stopes may reduce the amount of released seismic energy. Such a pillarless stoping sequence was used in Block 3 at the Golden Giant Mine in Canada and named pyramid retreat, as mining progresses in a triangular shape (Potvin and Hudyma, 2000). The Golden Giant Mine pyramid retreat sequence is illustrated in Figure 3.18. Although a continuous advancing stoping sequence is an attractive idea, it is very difficult to implement in practice, especially when fill is introduced into the system (Grice, 1999). The overall productivity is severely constrained by the individual stope cycle times as stopes must be mined, filled, and cured before an adjacent stope can be extracted. With active mining on a large number of sublevels, substantial development, scheduling, and logistic challenges are experienced throughout the stoping block (Potvin and Hudyma, 2000). As an example, the extraction of stope No. 6

M 654

4

M 657

1 Extraction sequence

CHF

6 7

M 660

5

0

in meters

1

M 667

Scale 25 50

3

M 662

2

11

100

M 676 9

M 674 8

M 683

M 678 N

3 10 4

Cut and fill 15 L

15 B

14 L

HF

J 668 CHF

7

J 672

3

1 Extraction sequence

4

J 679 5

4

J 684 2

100

J 686

in meters

Scale 0 25 50

1

8 6

12

13

9 4 11 10

J 694 J 696 J 698

N

FIGURE 3.14 Longitudinal section view showing secondary stopes mass blasted into the void created by a primary stope, Mount Isa Mines. (From Bywater, S. and Fuller, P.G., Cable support of lead open stope hangingwalls at Mount Isa Mines Limited, in O. Stephansson, ed., Rock Bolting: Theory and Application in Mining and Underground Construction, Proceedings of the International Symposium on Rock Bolting, Abisko, Sweden, 1983, pp. 539–555, Balkema, Rotterdam, the Netherlands.)

HF

15 B 15 L

M 651

6 500 N

14 C 2800 RL 14 L

M 665

6 600 N

M 671

6 700 N

M 680

6 800 N

13 L 14 C

6 700 N

J 670

J 675

13 L

6 000 N

J 682

6 800 N

J 701 7 000 N

J 689 J 691 6 900 N

J 704

Cut and fill (11L–13L)

Planning and Design 63

64

Geotechnical Design for Sublevel Open Stoping

P1 P1

S1 S1

P2 P2

S2 S2

P3 P3

S3 S3

P4 P4

S4 S4

FIGURE 3.15 Primary and secondary stope extraction sequence for a shallowly plunging orebody. P1, primary stope cemented fill; S1, secondary stope unconsolidated fill. The numbers show the sequence of stope extraction.

(Figure 3.18), although very early in the sequence, requires seven operational sublevels. Pillarless stoping sequences are more suited to paste fill as they require rapidly curing cemented fill with minimal drainage delays in all the stopes, thus potentially increasing the operating cost. In addition, tight backfill of the stope crowns is rarely achieved, especially when cemented rock fill is used (Figure 3.19). Introducing hydraulic fill to achieve tight fill is timeconsuming, expensive, and sometimes not practical. Consequently, large crowns that require extensive rock reinforcement are exposed by this method. In some cases, damage from stress concentrations (cracking through intact rock or slip on geological structures) in the stope brows is also experienced. This may create difficulties during drilling and blasting and make the reinforcement schemes inefficient, as very large slabs of rock parallel to the stope edges may be released. Figure 3.20 shows a pillarless stope extraction sequence where the stopes are partially mined under cemented paste fill. This sequence was implemented to allow the extraction of stopes under very high stress at the Junction Mine, Kambalda, Western Australia. The pillarless sequence was facilitated by the large initial extraction already under way within the center of the orebody at the Junction Mine. The pillarless extraction sequence actually evolved around the edges of the previous extraction. Figure 3.21 shows a typical view inside one of the open stopes, with paste fill constituting the back or roof of the stope.

65

Planning and Design

Cut and fill (11–13/L)

L698

L695

L692

L690

L687

L685

L683

13/L

Planned stopes

14/C

Cut and fill

14/L 15/D 15/B 15/L

MPa

Extracted stopes

Depth stress

Extracted

28

31

Extracted stopes

Extracted stopes

Induced normal stress

+70 MPa

50–59 MPa

60–69 MPa

40–49 MPa

0

25

50

100 m

FIGURE 3.16 Induced stress as stope extraction progresses. (From Bywater, S. et al., Stress measurements and analysis for mine planning, Proceedings of the Fifth Congress of the International Society for Rock Mechanics, Melbourne, Victoria, Australia, April 11–15, 1983, pp. D29–D37, Balkema, Rotterdam, the Netherlands.)

In practice, continuous retreating sequences can only be applied to individual stoping blocks that are separated by crown or waste pillars. An increased number of advancing fronts increases extraction flexibility, but also increases the number of pillars that must be dealt with at a later stage. Extraction of the pillars between the continuous fronts may be complicated in areas where high induced stresses are experienced. Figure 3.22 shows a proposed longitudinal view of two stoping blocks extracted using a continuously advancing front and single-lift stopes.

66

Geotechnical Design for Sublevel Open Stoping

4

4

3

3

2

3

2

1

2

3

4

FIGURE 3.17 A conceptual pillarless stoping sequence, center-out extraction. (From Morrison, D.M., CIM Bull., 89, 46, 1996.)

3.4.2.2.2  Primary and Secondary 1-3-5 or 1-5-9 Stoping Sequences A compromise to a pillarless sequence is to use a general triangular retreat shape but with a short lift primary and secondary stope arrangement. This system has been implemented at the Williams mine in Canada and is illustrated in Figure 3.23. This methodology allows for a number of primary stopes to be mined simultaneously, hence increasing the productivity within the stoping block. Because of the detrimental effects of the stress redistributions on the pendant pillars formed in the sequence, secondary pillar stopes must be recovered as early as possible in the extraction sequence. In general, no more than two sublevels are mined ahead of a pillar before recovering it and both sides of a pillar cannot be mined simultaneously (Potvin and Hudyma, 2000). In practice, however, stoping blocks are likely to interact with one another, making extraction of sill pillars extremely difficult and costly using this method. Figure 3.24 shows a longitudinal section view of a sill pillar extraction at the Williams mine, where a single seismic event required over $4 million expenditure on rehabilitation and additional development in order to resume mining. In addition, significant delays were incurred. A variation of this method was proposed for the George Fisher Mine, Queensland, Australia, where a 1-5-9 stoping sequence was selected for extraction (Neindorf and Karunatillake, 2000). Stopes 1-5-9 are extracted as two-lift primaries and filled with consolidated fill (Figure 3.25). This is followed by another set of primary two-lift stopes (3-7-11), also filled with consolidated fill. Following the fill cure within the primary stopes 1-3-5-9-11, a set of single-lift stopes (2-6-10) is then extracted and filled with unconsolidated fill. This creates a pendant pillar, which has many degrees of freedom and relies on the fill support from the primary stopes for stability. Finally, the single-lift stopes 4-8-12 are extracted and filled with unconsolidated fill

67

Planning and Design

18

27

19

10

20

21

13

6

12

22

23

15

9

4

8

14

24

25 17

11

7

1

5

2

16

66 m Stope height

4600 level

4533 level

4466 level 26

3 4400 level 1 Extraction sequence FIGURE 3.18 Pyramid retreat at the Golden Giant Mine, Canada. (From Potvin, Y. and Hudyma, M., Open stope mining in Canada, in G. Chitombo, ed., Proceedings of the MassMin 2000, Brisbane, Queensland, Australia, October 29 to November 2, 2000, pp. 661–674, AusIMM, Melbourne, Victoria, Australia. With permission.)

68

Geotechnical Design for Sublevel Open Stoping

Cutoff raise used as fill pass Filling

Drilling

Producing

FIGURE 3.19 Conceptual continuous advance for single-lift stopes. (From Grice, T., Mine backfill—course notes for the masters of engineering science in mining geomechanics, MEngSc thesis, Western Australian School of Mines, Curtin University of Technology, Kalgoorlie, Western Australia, Australia, 1999.)

5 9 13

4 11

3 8 14

7

3 Extracted area

7 10

6 4

6 5

8

9

8

12

11

10 13

12

FIGURE 3.20 Top-down, pillarless stoping sequence, mining under paste fill at the Junction Mine, Kambalda, Western Australia.

Planning and Design

69

FIGURE 3.21 Top-down stoping under paste fill at the Junction Mine, Kambalda.

before the entire sequence is repeated up-dip. The extraction of stopes 4-8-12 also creates pendant pillars. A major disadvantage of 1-5-9 (or 1-4-7) extraction sequences using short lift stopes is their inefficient production mucking characteristics. The method effectively requires (an upward) moving drawpoint sequence (even in primary stopes), which necessarily follows the vertical retreat of the stopes, as shown in Figure 2.15. This implies that production mucking is carried out in areas that had previously been subjected to stress redistribution and stope blasting at the stope crowns. Each stope access becomes a stope drawpoint and a significant amount of reinforcement using cablebolting is required in all the stope accesses and exposed backs to minimize large-scale back failures (Figure 3.26). Reinforcement can be largely inefficient within the bottoms of pendant secondary pillars where remote mucking is required for 100% of the tonnage. Furthermore, additional footwall development access in waste may be required on each sublevel, as more than one access may be required for effective production mucking of each individual stope. 3.4.2.2.3  Multiple Steeply Dipping Orebodies The extraction sequence for multiple, steeply dipping parallel orebodies, which are accessed by a common crosscut off a footwall ramp, requires

70

Geotechnical Design for Sublevel Open Stoping

22

20

18

16

8

6

4

2

Stoping block 1 12

14

10

Crown pillar between stoping blocks

21

19

17

5

3

15

Stoping block 2 13

11

9

7

Waste pillar

FIGURE 3.22 Bottom-up continuous extraction sequences on each stoping block.

1

71

Planning and Design

P

S

P

S

P

S

P

Production blasting

1

3

Development

P

Slot raise

Filling uncemented rockfill

Cablebolts

S

25 m

25 m 5

Production blasting

25 m

Cemented rockfill

FIGURE 3.23 A conceptual longitudinal section view showing a 1-3-5 extraction sequence. (From Bronkhorst, D. et al., Geotechnical principles governing mine design at the Williams Mine, in W.F. Bawden and J.F. Archibald, eds., Proceedings of the International Congress on Innovative Mine Design for the 21st Century, Kingston, Ontario, Canada, August 23–26, Balkema, Rotterdam, the Netherlands, 1993, pp. 433–442.)

31

30

29

28

27

26

25

24

23

9475

20

19

17

16

15

14

13

12 9475 9450 9415

Major damage Caved zone

9380

Minor damage

9370

9345 9310

18

Moderate damage December 17 2.6 Mn rockburst March 29 rockburst

9415

9370

21

Filled zone

9450

9380

22

9345 Filled zone

9310

FIGURE 3.24 Longitudinal section view of crown pillar between stoping blocks 3 and 4 at the Williams mine. (After Bawden, W.F. et al., Lessons in control of mine costs from instrumented cablebolt support, in J. Girard et al., eds., Proceedings of the Fourth North American Rock Mechanics Symposium, Seattle, WA, July 31 to August 3, 2000, pp. 633–642, Balkema, Rotterdam, the Netherlands.)

additional consideration, as the stope extractions at particular locations can be interrelated. In those cases where the thickness between orebodies is less than half the stope height, the stopes are likely to interact and the stope hangingwall deformations are minimized by extracting the orebodies from footwall to hangingwall. The extraction sequence shown in Figure 3.27 aims

72

Geotechnical Design for Sublevel Open Stoping

D orebody 1, 5, 9 sequence 11L 12C 12L 13C

6

5

8

7 1

13L 1P

3 2S

6

5

8

7 1

2

3

3P 4S

4

3

4 8S

2 Fill stopes 1, 5, 9 and extract 3, 7, 11 3 Fill and cure stopes 1, 3, 5, 7, 9, 11 and extract 2, 6, 10

2

1

5P 6S 7P

1 Extract stopes 1, 5, 9 8

7

2

4

6

5

4 Fill stopes 2, 6, 10 and extract 4, 8, 12 North

9P 10S 11P 12S

3-22

3-12

3-06

3-11

3-21

3-31

3-14

3-08

3-03

3-07

3-13

3-23

3-02

3-10

FIGURE 3.25 Longitudinal section view of George Fisher conceptual orebody extraction. (From Neindorf, L.B. and Karunatillake, G.S.B., George Fisher Mine—Feasibility and construction, in G. Chitombo, ed., Proceedings of the MassMin 2000, Brisbane, Queensland, Australia, October 29 to November 2, 2000, pp. 601–609, AusIMM, Melbourne, Victoria, Australia. With permission.)

3-04

3-01

3-05

3-09

Extracted

Longitudinal view stope extraction sequence

Potential failure surface under investigation Open slot

Stope

3-11

Filled

Mucking level Filled Isometric view stope 3-11 FIGURE 3.26 Brow instability in highly stressed single-lift open stopes at Hemlo Gold Mine, Canada. (From Milne, D. and Gendron, A., Borehole camera monitoring for safety and design, Presented at 92nd CIM Annual General Meeting, Ottawa, Ontario, Canada, May 6–10, 1990, 13pp.)

73

Planning and Design

σ1

Hig

hly st zon ressed e

σ1

4

3

2

1

Lea ext ding rac tion

5

Extraction sequence Hangingwall orebody

Footwall orebody

1

Filled

FIGURE 3.27 Footwall stope extracted ahead of other stopes in the same lift. (From Villaescusa, E., Trans. Inst. Min. Metall., Sect. A Min. Ind., 105, A1–A10, 1996.)

to minimize the effects that stopes might have on each other. The stopes interact as the block extraction sequence advances up-dip toward a region of high induced stress below a mining block extracted earlier. Within this sequence, the footwall stopes are always extracted one or two lifts ahead of the hangingwall stopes, effectively creating a “leading” stope geometry. The sequence is devised to “shield” the rest of the stopes in a particular lift from excessive stress-induced damage, as well as to minimize the effects of blasting, as most hangingwalls are mined in undisturbed ground. In some cases, the leading orebodies may experience stress-related crown damage, and adequate rock reinforcement must be provided to minimize failures. Alternatively, the leading orebody must be selected following considerations of rock mass strength, orebody width, and orebody grade. In such cases, it may be advisable to select a narrow orebody (located anywhere in the sequence) as the leading orebody. 3.4.3  Numerical Modeling Induced stresses from a particular extraction sequence can be determined using numerical modeling (Beck and Duplancic, 2005; Wiles, 2006). The inputs required are an estimate of the stress field orientation and magnitude with depth, the rock mass deformational properties, the initial excavation

74

Geotechnical Design for Sublevel Open Stoping

geometry, and the chosen overall stope extraction sequence. The limitations of linear elastic modeling include the inability to predict movement, falloff, or dilution from fault or shear zones. Consequently, the results must be used in conjunction with structural information, for example, large fault behavior, in order to interpret the different stoping sequences. Alternatively, nonlinear modeling which is able to predict rock mass failure and any stress redistribution resulting from such failures can be used (Beck and Duplancic, 2005). Progressive orebody extraction may induce several phases of post-peak behavior in a rock mass, and small changes to the stress field induced by distant stope extraction may cause significant rock mass damage around the stope boundaries. Typical outputs from numerical modeling include stresses and displacements, which in turn can be compared with empirical failure criteria established for the different domains within an orebody (Brady and Brown, 2004). Any predictive models must be validated against field data and observations. Modern numerical modeling tools allow realistic assessments to be made of mine-wide extraction sequences (Figure 3.28). The model preprocessing is usually linked to a three-dimensional model of the excavation geometries in σ1 (MPa) 50.0 45.0 40.0 35.0 30.0 25.0 20.0 15.0 10.0 5.0 0.0

z

y

x

FIGURE 3.28 Major principal stress distribution in a stoping block using the program MAP3D. (From Villaescusa, E. et al., Open stope design and sequences at great depth at Kanowna Belle, Unpublished Research Report for Placer Dome Asia Pacific, 2003a, 217p.)

Planning and Design

75

order to reduce mesh generation times. A link to mine scheduling is required in order to analyze the different extraction sequencing options. 3.4.4  Regional Pillars The use of regional pillars is sometimes required to control the overall stability and to provide safe access to active stoping areas across an existing orebody. In some cases, the pillars are required for permanent access throughout the entire life of a stoping block. The use of transverse pillars to control the overall stability of massive orebodies, such as the 1100 orebody at Mount Isa Mines is well documented (Alexander and Fabyanczyk, 1981). Transverse pillars are an efficient way of controlling overall crown subsidence, while ensuring safe access through the orebody (Figure 3.29). Regional pillars were also used to provide permanent access to multiple, steeply dipping orebodies in the Lead Mine at Mount Isa Mines. Access to each of the orebodies was provided through crosscuts centrally located within 40 m wide pillars. Extraction of the orebodies retreated toward the pillar edge as shown in Figure 3.30. Provision for the recovery of such permanent pillars can be designed for late in the extraction sequence of a mining block. Stress redistributions from a global stoping sequence may cause damage to transverse or regional pillars. This damage may require rehabilitation or loss of access development through the pillar. Extension strain cracking (Stacey, 1981) parallel to the direction of the major principal stress orientation may be experienced, especially in rock masses exhibiting a high modulus. Consequently, an eventual recovery of transverse pillars must be planned carefully, ideally with the initial pillar stope located in the best-quality rock mass area. Extraction of the initial stope may allow an overall stress reduction within the pillar, as a stress shadow is likely to be created for the adjacent transverse pillar stopes. In the example shown in Figure 3.31, extraction of stope A, as the first stope in the transverse pillar, may actually activate the fault causing shearing and failure into the stope. On the other hand, extraction of stope B as the first stope within the transverse pillar may cause a reduction of stresses through the pillar, minimizing the potential for shearing along the fault. Damage to permanent pillars is not entirely determined by stress-induced behavior, as preexisting geological discontinuities can also influence the performance of a pillar. Monitoring has linked stoping activities and instability in concurrent extraction areas along the strike lengths of large fault zones (Logan et al., 1993). The resulting behavior can be linked to induced stress relief along the structures with increased loading and degree of freedom. Large stope blasts can transmit energy along continuous fault zones, and fill drainage may introduce water into fault systems. As a result, production and

4000 mN

V405

S434

Y434

U434 U438

R454

T446 U450 T454

S447 T450 S454

S450

R450

Q451 Q455

Recently filled stope

U442

T438 T442

S442 S446

R442

Q442 Q446

Q450 Q454

P450 P454 P458

P465

Q465

Q465

Q461

P461

5

Scheduled stope

T4

Producing or empty stope

W426

V430

Q438

P438

R432 S438

R434

T430 T434

S430

Q431

R430

Q435

P442 P446

N462 O458 Q461

°

Filled stope

V409

T422 T426

R426

Q426 Q430

Q434

P426 P430 P434 O438 O442 O446

N454

P471

M465 M469 66° J46

N461 N465 N458

80

V401

U418

M444

L473 L473

S4

5000 mN

65 ° 8

FIGURE 3.29 A plan view of the Mount Isa Mines 1100 orebody showing transverse pillar access, large-scale discontinuities, and scheduled stopes. (From Villaescusa, E., Extraction sequences in sublevel stoping, Proceedings of the 12th International Symposium on Mine Planning & Equipment Selection, Kalgoorlie, Western Australia, Australia, April 23–25, 2003, pp. 9–18, AusIMM, Melbourne, Victoria, Australia. With permission.)

55° P41

U409

N430 N434 M438

4500 mN O426 O430 O434 N438 N442 O447

N426

° 70 44 M

U403

ary rim 6 p llar M405 M409 M413 M418 39 i p N422 M 99 N3 01 N401 N405 N409 N413 N418 N3 2 9 97 N3 N3 O418 O422 5 39 O401 O405 O409 O 2 O413 9 O3 P418 P422 P397 P401 P405 P409 94 P413 3 P Q397 Q418 Q422 Q401 Q405 Q409 83 Q413 Q3 Q398 R401 R405 R409 R418 R422 96 R3 7 9 S408 R413 S405 3 Q421 S S400 95 S3 S413 S409 S409 S418 S422 S401 60° J46 T405 T409 T413

M422

76 Geotechnical Design for Sublevel Open Stoping

77

Planning and Design

N

S Drilling horizon

To additional orebodies

(Sill drive) Ring blasting

Access crosscut

Broken ore

Mucking horizon

Permanent pillar

Drill holes

Sill drives

To access ramp

Crosscut

Previous stope filled

To access ramp

Previous mucking horizon Longitudinal section

Cross section looking north

FIGURE 3.30 Longitudinal and cross-sectional view of a typical permanent pillar in the Lead Mine, Mount Isa. (From Kropp, W. and Villaescusa, E., Development of mining practices in the Lead/Zinc Mine, Mount Isa, Proceedings of the Ninth Australian Tunnelling Conference, Sydney, New South Wales, Australia, August 27–29, 1996, pp. 461–466. With permission.)

High stress

Stope A

Fault zone

Stope B

Extracted High stress

Pillar stope

FIGURE 3.31 Plan view of an initial extraction in a transverse pillar stope. (From Logan, A.S. et al., Geotechnical instrumentation and ground behavior at Mount Isa, in T. Szwedzicki, ed., Geotechnical Instrumentation and Monitoring in Open Pit and Underground Mining, Proceedings of the Australian Conference, Kalgoorlie, Western Australia, Australia, June 21–23, 1993, pp. 321–329, Balkema, Rotterdam, the Netherlands.)

78

Geotechnical Design for Sublevel Open Stoping

filling strategies must minimize stope interaction along common faults that intersect permanent pillars (Logan et al., 1993). 3.4.5  Block Development The purpose of a block development is to provide suitable access for stoping and ore handling, fill reticulation, ventilation, mine services, as well as gaining further and more detailed information about the nature and size of the orebody. The two main factors to be considered are the mode of entry into the underground workings and the related lateral development required to extract the orebodies. The layout of the basic development depends upon the orebody characteristics, the nature of the host rock, and the stoping method chosen for extraction. Properly designed block development is critical to the ongoing success of a stoping operation. Figure 3.32 shows the ore-handling flow from sublevel stoping in the 1100 orebody at Mount Isa Mines, with some key infrastructure being illustrated (Grant and DeKruijff, 2000). 3.4.5.1  Shaft Stability Vertical shafts are the most common type of access for deep underground orebodies. Shaft sinking and equipping is a specialized, complex procedure usually costing millions of dollars. Consequently, it is economically justifiable to spend a significant amount of time and money on shaft site selection and characterization. The rock mass investigations require geotechnical drilling to assess the presence of large-scale geological discontinuities, the hydrological regime, the nature and strength of jointing, and the physical properties of the rock types intersected. This is likely to indicate any potential instability problems during shaft sinking and the subsequent access maintenance. A shaft is sunk to a depth that will ensure many years of production during the life of a mine. Shaft location is controlled by the mining method used as well as the rock types present on a particular site. In sublevel stoping, the location of the shaft is usually at the footwall of the orebodies, where it is likely to be outside the influence of any ground disturbance caused by the stoping operations. In cases where the shaft is located within an orebody, a large amount of level development can be carried out within the orebody. However, a large amount of ore around the shaft must be left unmined as a shaft pillar (Figure 3.33). For example, the main and services supply shafts of the 1100 orebody at Mount Isa Mines have a shaft pillar that exceeds 200 m in diameter (Grant and DeKruijff, 2000). The design and monitoring of shaft pillars usually include the prediction of elastic/plastic strain profiles as a first-pass design, followed by physical monitoring of rock mass response to mining in order to identify displacement on preexisting geological discontinuities intersecting the shaft. A maximum

15 level conveyor Fill passes to stopes

CHF and HF lines Ground support

Development + mucking

23E Sub 6.2 Ore bin

Ore bin

Truck haulage

Production mucking

S60

5.2 C

/V

Loading flasks

No.3

Skips

Concentrator

22/L

21/L

20/L

19/L

Crude ore bin

U62 shaft Surface bin

Copper mine process flow diagram

lt

bo

ble

Ca

Crushers

ipples

No.4 T

Production and blasting superintendent

Drilling and services superintendent

Development superintendent

Copper planning team leader

Ore pass 19L Hallage

21C Sub

Ore pass

Note: Most stopes are Filled adjacent to at least Stope one filled stope

Production drilling

24D Sub

P41 Crusher

22D Sub

Spyder passes

Blasting

Ore

Exhaust vent, shafts V33, M37, L44, K48, I54 V33, Y37 water to death adder gully

Geological model and planning

Diamond drilling

Copper mine accountability

FIGURE 3.32 Schematic of ore process flow throughout the stoping cycle in the 1100 orebody, Mount Isa Mines. (From Grant, D. and DeKruijff, S., Mount Isa Mines—1100 orebody, 35 years on, in G. Chitombo, ed., Proceedings of the MassMin 2000, Brisbane, Queensland, Australia, October 29 to November 2, 2000, pp. 591–600, AusIMM, Melbourne, Victoria, Australia. With permission.)

Surface fill passes

Wet fill plant X41 Shaft Surface fill conveyor (ex. K.S.O.C.)

Planning and Design 79

80

Geotechnical Design for Sublevel Open Stoping

60

F

61

62

63

64

G H I J K

N643

M N

N645

Restricted mining area

L

O P

S

R62 supply and ore shaft No mining

6500 N

R

6000 N

Q

FIGURE 3.33 Plan view of no mining and restricted mining pillars around the R62 shaft complex at Mount Isa Mines. (Courtesy of Mount Isa Mines, Mount Isa, Queensland, Australia.)

strain criterion of ±150 με in any direction was historically established for shaft conveyance at the R62 shaft complex at Mount Isa Mines. Numerical modeling can be used to predict movement on regional faults intersecting a shaft complex. The change on the stoping geometries, the fault locations, and the history of shaft problems must be considered in the analysis. The results and predictions of the numerical analysis must be supported by the actual shaft inspection results. Regular direct shaft inspections coupled with kinematic shaft surveys can provide a baseline for monitoring the actual shaft deformation with time. Reusch and Beck (2007) have used results from nonlinear numerical modeling to compare plastic strain with measured shaft deflections. Their results are shown in Figure 3.34, where the simulated magnitude of the shaft deflection matches measured values with an error of less than 10%. The main deviation between model results and measurements occurs over a short section of the shaft, where some significant perturbations exist. The local change in shear strain usually corresponds to

81

Planning and Design

West 500

Shaft deflection (mm)

400

300

200

100

0

East –100 –200 250.0

WE WE FE

150.0 50.0

–150.0 –250.0 –350.0

Shaft depth (m)

–50.0

–450.0 –550.0 –650.0 –750.0 –850.0

FIGURE 3.34 Modeled and measured shaft deflections (left) and modeled plastic strain (right) due to sublevel open stoping. (From Reusch, F. and Beck, D., Simulating shaft and crusher damage in deep mines, in Y. Potvin, ed., Proceedings of the Fourth International Seminar on Deep and High Stress Mining, November 7–8, 2007, Australian Centre for Geomechanics, Perth, Western Australia, Australia, pp. 65–79.)

significant mechanical difficulties related to deflection of the rails or damage to shaft lining. The largest amount of the modeled plastic strain corresponds with an area of significant damage in the hangingwalls of the stopes in close proximity to the shaft (Reusch and Beck, 2007). 3.4.5.2  Ramp Access In some cases, major access to stoping blocks is provided by ramps, which are usually located within the footwalls of the orebodies (Figure 3.35). Access and trucking ramp systems are generally used, with major trucking ramps usually graded and designed with enough radius of curvature to preserve sight distance, enable maneuverability, and minimize tyre wear. Ideally, ramps are designed anticlockwise downward in order to provide optimum sight distance to left-hand drive (LHD) drivers, which must descend bucket first. Ramps must not lead directly into accesses to major mining excavations such as workshops, fueling bays, etc. The ramp dimensions are determined by the size of the mining equipment utilized. In particular, the design of ramp intersections with other roadways is important, as they must remain stable. Ramps may undergo high stress redistributions as the stopes are usually retreated toward

82

Geotechnical Design for Sublevel Open Stoping

N

Portal 1340RL

1240 mL L7 stope

1200 mL

O8 stope 1160 mL 1120 mL L8 stope

1080 mL

N8 stope

1040 mL 1000 mL 960 mL

H7 stope K7 stope

N7 stope

920 mL

FIGURE 3.35 Ramp access at the Mount Wright Mine, Queensland. (From De Vries, R., Sublevel shrinkage at the Mount Wright underground gold mine, MEngSc thesis, Western Australian School of Mines, Curtin University of Technology, Kalgoorlie, Western Australia, Australia, 2013, 89pp.)

crosscuts off a ramping system. The locations and geometries of the ramps must take into account factors such as the orebody geometry, the rock mass strength, and the stress loading as a result of the overall extraction sequence. Typical horizontal distances for ramp location from an orebody range from 50 to 70 m. 3.4.5.3  Crown Pillar In some cases, a major crown pillar is left in place to separate open pit and underground excavations within the same orebody (Figure 3.36). Conse­ quently, crown pillar stability is then critical to ensure safe underground extraction. The crown pillar dimensions and stability are a function of a

83

Planning and Design

Open pit extraction Crown pillar under open pit

–150 m

–250 m

–350 m

100 m

–600 m Planned delineation drillhole

FIGURE 3.36 Crown pillar at the Kundana Gold Mine, Western Australia.

number of parameters. The most important are the width of the orebody, the stress regime, the blasting practices, the rock mass strength within the pillar, the overall extraction sequence (top-down or bottom-up), and whether backfill will be introduced into the system. The actual crown pillar dimensions will depend upon the stress environment. Indications of high stress could include obvious signs of mininginduced stress fracturing or rock burst activity. High stresses may also be induced in otherwise low stress environments near the surface, due to the geometry of the orebody and the extraction ratio below and above the pillar. In addition, if a crown pillar is situated within a stress shadow environment, consideration must also be given to potential unraveling due to loss of clamping across the pillar. As a general rule of thumb, for narrow orebodies (90

80–90

70–80

60–70

50–60

40–60

30–40

FIGURE 3.42 (a) Extracted and scheduled stopes and (b) induced normal stress following extraction, Mount Isa Mines. (Courtesy of Mount Isa Mines, Mount Isa, Queensland, Australia.)

frequency of preparation, time periods, level of detail, format information, and communication process for mine scheduling may differ between mine sites (Trout, 1997). 3.4.6.1  Long-Term Production Scheduling Production scheduling is the highest level of scheduling and provides a longterm view of the mining process by focusing on issues such as ore grade, extraction sequences, and production quantities. Production schedules typically extend over a number of years and are expressed in terms of ore sources relating to stoping blocks (Trout, 1997). These schedules can extend through to the life of a mine, depending upon which event comes first. The items included in a scheduling exercise are long-term production targets, fill, development, raising, and diamond-drilling requirements. Annual estimates for equipment replacement, capital, and operating expenditure may also be determined. The most common restrictions imposed on scheduling may include capital availability, expected life of the mine, infrastructure, and equipment life. 3.4.6.2  Medium-Term Activity Schedules The second level of scheduling undertaken in underground mining is called medium-term activity scheduling. This schedule usually consists of a 2-year

91

Planning and Design

(or similar length of time) production period. Similarly to a long-term schedule, production targets, backfill, development, raising, and diamond drilling requirements are considered within this schedule. However, the activities are updated (using a rolling format) and issued every 3 months. Usually, a 1-year budget schedule is developed and adopted within a medium-term activity schedule. This full-year forecast is a critical document that sets the formal budget for the subsequent production year. The forecast is based on preliminary stope designs, in order to ensure that the budget metal, capital, and operating expenditure can be effectively achieved. Depending on the size of the mine and the number of ore sources, mine size and number of sources, the full-year forecast may be reviewed and updated each month. Priorities are then determined to ensure that the budget targets are met. 3.4.6.3  Short-Term Activity Schedules Short-term activity scheduling plays a tactical role while providing a detailed schedule over a short time horizon. The activity schedule contains sufficient details to allow underground personnel to plan and perform their work (Trout, 1997). Usually, this schedule considers the production activities within a 3-month period. It is updated and issued each month, primarily to assist production personnel in identifying the short-term activities (day-to-day mine operation) required to fulfill yearly budget targets. The short-term activity schedules are usually presented during a meeting between the planning and production personnel, where stope preparation (stope access development, ground support, services installations, stope drilling) and production issues (blasting, material handling, and filling) are discussed (Figure 3.43).

Stope production phase

Preparation phase

Extraction phase

Filling phase

FIGURE 3.43 A time-based representation of stope mining phases. (From Trout, P.L., Formulation and application of new underground mine scheduling models, PhD thesis, The University of Queensland, Brisbane, Queensland, Australia, 1997, 344pp.)

92

Geotechnical Design for Sublevel Open Stoping

3.4.7 Ventilation A mine ventilation system is related to the magnitude and direction of air movement through the various working places in the mine. The supply of air is referred to as air distribution, and it is accomplished by adopting a ventilation circuit suitable for the particular mining method used for extraction. In sublevel stoping, primary development openings such as shafts and ramps are used for main airways for ventilation, while the individual levels can be used as intakes and outlets using unidirectional air distribution. Sublevel stoping mines are likely to have extensive workings on each level, as well as between levels, and therefore require ventilation from combined vertical and horizontal circuits. The stopes are designed to allow flowthrough ventilation between the sublevels connected by the stopes. The overall objective is to supply fresh air to each level from a downcast pressure source, radiating outward and upward through the working places to exhaust airways leading to upcast shafts (Figure 3.44). In general, the airflow should be in an opposite direction to the stope retreat direction, so that dust and fumes are kept away from the operators. Consequently, the ventilation design for a stoping block will consist of access to fresh air, either from fresh air raises or a decline, as well as a return air exhaust system. The preferred approach is to ventilate each stope with a separate split of air, with the air introduced to the working places from the lowest level. Separate exhaust openings may be required to prevent contaminated air from entering other stopes in a stoping block. Ventilation shafts and airways must be located and maintained in ground which will not be caved and lost during the lifetime of the operation. In addition, short

Exhaust fan

Escape way Decline

VR

Development Stope

Intake Return Bulkhead Security door VR Vent raise Production Stope

Stope

Development FIGURE 3.44 Schematic of primary ventilation, Konkola deep mining project, Zambia. (From Calizaya, F., Schematic of a primary ventilation network, pers. commun., 2013.)

Planning and Design

93

circuitry and dust hazard created by air leakage up or down partially filled orepasses must be prevented. 3.4.8  Global Economic Assessment A number of global design considerations must be analyzed and economically evaluated to arrive at the optimum design for a stoping block. The outline of the orebody is determined by cutoff grade evaluations that account for the cost of block development, mining cost, haulage, surface cartage, mineral processing, and general overhead costs. A financial model is used to determine the viability by comparing the unit cost of all the steps involved in mining and processing with the estimated revenue. This could be an iterative process as, once the cost of development is included, some stoping blocks may prove not to be economical. However, they may become economic if development is carried out through those blocks to access other more economic areas. Thomas and Earl (1999) have described a computerized stope optimization tool that can be applied in the strategic planning of underground stopes. The technique can be used to generate an extraction sequence in conjunction with an optimum stope configuration that maximizes the net present value of an operation. The tool is used to generate inventories for a series of cutoff grades, and the results are scheduled to produce net present value (NPV) versus tonnage relationships.

3.5  Detailed Stope Design Detailed stope design relates to the extraction of individual stopes within a stoping block or global area. Detailed design is the process of establishing an optimum extraction method for an individual stope, subject to a number of variables and constraints. Blasthole geometry, firing sequence, ground support, ventilation, and economics are some of the key variables considered. The constraints include the orebody boundaries, the geological structures, any existing development and, in some cases, any adjacent fill masses (Figures 3.45 and 3.46). Figure 3.47 shows a typical process for taking an open stope from conceptual design through to production at the then WMC Resources, Australia (Teasdale, 2001). The detailed design process begins when the geological team undertakes detailed orebody delineation for a particular stope extraction. In-fill delineation drilling, mapping, sampling, and geological interpretations on a stope scale are then completed. The mine planning engineer uses geological sections from a mine design package to do a preliminary stope design, while the rock mechanics engineer completes a rock mass

94

Geotechnical Design for Sublevel Open Stoping

16B

16B

17D

18E

18E

18B

19C 19C

19A

19/L 19/L FIGURE 3.45 Isometric view of the P446 stope in the 1100 orebody, Mount Isa Mines. (From Grant, D. and De Kruijff, S., Mount Isa Mines—1100 orebody, 35 years on, in G. Chitombo, ed., Proceedings of the MassMin 2000, Brisbane, Queensland, Australia, October 29 to November 2, 2000, pp. 591–600, AusIMM, Melbourne, Victoria, Australia. With permission.)

characterization program, providing guidelines for stope stability, dilution control, reinforcement, and blast sequencing. At this stage, extraction factors that account for dilution as well as back analysis of performance from any adjacent stopes are taken into account. Drill and blast design is undertaken considering the equipment capabilities to ensure that the designed stope shape is achievable. This is then followed by an economic analysis that determines stope viability by considering the break-even revenue cutoff figures including a calculation of net revenue versus total mining, concentrating, and overhead cost. Finally, a stope design document that includes detail of the overall extraction philosophy, plans of sublevel development, sections showing blasthole design concepts and drilling and blasting parameters, ore- and waste-handling systems, ventilation, geology, rock mechanics, and overall firing sequence is issued to the operating personnel. All the topics included in a stope design document are interrelated. The extraction philosophy provides a general overview of the design, safety,

95

Planning and Design

4500 N

1800 E

Q450 filled

4480 XC Production rings

4457 XC

Cutoff slot

N Western cutoff slot

1800 E

DP T

2

4500 N

4450 N

4450 N Eastern cutoff slot

P442 filled (a)

(b)

4500 N

Q450 filled

Q450 filled

Cutoff slot

Production rings

1800 E

1800 E

4500 N

Production rings 4500 N

4500 N

P442 filled (c)

Cutoff slot

P442 filled (d)

FIGURE 3.46 Plan view of several sublevels through stope P446 showing drilling layouts and adjacent fill masses. (a) Extraction level, (b) mid height sublevel 18B, (c) mid height sublevel 18E, and (d) top sublevel 17D. (From Grant, D. and De Kruijff, S., Mount Isa Mines—1100 orebody, 35 years on, in G. Chitombo, ed., Proceedings of the MassMin 2000, Brisbane, Queensland, Australia, October 29 to November 2, 2000, pp. 591–600, AusIMM, Melbourne, Victoria, Australia. With permission.)

96

Geotechnical Design for Sublevel Open Stoping

Drilling and sampling

Kriging and wireframe

Preliminary design Final design Survey pickup

Development and ground support

Ring design

Face mapping, geological mark-up

Geological wireframe

Production drilling Blasting, mucking CMS survey Filling Reconciliation FIGURE 3.47 Typical process for open stope design, WMC Resources. (From Teasdale, P., Open stoping mining method of mining at WMC Resources Gold Business Unit operations, design process and operating practices, MEngSc thesis, Western Australian School of Mines, Curtin University of Technology, Kalgoorlie, Western Australia, Australia, 2001, 68pp.)

and production issues for a particular section of an orebody. Properly reinforced stope development is required to allow access for drilling, blasting, and mucking. Development size is a function of the stoping method and the equipment utilized. Development allows for the drill geometries to be designed, as well as subsequent ring firers’ access to charge the rings. Knowledge of the nature and stability of the adjacent fill masses is needed to design cleaner rings or to avoid toeing of blastholes into the fill. Geological considerations such as the presence of major geological discontinuities often influence the blasting sequences. Other factors considered are the stress redistributions within and around a stope that are likely to control falloff behavior on the exposed walls. In addition, the retreat direction of the blasthole rings must take into account the stope ventilation network, with a retreat direction into fresh air. Progress through a detailed design process can be tracked using a stope control sheet that can be used to track progress with preliminary design, production, and filled stopes (Figure 3.48).

97

Planning and Design

Stope Control Sheet Stope name: _______ Orebody: _______ Upper level: _______ Lower level: _______ Task

Responsible

Development completed

Development superintendent

Rock mass characterization completed

Rock mechanics engineer

Survey pick-up completed

Survey department

Wireframe geology

Geology department

Ring design completed

Mine planning department

Ring grade

Geology department

Update database

Mine planning department

Drilling completed

Production superintendent

Production completed Cavity monitoring completed

Production superintendent Survey department

Filling completed

Fill superintendent

Stope finished Update database

Mine planning department

Stope reconciliation note

Mine planning department

Update ore reserve long section

Geology department

Initial

Date

Comments: FIGURE 3.48 A stope control sheet developed at Mount Isa Mines. (Courtesy of Mount Isa Mines, Mount Isa, Queensland, Australia.)

3.5.1  Geological Information The typical geological information required for stope design consists of grade and tonnage, dilution factors, and delineation of the main geological features intersecting a stoping area. The initial information is usually collected from the diamond drillholes intersecting the area of interest. This information is used to create a conceptual three-dimensional orebody delineation design with ore tonnages and grades. Empirical extraction and dilution factors that account for the expected tonnage and dilution are issued prior to the preliminary economic analysis. A 110/96 factor indicates that up to 10% additional tonnes are expected from the stope. In addition, a 4% reduction on the grade is also expected. Major geological structures provide the greatest potential for large falloff in a stope void. Thus, information on the major geological structures anticipated can be used to delineate potentially unstable zones adjacent to an exposed stope wall. Unraveling and block release is possible along major structures, resulting in a zone of disturbance. In some cases, failure can

98

Geotechnical Design for Sublevel Open Stoping

17D

Overbreak

80 m

S48 fault zone

O434 stope

18B Stope design outline FIGURE 3.49 O434 stope hangingwall failure, Mount Isa Mines. (From Logan, A.S. et al., Geotechnical instrumentation and ground behavior at Mount Isa, in T. Szwedzicki, ed., Geotechnical Instrumentation and Monitoring in Open Pit and Underground Mining, Proceedings of the Australian Conference, Kalgoorlie, Western Australia, Australia, June 21–23, 1993, pp. 321–329, Balkema, Rotterdam, the Netherlands.)

progress beyond the weak zone itself. If a stope design requires blasting to the top of a fault or potential failure zone (by considering that such material has a high probability of failure), the design dilution factors are actually increased. However, the problems related to poor fragmentation from fault falloff may actually be minimized. In other cases, stope designs attempt to leave weak faults in place by leaving a beam of good-quality material against a fault or potential failure zone in order to improve the stability. An accurate assessment of fault location and knowledge of the likely behavior and deformational characteristics of the rock beam are required. The reduced fragmentation problems due to minimal falloff must be balanced against the ore loss occurring within the rock beam. A successful outcome during stope extraction involving a large weak zone is not always guaranteed, even by leaving ore beams, as shown in Figure 3.49. 3.5.2 Development The orebody characteristics and the type of equipment used are likely to influence the locations as well as the final sizes and shapes of the stope

Planning and Design

99

FIGURE 3.50 Development access prior to stope drilling at the Mount Marion Mine, Kalgoorlie, Western Australia.

development accesses. Geological control during development of the ore drives is required to minimize undercut and blast damage at the orebody boundaries. Geological mapping and orebody contact markup are undertaken at every development cut through a stope. This information is entered into a computerized database that can be used for orebody delineation purposes. Stope production drilling is facilitated when the development drill drives exhibit straight walls and good floor profiles (Figure 3.50). Development inside the stope, namely, cutoff slot and production blasting access, does not require such tight control as does the development located on the ore/waste contact, and so can be mined under survey control. The length of the development rounds must be compatible with orebody boundary variations along strike. Long round development may not be compatible with orebodies that pinch and swell along strike, as the chances of hangingwall and footwall undercutting may actually increase. In addition, the excavation size and shape must suit the equipment used during each task within the stoping cycle. Strike drives and crosscuts must take into account the dimensions and capabilities of development jumbos, longhole drilling, and production mucking equipment. When possible, stope drill drives are developed along the stope boundaries to limit the subsequent maximum drill hole length during production drilling within the stope. This may also

100

Geotechnical Design for Sublevel Open Stoping

prevent the blastholes toeing into stope footwalls or hangingwalls. However, depending on the orebody width, sometimes it is not always possible to locate twin drill drives at the boundary of a stope. 3.5.3  Geotechnical Assessment Geotechnical assessment for a stope design is carried out following completion of the strike and crosscut development within the stope limits. Geotechnical data can then be collected from direct mapping of the exposed walls within the stope development. These data are used to complement the initial data collected from core logging of the exploration diamond holes intersecting the rock mass within and around the stope. The data from mapping, logging, and the information on major geological discontinuities and rock type variations provided by the geologists are fundamental to the assessment of structurally controlled stope wall behavior. Figure 3.51 shows three major shear zones intersecting a planned stope design boundary. An initial interpretation from diamond drilling was confirmed with

ear

25A

he ar

W6

0 sh

Recrystallized shale

Ur

qu ha

rt s

25B

W 63

she

ar

Interpretation from mapping and diamond drilling 27C Interpretation from diamond drilling

28D

FIGURE 3.51 Interpretation of the main geological discontinuities on a stope scale, Mount Isa Mines. (Courtesy of Mount Isa Mines, Mount Isa, Queensland, Australia.)

101

Planning and Design

direct access mapping within the upper portion of the stope (25A and 26B sublevels). Experience with similar structures was used to predict a potential failure zone in the stope crown (Logan et al., 1993). The determination of stable stope wall dimensions is a critical aspect of the geotechnical assessment for a particular stoping area. Experience has shown that localized dilution as well as large block failures can be experienced in poorly dimensioned (very large) stope walls. On the other hand, designing for a worst-case geological scenario (small stopes) means that stope productivity may be unnecessarily affected throughout the operation. In most mines, the maximum stable stope wall length (or width) dimension is influenced by the height of the sublevel interval chosen. As the dimensions for the sublevel interval are systematically applied throughout a design block, considerations of stope wall stability are used to calculate the maximum permissible length or stope width for a particular stoping scenario. Stope wall dimensions become a very important economic parameter within individual stope design as they also control the size of the exposed spans at the stope crowns. The effect of external dilution (due to failures) or any unrecovered ore that must remain in pillars required to stabilize large spans is a key factor that requires consideration during stope size determination. The dimensions of a maximum stable length or width for a particular stope area are usually determined using local experience or an empirical rock mass classification system (Potvin et al., 1989). A geotechnical model of the maximum permissible stope lengths (or widths) for a fixed sublevel interval can be established for each particular stoping area (Figure 3.52). The model is based on geotechnical Hangingwall stability (55° dip)

100 90

Floor to floor

Permissible length (m)

80 70

50

12 15 22.5 30 45

40

Meters

60

30 20 10 0 3.5

4.0

4.5

5.0

5.5

6.0

6.5

7.0

7.5

8.0

8.5

9.0

Hydraulic radius (m) FIGURE 3.52 Maximum permissible stope length for different (floor-to-floor) sublevel intervals.

102

Geotechnical Design for Sublevel Open Stoping

mapping of stope development exposures and the localized stope delineation drilling. An iterative process is required to determine the optimum stope dimensions, as different degrees of stability may be predicted along and across the strike of an orebody (see Chapter 5). Geotechnical assessment also requires numerical modeling of the blasting sequences within the stope itself. This will determine the stress redistribution in and around the stope area. The likely damage due to stress changes, either compressive or tensile, in the stope brows, exposed hangingwalls, and adjacent pillars can then be predicted from back analysis in other similar areas. 3.5.4  Stope Design Philosophy Stope wall falloff and its subsequent influence on production efficiency is largely controlled by geology. Consequently, the stope design philosophy must consider the influence of any large geological feature intersecting a stoping area. The stopes must be designed to minimize falloff, rather than to maximize direct short-term cost savings by utilizing existing development that may force unfavorable positioning of blasthole rings, cutoff slots, retreat directions, and sequencing of the blast within the stope itself. Priority must be given to the analysis of the geological structures and their influence on production blasting and stope wall stability. Short-term savings on development may subsequently lead to poor fragmentation, falloff, and production losses several orders of magnitude higher than the savings on development. 3.5.4.1  Production Rings Ring blasting establishes the location of blastholes in relation to the drill drives, the orebody, and, most importantly, the planned stope outline. The designed volume and shape of the stope to be blasted as well as the positions and shapes of the drill drives and production mucking horizons are established for each ring section. Each individual ring design layout consists of a section through the orebody. The information presented in a ring design consists of the collar positions and the lengths and angles of the holes to be drilled and blasted. The hole size, the amount of explosives used on each hole, and the tonnes fired in each ring can also be indicated. In addition, the lengths of any uncharged collars on the holes are also provided. A plan view of the drill drives is used to determine the position of the cutoff slot in relation to the rings as well as the burden on the rings (Figure 3.53). A flexible blast design is one that allows the engineer a choice of single or multiple ring firings avoiding significant undercutting of stope areas. Blasting of the initial rings around the cutoff slot creates enough room for the remainder of the stope to be blasted. Considerations such as the level of the induced stresses and production and access constraint requirements are taken into account to determine the number of rings to be blasted together.

103

Planning and Design

Mount Isa Mines Limited, Mount Isa

16A sublevel

0.5m

Cross section view taken at ring 31 looking north at 6684.9 m Machine Hole diameter Explosive density kg/m ANFO Explosive density kg/m LD450 Explosive density kg/m LD425 Tonnes broken Task code Scale 1:250 Designed Drawn Checked Approved

Number of holes: 5/4 Footwall contact

Hangingwall contact

1.5 m 1.5 m

m

m

12. 0° 3m 64. 1 5° 2.8 m 65. 13. 0° 2m 65. 13. 0° 7m 65. 14. 0° 2m

64.

0.8 m

1.5

1.5

0.8 m

1.5

16B sublevel

Average orebody width: 9.3 m

m

Rings 29–33

R39 0.5 m

R38

2.0 m

0.8 m

L 4

L 3

R30 L 4

R29

L

0.5 m

Plan view drilling layout guide SIMBA 70 mm 4.30 2.15 1.58 820 2005

Notes: Dashed line represents the orebody outline All collar and breakthrough positions are relative to Orebody All holes are blow loaded Explosive densities shown are for blow loading Hangingwall hole to be loaded with LD450 12 Orebody Bench stope 12 C8

FIGURE 3.53 Typical section and plan view of drilling layout for bench stoping at Mount Isa Mines. (From Tucker, G. et al., Bench stoping at Mount Isa Mine, Mount Isa, Queensland, Proceedings of the 7th Underground Operators Conference, Townsville, Queensland, Australia, 30 June–3 July, 1998, pp. 135–147, AusIMM, Melbourne, Victoria, Australia. With permission.)

Important information such as the actual firing sequence, blasting results (fragmentation, freezing of holes, misfires, etc.), and any stope wall failures related to blasting must be recorded during ring blasting. 3.5.4.2  Diaphragm Rings Diaphragm rings are used where there is a moderate to high probability of fill exposure failure. Diaphragm ring design is complicated by issues such as different drilling and blasting techniques, different exposure sequences, varying stress regimes, and containment of anything from cemented to

104

Geotechnical Design for Sublevel Open Stoping

uncemented fill. Diaphragms are potentially unstable where undercutting of the diaphragm by the main rings is experienced. This may occur due to poor drilling resulting in hole deviation. In addition, failure may occur when a weak geological structure intersects a diaphragm in an unfavorable orientation or when extraction from previous stoping has damaged the rock mass within the diaphragm sufficiently to reduce stability. Another factor that assists diaphragm design is accurate knowledge of the backfill–rock interface. This knowledge would allow a proper determination to be made of the diaphragm thickness, in cases of uneven and sometimes overhanging fill masses. Stope surveys using the cavity monitoring system must be conducted following stope completion. However, stope wall falloff may still occur after the final stope survey, and probe drilling may be needed to accurately determine the actual rock–fill interface. 3.5.4.3  Cutoff Slot Design A cutoff slot is a very important element in a stope extraction sequence as it provides a free face and the required void for the rest of the stope to be blasted. The cutoff is created by the sequential enlargement of a long hole winze (LHW) geometry or a raise-bored opening. The decision to use one or the other is controlled by equipment availability, the height of the stope, the position of the existing development, and the desire to minimize damage from blasting. In multiple-lift sublevel stopes, existing development may be offset on alternative sublevels, making one straight raise-bored hole impossible to accommodate (Rosengren and Jones, 1992). LHWs are more flexible, but they limit the speed with which a stope can be brought into production. Stope ventilation requirements must also be considered as raise-bored holes improve the initial stope ventilation circuits. In some cases, a combination of raise-bored and LHW can be used within a single stope design. Figure 3.54 shows a stope design that incorporates an LHW at the western lower boundary (27C–28D) with cutoff holes retreating east. The top section of the stope was designed using a 1.8 m diameter raise-bored hole. Cutoff holes retreat west from 25A to 27C. 3.5.4.4  Drawpoint Design Production mucking can be carried out either longitudinally or transversely, across the strike of an orebody. Longitudinal mucking requires exposure of the loader under a retreating bench stope brow, while transverse mucking requires the use of fixed specialized drawpoint geometries that may be located outside an orebody boundary. In longitudinal mucking, the stability of a retreating brow is a function of the orebody width, the nature and strength of the geological discontinuities, the blasting practices, and the induced stresses. Mucking is carried out

105

Planning and Design

East

West

Ha

ng

ing

wa ll

25A level

Raisebore

26B level

Raisebore

27C level

LHW

28D level FIGURE 3.54 Cross section view showing a cutoff slot design parallel to a hangingwall to prevent blast damage. (Courtesy of Mount Isa Mines, Mount Isa, Queensland, Australia.)

by remote control because of the dangers of rock falls near the brow. The operator is positioned in a safe location just outside the stope and in constant visual contact with the loader. The loader is driven to the muck, loaded, and returned to the safe area under remote control. The operator then boards the loader and completes the mucking cycle under manual control. Longitudinal mucking of highly stressed, narrow, and long orebodies (where the muck is typically thrown up to 20 m away from the brow) is well suited to teleoperated mucking. Mucking is carried out from transportable, air-conditioned control stations that may be located up to 300 m from the stopes (Figure 3.55). The operator is seated on a comfortable chair that has joy stick control in each arm rest. Color video cameras are mounted at the front and rear of each loader and the images are continually transmitted back to the operator via a radio link. Teleoperated mucking provides improved

106

Geotechnical Design for Sublevel Open Stoping

FIGURE 3.55 A modern cabin for teleoperated mucking. (From McHugh, C., Introduction of autonomous loaders to Olympic dam operations, Australia, Proceedings of the Ninth Underground Operators Conference, Perth, Western Australia, Australia, 7–9 March, 2005, pp. 127–132, AusIMM, Melbourne, Victoria, Australia. With permission.)

occupational health for the operators, especially in highly stressed, seismic conditions. In addition, increased production rates, greater tolerance of hot and dusty conditions, and a reduced loader fleet have been accomplished with this method (Villaescusa et al., 1994). For transverse mucking, a number of factors must be considered during drawpoint design, including the size of equipment, tramming distances from access drives, as well as gradient and orientation with respect to a stope boundary. The drawpoint dimensions must be sufficient to suit the equipment, but kept as small as possible to minimize instability. 3.5.5  Stope Design Note A stope design note covers many factors involved in the development of, and production from, a stope (Table 3.3). Technical presentations are required to encourage technical input from all the members of the design team (geology, rock mechanics, planning, operations, and management). They usually occur twice within the design process, at the conceptual design stage and prior to the issue of the final drill and blast design. Feedback from both meetings should be incorporated into the final stope design. Once a final stope design status has been achieved, the blasthole design is undertaken by considering the production rigs that will be used, the ore limits, the survey pickup of the access development, the extent and sublevels of the stope, as well as the ring burden and toe spacing. The ore limits are usually updated in accordance with the completed stope development. A scaled floor plan showing details of the latest survey information including any vertical openings and status of surrounding stopes is provided to assist drilling. Locations of hangingwall, footwalls, cutoff slot detail, and locations of the production rings are also included (Figure 3.56). A long section that

Planning and Design

107

TABLE 3.3 Stope Design Presentation Issues Geological structures Stope access and development requirements Ore passes, loading bays, etc. Stope cutoff location Selection of drill rig and hole size Selection of explosive type Blasting sequence Stability issues, ground support requirements Stress redistributions assessment Fill requirements or permanent pillar demands Production schedule Ventilation requirements Detailed economic analysis

includes a schematic view of the stope cutoff raise, the cutoff slot, the production rings, and the trough undercuts is also completed. This section helps to explain the stope design philosophy, and becomes a useful tool during drilling and blasting of the stope. Table 3.4 lists a number of issues that should be considered during stope design. 3.5.6  Stope Firing Sequences The actual firing sequence used to extract individual stopes is likely to influence the stress redistribution as well as blast-induced damage within a stope. Stress and blast-induced falloff within a stope boundary may lead to poor mucking performance during extraction. Although falloff resulting from stope firing is not the only source of poor fragmentation, it can be minimized by avoiding excessive undercutting of the stope walls. Stope undercutting is usually linked to single-lift stopes (Figure 3.57). As a guideline, undercutting should not be undertaken when the stope is well advanced, and should never be attempted in poor ground where large-scale structures are present. Unfortunately, in single-lift stopes, where in most cases a cutoff slot is not available, undercutting is required by the method, regardless of the rock mass conditions. A number of design options can be used to reduce stope undercutting including firing the cutoff slot to the full height of the stope before blasting of the main rings commences. This can be followed by the sequential blasting of the main rings to the full stope height (Figure 3.58). The objective is to reduce the number of stope faces exposed, thereby reducing the potential for time-related structurally controlled falloff. Undercutting of the main rings can be avoided by designing the troughs to be blasted with coinciding faces.

108

Geotechnical Design for Sublevel Open Stoping

6750 XC

16B

16A

Bench limit 6730 N

Bench limit 6730 N

13C8 Sill drive

N

12C8 Sill drive

13C9 Sill drive

6700 N

12C9 Sill drive

11C9 Sill drive

6700 N

6650 N

6650 N

Bench limit 6620 N

6600 N

Bench limit 6620 N

6601 XC

6600 XC

Note: Bottom sill is shown to the left

Revision

6600 N

Mine design 12C8 bench stope Floor plan 16B-16A Scale 1:500

FIGURE 3.56 Floor plan of 12CB bench stope showing cutoff slot position and main rings, Mount Isa Mines. (From Tucker, G. et al., Bench stoping at Mount Isa Mine, Mount Isa, Queensland, Proceedings of the 7th Underground Operators Conference, Townsville, Queensland, Australia, 30 June–3 July, 1998, pp. 135–147, AusIMM, Melbourne, Victoria, Australia. With permission.)

Planning and Design

109

TABLE 3.4 Stope Design Checklist Location, orientation, and strength properties of large-scale geological structures Size of existing development and suitability for available drilling rig Additional development requirements, size, shape, and gradient Ground support requirements for development and stope walls Equipment needs for development including drilling, mucking, charging, and ground support Water drainage Tramming distances and alternate ore and waste passes Emergency escape routes during development and production Drill drive layout, blasthole design, and firing sequence Ring firers’ access to stope Drawpoint brow location and ground support requirements Ventilation requirements during development and stope production Bomb bays for storage of oversized rocks and secondary blasting Explosive types for development and production blasting Location, size, and orientation of pillars Overall rock mass (and fill mass) stability of the area prior to, during, and after stope extraction Detailed scheduling of stope development, production blasting, and filling Cost comparison of alternative designs Fill requirements including fill passes, reticulation, and delivery to stope Continuing stope performance monitoring during extraction Undertaking stope performance review after stope extraction

A stope firing sequence also determines the rate of exposure of the main geological discontinuities intersecting a stope. Rapid exposure of a large fault may occur after mass blasting or after progressive firing to a fault. Such exposures may not allow sufficient time for gradual stress relief. If the orientation of the stress field is unfavorable, large shear stresses may result inducing local and regional fault movements leading to stope falloff. 3.5.7  Production Monitoring Regular inspections of a producing stope are required, especially after each firing, in order to monitor wall, crown, and drawpoint conditions. Any significant rock noise, falloff, or underbreak should be documented. In addition, dilution exceeding more than 10% should be reported so that the actual stope grade can be adjusted accordingly. Geologists should conduct drawpoint investigations to estimate the grade of the ore being produced. Secondary blasting of oversized rocks and hung-up drawpoints may be required. In some cases, a bomb bay may be available for stockpiling oversized rocks and undertaking secondary blasting.

110

Geotechnical Design for Sublevel Open Stoping

33 2

1

2

FIGURE 3.57 Stope wall undercutting within a stope-firing sequence. 1–3 indicate blasting sequence for a single stope.

Broken ore is mucked conventionally when the drawpoints are full, but it is sometimes required to remote muck the last ore remaining on the floor of a stope, especially in large flat-bottomed stopes with retreating drawpoints. Significant disruptions to mucking productivity can occur when excessive delays are experienced during a stope extraction. Stopes left open over long periods of time may be influenced by time-dependent regional fault behavior. Stress redistribution, production blasting, and backfill drainage from adjacent stopes are likely to influence stope stability over a period of time. Blast damage and the effects of water from backfill can be transmitted along common fault structures intersecting a number of stopes. Instability may create difficult remote mucking conditions due to large-sized material falling off into the stope. These delays (stope production tails) actually extend the stope life, which in turn may contribute to more overbreak and more mucking delays. 3.5.8 Ventilation Stope ventilation is required during stope development and during stope production. Ventilation during development requires auxiliary fans that are used to force ventilate before a circuit is established. In steeply dipping orebodies with a single ramp access, the fresh air usually flows through the access ramp, where it can be force ventilated to the crosscut and ore drives using auxiliary fans. The fans are equipped with flexible ventilation

111

Planning and Design

Ring

blast

ing

Stop

e vo

id

17 level

18B sublevel Ring

Blasted

4

3-1-83

5 and 6 12-1-83 6 and 7 19-1-83 19C sublevel FIGURE 3.58 Full stope height blasting with matching trough undercut geometries to minimize undercut. (Courtesy of Mount Isa Mines, Mount Isa, Queensland, Australia.)

bags running along the crosscuts and the ore drives during development and production drilling. Once a stope raise is blasted and a ventilation circuit is established, the air is exhausted by means of specialized drives connected to return air raises. Stopes with interlevel connection are usually ventilated with air introduced in the lowest level and exhausted through the top level. 3.5.9  Financial Analysis The estimated cash per tonne of extraction reserves is calculated using the delineated mining reserve (tonnes and grade), the metal prices, and the extraction and dilution factors expected. The total cash profit (or loss) is determined using a proper ore value model suited to the particular economics of a mine site. The input factors may include tonnes mined, grades

112

Geotechnical Design for Sublevel Open Stoping

and metal prices, mining, milling, smelting, overheads and royalties, and exchanges rates. In periods of excess mining, hoisting, and milling capacity, the total net cash revenue can be increased by mining marginal stopes or marginal ore within stope boundaries. Marginal ore can be included within a stope design provided that little or no extra cost (no excessive extra development or additional reinforcement, etc.) will be incurred. An individual stope should be extracted if it can return a positive total net cash revenue after covering the costs of the remaining work required for extraction. Specific stopes may not break even but may be sufficiently advanced in terms of development and ground support to warrant a reduction in the break even value.

4 Rock Mass Characterization

4.1 Introduction A rock mass is a three-dimensional discontinuous medium that can be thought of as an assembly of potential blocks that can be disaggregated by the excavation process. The size distribution, shape, and degree of interlock of the blocks are functions of the distribution and nature of usually at least three main discontinuity sets. Rock masses are rarely uniform or isotropic; even within the confines of a design area, there are likely to be major geological structures, significant changes of lithology, and a prevailing anisotropy. The nature and degree of this anisotropy and the heterogeneity of the rock mass properties are likely to exert considerable influence on the extent of damage to, and dilution from, the final stope walls. During the last 35 years or so, a great deal of effort has been devoted to the characterization of discontinuity networks and to modeling them quantitatively (Call et al., 1976; Hudson and Priest, 1979; Villaescusa, 1991; Brzovic, 2010; Cepuritis, 2011a). Systematic collection of geotechnical information in conjunction with an appreciation of rock mechanics and geological factors are essential in planning and designing stable stopes. The structural data are initially utilized in the design of ground support configurations for the stope infrastructure, including access drives, crosscuts, and drawpoints. These data are used to develop an understanding of the various structural domains within an orebody, which can be used to predict the likely wall behavior during stope extraction. An optimized stope extraction sequence can be determined from this information. Some of the most important geological factors influencing a rock mass are shown schematically in Figure 4.1. The main features include the following: 1. Intact rock: This is the solid material between the discontinuities. Failure modes may involve failure of intact rock bridges. 2. Rock stress: The vertical stress caused by the weight of overlying strata and the horizontal stress caused by tectonic forces within the earth’s crust. 113

114

Geotechnical Design for Sublevel Open Stoping

Termination

Wall strength

Large discontinuity

Infill Waviness or planarity

g in

ac Sp Discontinuity

In situ stress

ce en ist rs Pe

Block size (intact rock)

σ1

σ3

σ2

Dip and dip direction

set

Water seepage FIGURE 4.1 Some of the major geological factors influencing the engineering behavior of a rock mass.

3. Number of discontinuity sets: A discontinuity is a mechanical break (of geological origin) within the rock mass. Because of geological process, discontinuities are formed in sets. In addition, a rock mass may be divided by single, large-scale geological discontinuity. 4. Discontinuity orientation: The three-dimensional attitude of a discontinuity in space is measured using dip direction (azimuth from north to the steepest line on the plane measured in a horizontal plane) and dip angle (the angle that the steepest line makes with the horizontal plane). 5. Discontinuity frequency and spacing: The frequency is the number of discontinuities per unit distance in space. It is the reciprocal of the spacing and can be defined globally for all discontinuity sets or by individual sets. 6. Discontinuity persistence and termination: Persistence is the observed trace length of a discontinuity within a rock mass. It provides a measure of areal extent or penetration for each discontinuity. Termination of a discontinuity can be either in intact rock or against another discontinuity. 7. Block shape and size: The shape and size of an intact rock block within a rock mass. The block size is a function of the number of sets, frequency, orientation, size, and termination of the geological discontinuities present within the rock mass. 8. Discontinuity roughness and planarity: Inherent surface roughness and planarity (or waviness) with respect to the naturally occurring mean

Rock Mass Characterization

115

plane defining a discontinuity. Both roughness and planarity contribute to shear strength. 9. Aperture: Perpendicular distance across adjacent walls of a discontinuity. 10. Wall strength: Compressive strength of adjacent walls of a discontinuity. Usually lower than the rock block strength due to alteration of the walls (by migrating fluids). Constitutes a key component of discontinuity shear strength if the walls are in contact. 11. Infill: Material that separates adjacent rock surfaces of a discontinuity. The material may be weaker (usually) or stronger than the adjacent rock walls. 12. Water seepage: Moisture or water flow within individual discontinuities or through intact rock. Some aspects of rock mass structure, strength, and stress can be measured by the logging of drill cores, directly by the structural mapping of exposed faces, or can be deduced from indirect measurements made using geophysical techniques. At most mining sites, conventional geological mapping is completed for all horizontally developed excavations, while geotechnical mapping is restricted to areas of specific concern where greater characterization of the rock mass is required. However, the largest amount of information in terms of areal coverage across an orebody is collected from diamond drilling during the several stages of the orebody delineation process. This process includes data collection from widely spaced surface drilling programs and any subsequent underground drilling for detailed stope design purposes.

4.2  Characterization from Exploration Core Diamond drilling, with geological core logging, is the most commonly used method for orebody delineation. Information obtained from drill intersections is extrapolated hole-to-hole using geological assumptions to provide estimates of lithological boundaries, alteration, weathering, hydrogeology, orebody size, shape, grades, continuity, tonnage, and some geotechnical characteristics (Figure 4.2). The advantages are the depth to which the information can be obtained, and a relatively routine data analysis and interpretation. Holes near the center of the mineralization provide critical information for stope design, while holes near the periphery are critical to the design of mine infrastructure such as shafts, access declines, and crusher chambers.

116

Geotechnical Design for Sublevel Open Stoping

FIGURE 4.2 Core details showing shear zones and faults intersecting orebodies at depth.

2200 RL

9000 E

Orebody boundary

8800 E

8600 E

2000 RL

HW intercept FW intercept

FIGURE 4.3 Longitudinal section view showing typical exploration drillholes.

Another advantage of geological logging is that characterization encompasses every drilled hole through a geological deposit (Figure 4.3). If some relevant geotechnical parameters are collected within this program, an extensive and representative database within and across the immediate boundaries of an orebody can be established. Parameters such as discontinuity linear frequency and rock mass classification data can be used to determine spatial variations in rock quality across an orebody. A perceived disadvantage is that a large number of individuals may perform the geological and geotechnical logging, introducing the chance of bias arising from different practices and interpretations. In addition, some of the drilling data

117

Rock Mass Characterization

may be collected from small-sized unoriented core that is not ideally suited for geotechnical logging. The approach suggested here is to carry out geotechnical logging on a number of selected holes within each exploration ring as part of the orebody delineation drilling program. The approach does not require oriented core to carry out the geotechnical logging, with the level of detail required during geotechnical investigations usually depending upon the stages of a particular project (mine prefeasibility, feasibility, etc.). Estimates of the likely stable stope sizes and shapes, dimensions of regional pillars, the best locations for underground infrastructure, and reinforcement schemes can be provided by such investigations. Figure 4.4 shows a cross section of a typical exploration ring where horizontal, steeply inclined, and steeply declined holes were logged systematically across the orebodies. Experience has shown that the choice of data format is important to facilitate the subsequent stages of the stope design process. In some cases, the computerized geological and geotechnical data are meshed as a threedimensional model. In some mines, such a geological/geotechnical model is not available, and the information is presented on paper plans/sections from which it can be digitized for printing purposes only. It is important that the initial geological model built is not only for grade control purposes but also intended for use in predicting the likely engineering performance of the excavations. The following sections describe a procedure that can be followed to carry out a rock mass characterization program from routine underground orebody delineation drilling.

Surface holes

Underground drilling

Core logged for geotechnical purposes FIGURE 4.4 Cross-sectional view showing typical underground exploration ring.

118

Geotechnical Design for Sublevel Open Stoping

4.2.1  Drilling Layout Design The drilling layout is based on information obtained from the initial surface delineation and a subsequent geological evaluation program. Section spacing and the distance between intersections down dip are determined based on local orebody complexity and the experience of the site geologists. An understanding of the associated risks incurred by the inability to interpret the orebody geometry and grade must be developed. Table 4.1 shows three stages of diamond drilling and the required confidence levels associated with each stage. Detailed stope design usually requires a 20 m average spacing between sections. With such a drilling spacing, the number of holes per stope is likely to be sufficient for an effective rock mass characterization process. Development access must be maintained ahead of production in order to have sufficient time and locations to undertake the proper orebody delineation. In most cases, development is maintained at least a year ahead of production, providing enough time to complete the task. 4.2.2  Underground Drilling Following the completion of the drill layout design, the holes are drilled from a suitable underground access (footwall access or ramps, see Figure 4.5). In most cases, the borehole diameter used during the underground delineation stages ranges from AQ to BQ (27–40.7 mm core, see Table 4.2). These hole sizes may not always be appropriate for the collection of geotechnical parameters without proper correction for the mechanical effects of drilling upon the core. 4.2.3  Core Transfer to Surface Following the completion of drilling, the recovered core must be transferred to surface to a core shed or similar, for logging. Significant damage to the TABLE 4.1 Typical Drill Spacing during Orebody Delineation Orebody Nature Tabular Stope Design Stage Feasibility Block design Detailed design

Structurally Complex

Drill Spacing (m × m)

Confidence (%)

Drill Spacing (m × m)

Confidence (%)

80 × 80 20 × 40 20 × 20

50 70 90

80 × 80 20 × 20 10 × (10 or 20)

50 80 90

Source: Teasdale, P., Open stoping mining method of mining at WMC Resources Gold Business Unit operations, design process and operating practices, MEngSc thesis, Western Australian School of Mines, Curtin University of Technology, Kalgoorlie, Western Australia, Australia, 2001, 68pp.

119

Rock Mass Characterization

FIGURE 4.5 Delineation drilling prior to detailed stope design.

TABLE 4.2 Nominal Hole and Core Diameters from Wireline Drilling Drill Size AQ BQ3 BQ LTK60 NQ3 NQ HQ3 HQ PQ3

Nominal Core Diameter (mm)

Nominal Hole Diameter (mm)

27 33.5 36.5 43 45.1 47.6 61.1 63.5 83.1

48 59.9 60 60 75.7 75.8 96.1 96.1 122.6

recovered core can occur at this stage due to mishandling of the core trays. This damage must be minimized so that an accurate estimation of the geotechnical parameters can be facilitated. 4.2.4  Drill Core Logging The geologists, geological technicians, or geotechnical engineers log the recovered drill core. The logged data are (manually or computerized) entered into a geological database. Mineralized zones are identified and prepared for assaying. The logged core is then photographed, split, bagged, and sent to

120

Geotechnical Design for Sublevel Open Stoping

FIGURE 4.6 Core splitting within an ore zone and immediate boundaries.

be assayed (Figure 4.6). If, at this stage, geotechnical logging has not been undertaken, critical information regarding the mechanical behavior of the rock mass will be lost permanently. Therefore, it is strongly recommended that laboratory assay data are obtained after all geological and geotechnical logging are completed. Geotechnical logging must be carried out over lengths of 1 m for at least the first 5–10 m immediately outside an orebody boundary. Logging intervals within an orebody are dependent upon the geological split defined by the general geological interpretations (Figure 4.7). Geotechnical logging must include interpretation and identification of major structures likely to form a discrete failure surface. 4.2.5  Geological Database Most mines have some type of database system, with data entry efficiency ranging from manual to highly computerized digital drillhole logging systems. Following the completion of data entry into a geological database, two-dimensional sections showing the geological logs and assays can be displayed on computer screens or paper plots. 4.2.6  Interpretation of the Orebody and Main Geological Features Based on the geologist’s experience and interpretative skills, orebody contacts and the main geological features such as faults, dikes, and shear zones are established between the boreholes on each section. This step is the

121

Rock Mass Characterization

DDH Orebody Logging each meter (for 5–10 m) DDH

Logging by geological split

Logging each meter (for 5–10 m)

Orebody Underground access FIGURE 4.7 Recommended core logging intervals across an orebody.

most important in terms of predicting the mechanical behavior of the stope walls and controlling the geological dilution (see Chapter 8). The quality of the information is important as additional data are not easily collected. The geologist must decide whether or not enough geological and assay information is available for an adequate interpretation of the mineralized zones and the main geological features. If a clear geological interpretation is not possible, then additional underground drilling may be undertaken (Figure 4.8). A second phase of underground delineation drilling (spacing between sections and down dip as close as 10 m) may be needed to facilitate the interpretations. Information from locations near the center of the orebody can be used for stope design, while information from the orebody periphery can be used to design infrastructure, such as shafts, ramps, and other related vertical infrastructure. 4.2.7  Orebody Meshing in Three Dimensions Once the orebody and the main geological features have been sufficiently sampled, the resulting shapes can be digitized in the two-dimensional paper sections. After that, the final three-dimensional orebody shape and volume can be established using computerized meshing tools. Computer manipulation and visualization of the meshed ore zones and controlling geological features can be used to establish geological and geotechnical models of a stoping block in three dimensions (Cepuritis, 2011a). 4.2.8  Problems with Data Analysis An orebody delineation process usually produces information that flows in a linear and sequential fashion. As the logging is undertaken, new information

122

Geotechnical Design for Sublevel Open Stoping

FIGURE 4.8 Two stages of drilling (global and detailed) for orebody delineation and characterization.

is being added to a geological database. However, very rarely a geological or geotechnical model, provided one actually exists, is centrally updated as soon as new information is obtained through data exchange between geology and mine planning. Relevant data required for long-term planning and detailed design may not be made available on time to the mine planning engineer. Data manipulation and visualization systems to update, access, retrieve, and display geological and geotechnical information with minimal effort are required (Cepuritis, 2011a).

4.3  Analysis of Logging Data 4.3.1  Discontinuity Linear Frequency Back analysis of unsupported hangingwall performance in open stoping carried out by Baczynski (1974) indicates that the number of discontinuities per meter within the first 3–5 m of a stope wall usually has a major control on the behavior of an exposed opening, including dilution control. The number of geological discontinuities is recorded for every split

123

Rock Mass Characterization

of core logged and then manipulated to determine the linear frequency per meter. In doing this, attention is required to identify and discount fractures caused by the drilling process or core handling. This process can be subjective, but most natural discontinuities have distinguishing characteristics such as mineral coating, while artificially broken core often has a rough, jagged appearance. Figure 4.9 shows a core logging sheet in which a common rock quality designation (RQD) data collection sheet has been modified to include information on discontinuity linear frequency. The rock mass class ranges have been established by back analysis of unsupported stope spans at Mount Isa Mines (Baczynski, 1974; Villaescusa et al., 1992). Discontinuity linear frequency is defined as the number of geological discontinuities per meter of a borehole through the rock mass. In three dimensions, the linear value depends on the orientation of the line with respect to the structural discontinuity network. The linear frequency can be calculated for a single joint set or a number of combined sets since the total number of joints encountered along a line is additive. It is calculated from



lL =

n T (4.1) LT

where nT is the total number of discontinuities intersected by a borehole of total length LT. Hudson and Priest (1983) have established that variation in the discontinuity linear frequency value, λL, when calculated in different directions in space, is a function of the existence of any anisotropies or preferred discontinuity orientations. The discontinuity frequency within the first 5–10 m adjacent to a stope hangingwall or footwall is calculated for each logged hole. The data for all the holes intersecting an orebody can then be interpolated and represented as a contour plot on a longitudinal section view (Figure 4.10). Interpolation techniques such as kriging can also be used to display discontinuity frequency data, which can be used to predict ground behavior following a core logging program. A kriged model of discontinuity frequency for an orebody can be produced using equivalent kriging weights. Figure 4.11 shows variograms of bedding plane frequency calculated from the closely spaced drilling fans in some of the orebodies at the George Fisher Mine in Mount Isa. A strong anisotropy ratio (across versus along bedding) was found for these orebodies, and the equivalent kriging weights used were based on a strong 9:1 anisotropy ratio. The advantage of this type of analysis is that estimates of ground behavior for an entire deposit can be made using a number of commercially available software packages. The estimated conditions can be predicted using geostatistical block model data and displayed on cross sections or plan views (Figure 4.12).

m

m

m

Fair 12

Poor 17

Discontinuity linear frequency

Very Exc Good 1.5good 4 7

FIGURE 4.9 Discontinuity frequency and RQD logging sheet.

Depth (m)

Lithologic log

Lithologic, Linear Frequency and RQD Logs– Project Name Very poor

Rock quality designation

Date

Page

Very poor Poor Fair Good Exc 10% 20% 30% 40% 50% 60% 70% 80% 90%

By

124 Geotechnical Design for Sublevel Open Stoping

125

Rock Mass Characterization

2

17

1

15.3 13.6 11.9 10.2 8.5 6.8 5.1 3.4 1.7 0

FF/m

FIGURE 4.10 Longitudinal section view showing contours of discontinuity linear frequency and large scale geological discontinuities for the first 5 m into an orebody hangingwall.

Variogram/(mean + 4.70)2

0.28

Cross bedding

0.24 0.20

Along bedding vertical

0.16

Along bedding horizontal

0.12 0.08 0.04 0.00

0

10

20

30

40

50

60

70

80

90

100

Distance (m) FIGURE 4.11 Relative variograms of bedding plane frequency.

4.3.2  Rock Quality Designation The concept of RQD as described by Deere (1964) is a quantitative index based on core recovery in which the measure of the quality of the core is determined incorporating only those pieces of intact sound core greater than a threshold value tc in length. This value is generally twice the core diameter

126

Geotechnical Design for Sublevel Open Stoping

Color

Condition Very good Fair Very poor

Disc/meter 1.5–4 7–12 >17

Color

Condition Disc/meter Good 4–7 12–17 Poor

FIGURE 4.12 Geostatistical display of bedding plane frequency for a number of orebodies (grid is 100 m × 100 m).

dimensions, and shorter lengths of core are usually ignored. RQD can be formally defined as



RQD = 100

n

ÂL i =1

x i (4.2) T

where xi is the length of ith length greater than the threshold value tc n is the number of such intact lengths greater than tc LT is the length of the borehole along which the RQD is calculated The original concept of RQD was based on data from NQ size core (Table 4.2) with tc = 100 mm. However, core from underground exploration drilling is typically smaller than NQ, and such core is likely to be more sensitive to drilling and handling conditions than larger diameter core. Consequently, the threshold value used in the evaluation of RQD should reflect the sensitivity to core diameter. Although the global statistics and distributional nature of the RQD values per stope surface (hangingwall, orebody, or footwall) can be determined (Figure 4.13), the mean values may not be relevant for individual stope designs. The spatial variability of the RQD values needs to be taken into

127

Rock Mass Characterization

N = 719 X = 83 SD = 19.5 Max = 100 75% = 97 50% = 91 25% = 79 Min = 0

25

Observed frequency

20

15

10

5

0

0

15

30

45

60

75

90

100

Rock quality designation (%) FIGURE 4.13 Histogram of RQD values for a hangingwall surface.

account when designing individual stopes. As shown in Figure 4.14, stope outlines can be superimposed on RQD contours, allowing local RQD values to be determined and used for design. The data for Figures 4.13 and 4.14 were collected for the first 5 m surface of the hangingwall at the Kanowna Belle Gold Mine, Western Australia. Priest and Hudson (1976) formulated the theoretical RQD as an integration of the probability density function of discontinuity spacing. If the spacings are negative exponentially distributed along a borehole axis, the RQD values can be approximated by

RQD t = 100 e(

-lL t c ){lL t c +1}

(4.3)

where λL is the linear frequency along a borehole of total length LT tc is the threshold value Equation 4.3 provides a theoretical link between RQD and the linear frequency, and provided the global spacing is negative exponentially distributed, it can be used to give a reasonable estimation of the actual RQD values. A practical alternative is to use the empirical correlation between linear frequency and RQD initially suggested by Baczynski (1980). Following a similar approach and based on the data collected using the logging sheet in Figure 4.9, the linear frequency and the RQD values can be calculated for

128

Geotechnical Design for Sublevel Open Stoping

100 90

Block A 10000 N

80

Block B

70 60

9800 N

50

Block C

40 9600 N

30 20

20400 E

RQD

20200 E

9400 N

19800 E

19000 E

0

20000 E

Block D

10

FIGURE 4.14 Longitudinal section view of the Kanowna Belle Mine showing contours of RQD values and individual stope outlines (grid is 200 m × 200 m).

each meter split of core logged along the axis of a borehole. Significant relationships can be found between the data sets using linear, polynomial, and logarithmic fits. Figure 4.15 shows an empirical best fit using a polynomial fit to a typical set of data as follows:

2 RQD = 100 - 6 {lL }+ 0.08 {lL } (4.4)

where RQD is the calculated rock quality designation λL is the observed linear frequency along the borehole axis As noted earlier, the initial guidelines for RQD calculation developed by Deere (1964) were based on core logging of NQ diameter core. Experience indicates that the larger the diameter of the core, the less likely the influence of drilling damage on core fractures. RQD values calculated from smalldiameter core typically used underground may be affected by the mechanical disturbance from drilling and effectively underestimate the actual ground conditions. Therefore, careful consideration must be given when using small-diameter core to calculate RQD values. An example of logged data from two closely spaced holes on the same area (1.5 m apart) at Mount

129

Rock Mass Characterization

100 90

RQD = 100 – 6LF + 0.08 (LF)2

80

RQD

70 60 50 40 30 20 10 0

0

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 Discontinuity linear frequency (J/m)

FIGURE 4.15 An empirical relationship between discontinuity linear frequency and RQD.

Isa Mines indicated that on average, up to two additional breaks per meter can be expected when BQTK rather than LTK60 drilling is used as shown in Figure 4.16. The effect of the core size is evident as an increased discontinuity frequency along the smaller-diameter core axis. Figure 4.17 shows two empirical relationships between RQD and linear frequency calculated from different core sizes and orientations at the New Celebration Mine, Western Australia (Cepuritis, 1987).

Bedding plane breaks/meter

25 LTK60 (45 mm)

Very poor

BQTK (38 mm) 17

Poor 12 Fair Good

7

Very 4 good 1.5 Exc 0

0

10

20

30

40

50

60

Drillhole depth (m) FIGURE 4.16 Comparison of observed linear frequency with different hole sizes.

70

80

130

Geotechnical Design for Sublevel Open Stoping

HQ3 Core (63.5 mm) — Hole HBG_1559_1 (50/64) 100

RQD = 100 –3.78LF + 0.011LF2

R2 = 0.7603

RQD (%)

80

60

40

20

0

0

2

4

6

8

10

12

14

16

18

20

Discontinuity linear frequency (J/m) LTK 56 Core (45.6 mm) — Hole UHS_3302_1 (–30/077) 100

RQD = 100 –3.081LF + 0.009LF2

R2 = 0.684

RQD (%)

80

60

40

20

0

0

2

4

6

8

10

12

14

16

18

20

Discontinuity linear frequency (J/m) FIGURE 4.17 Empirical relationship between RQD and linear frequency for two core sizes. (From Cepuritis, P.M., Hampton boulder haulage shaft geotechnical study. MEngSc thesis, Western Australian School of Mines, Curtin University of Technology, Perth, Western Australia, Australia, 1987.)

Rock Mass Characterization

131

Figure 4.18 shows a comparison of RQD and linear frequency for the hangingwall of the Kanowna Belle Mine, where both parameters were calculated over identical 1 m intervals down the hole. The parameters were calculated using a 10 m × 10 m grid model overlain on a solid model defined by the first 5 m into the orebody hangingwall, where the parameters were calculated. The inverse distance square method was used for interpolation. The data suggest that both methods predict similar variability of the rock mass conditions at the Kanowna Belle open stoping mine. 4.3.3  Rock Mass Classifications from Core Logging Rock mass classification systems are used in mining and civil engineering applications to characterize the rock mass and to determine maximum unsupported spans, support, and reinforcement requirements and estimated rock mass strengths. They are empirical methods that have been developed from the back analysis of excavation performance. Where multiple joint sets with differing discontinuity characteristics are present in a rock mass, a decision must be made as to the most important discontinuity set that is likely to control rock mass behavior and potential failure. A program of data collection for rock mass classification purposes can also take place when the geologists are logging the core for orebody delineation purposes. A number of conventional classification systems can be considered for rock mass characterization, the most widely used in the Australian mining industry being the Q (Barton et al., 1974) and the RMR (Bieniawski, 1976, 1989) systems. Basically all the holes can be logged in a manner consistent with obtaining information on discontinuity linear frequency and the information required to carry out a rock mass classification. For each split, the data collected may include the number of discontinuities per split, joint condition, joint set number, joint roughness, joint alteration, point load strength, and the position of faults. The rock mass classification systems (Tables 4.3 and 4.4) have been extensively described elsewhere (e.g., Hoek et al., 1995) and so are not described in detail within this book. However, a number of guidelines are presented to enable logging of rock mass classification parameters using standard exploration core. The individual parameters for RMR and Q using conventional exploration core can be estimated as follows: 1. UCS can be determined from standard uniaxial compression or point load testing of selected core samples. 2. RQD can be determined using the total number of discontinuities per meter or split to calculate the discontinuity linear frequency. The actual estimate of RQD can be obtained by evaluating Equation 4.3 or a locally established empirical relationship using core data as shown in Figure 4.15.

RQD

9400 N

9600 N

9800 N

10000 N

Block D

Block C

Block B

Block A

0 FF/m

1.7

3.4

5.1

6.8

8.5

10.2

11.9

13.6

15.3

17

9400 N

9600 N

9800 N

10000 N

19800 E

20400 E

20200 E

20000 E

19800 E

FIGURE 4.18 A comparison of RQD and discontinuity linear frequency contours for the same orebody (grid is 200 m × 200 m).

0

10

20

30

40

50

60

70

80

90

19000 E

100

19000 E

Block D

Block C

Block B

Block A

132 Geotechnical Design for Sublevel Open Stoping

20400 E

20200 E

20000 E

Strength of intact rock material

Drill core quality RQD

Spacing of discontinuities

Condition of discontinuities

1

2

3

4

Parameter

Rating

Rating

30

25

12 75–90 17 0.6–2 m 15 Slightly rough surfaces, separation 2 m 20 Very rough surfaces not continuous, no separation, unweathered wall rock

Rating

100–250

>250

Uniaxial compressive strength (UCS) (MPa) Rating

4–10

>10

Point load strength index (MPa)

Rock Mass Rating (RMR) System

TABLE 4.3

20

7 50–75 13 200–600 mm 10 Slightly rough surfaces, separation D/2 (a) L

P

W

L

De

D

P

W1 D

De

W2 L > D/2

(c)

0.3W < D < W

(b)

0.3W < D < W W = (W1 + W2)/2 (d)

FIGURE 4.51 Specimen shape requirements for (a) diametral test, (b) axial test, (c) block test, and (d) irregular lump. (From Brown, E.T., ed., Rock Characterization, Testing and Monitoring—ISRM Suggested Methods, Pergamon, Oxford, U.K., 1981, 211pp.; Int. J. Rock Mech. Min. Sci. Geomech., 22(2), ISRM, Suggested method for determining point load strength, 51–60, Copyright 1985, with permission from Elsevier.)

175

Rock Mass Characterization

500

UCS = 20.103Is(50) R2 = 0.4203

UCS (MPa)

400 300 200 100 0

0

5

10

15

20

25

Point load index—Is(50) (MPa) FIGURE 4.52 UCS versus point load index (Is(50)).

higher point load strengths, since those specimens are less likely to contain preexisting flaws, and a correction factor is required as follows (Brown, 1981; ISRM, 1985; Ulusay and Hudson, 2007): 0.45



ÊD ˆ I s(50 ) = Á e ˜ Ë 50 ¯

(4.13)

Although theoretical considerations show that Is provides a measure of tensile strength, the experimental results show that Is is also sufficiently related to σc as shown in Figure 4.52. The data in the figure were determined by calculating pairs of σc and Is from adjacent pieces of core for a large number of deposits and host rock masses in Australia. On average, σc is about 20 times Is(50), which agrees with the multiplication factor of 20–25 suggested by Broch and Franklin (1972) and the ISRM (1985). In some rocks, different Is(50) values are obtained when the core sample is loaded axially or diametrally. 4.6.3  Confined Compressive Strength As noted in the opening of Section 4.6, triaxial compression testing is one of the commonly used mechanical property tests for intact rock. This test is carried out on cylindrical samples subjected to a range of uniform allround confining pressures and loaded in axial compression (Figure 4.53). Discussion of the test procedures, the factors influencing rock behavior in

176

Geotechnical Design for Sublevel Open Stoping

σ1

Platens

σn

σ3

α

e

lan

τn

p re

ilu

Fa

σ3

Specimen

σ1 FIGURE 4.53 Schematic of triaxial compressive testing showing failure plane.

these tests, and the interpretation of the test results is beyond the scope of this book. Discussions of these types are given by Brady and Brown (2004) and Hoek (2007), for example. Figure 4.54 shows the complete axial stress (σa)–axial strain (εa) curves obtained by Wawersik and Fairhurst (1970) in a series of triaxial compression tests carried out on a marble. These and similar data for other rocks show that, with increasing confining pressure, the following results are obtained: a. The peak strength increases. b. A transition occurs from typically brittle (with a postpeak reduction in strength) to fully ductile (continued deformation at constant differential stress) behavior with the introduction of plastic mechanisms of deformation including cataclastic flow and grain-sliding effects. c. The region incorporating the peak of the σa–εa curve flattens and widens. d. The postpeak drop in stress to the residual strength reduces and disappears at high values of the confining pressure, σ3.

177

Rock Mass Characterization

Increasing confining pressure (MPa)

300

48.3

Axial stress (MPa)

34.5 27.6

200

20.7

13.8 100

0

6.9 3.4 0 0

0.10

0.20

0.30

0.40

0.50

0.60

0.70

Axial strain, εa (%) FIGURE 4.54 Complete axial stress–axial strain curves obtained in triaxial compression tests on Tennessee marble at the confining pressures indicated by the numbers on the curves. (After Int. J. Rock Mech. Min. Sci., 7(5), Wawersik, W.R. and Fairhurst, C., A study of brittle rock fracture in laboratory compression tests, 561–575, Copyright 1970, with permission from Elsevier.)

σ1

τn ϕ

ψ σc

c 2β (a)

σ3

σ1

σn

(b)

σ3

FIGURE 4.55 Mohr–Coulomb peak strength envelopes in terms of (a) shear and normal stresses, and (b) principal stresses. (From Brady, B.H.G. and Brown, E.T., Rock Mechanics for Underground Mining, 3rd edn., Kluwer, Dordrecht, the Netherlands, 2004, 628pp.)

As illustrated in Figure 4.55, the peak axial stress (σ1) reached at each value of confining pressure (σ3) in a series of triaxial compression tests may be plotted as Mohr’s circles of stress on shear stress (τn)–normal stress (σn) axes (Figure 4.55a), or as plots of σ1 against σ3 (Figure 4.55b). The resulting peak strength envelopes for intact rock are customarily curved and may be described by the nonlinear Hoek–Brown empirical strength equation to be introduced in Section 4.7.2 or by other empirical criteria (Brady and

178

Geotechnical Design for Sublevel Open Stoping

Brown, 2004). However, for many rocks, particularly over limited ranges of the stresses, σn and σ3, they may be approximated closely by straight lines. As shown in Figure 4.55a, the straight line peak strength envelope on τn–σn axes is a representation of the classical Coulomb (often referred to as the Mohr–Coulomb) shear strength criterion:

tn = c + s n tan j (4.14)

where c is the cohesion φ is the angle of internal friction Note that the intercept of the principal stress envelope on the σ1 axis (Figure 4.55b) gives the UCS, σc. The slope of the σ1 versus σ3 envelope, ψ, is a function of the angle of internal friction, φ, as follows:



tan y =

1 + sin j (4.15) 1 - sin j

The UCS, σc, is related to cohesion, c, and the angle of internal friction, φ, as follows (Brady and Brown, 2004):



sc =

2c cos j (4.16) 1 - sin j

The Mohr–Coulomb criterion is also used as the basis of a range of expressions used to describe the shear strengths of smooth, rough, and infilled discontinuities in rock (Brady and Brown, 2004).

4.7  Mechanical Properties of Rock Masses In analyzing geotechnical problems encountered in the design of open stoping layouts and sequences, often by using numerical modeling, it is necessary to estimate the mechanical properties of the rock mass, usually represented by its stress–strain behavior. Important aspects of this behavior are the constants relating stresses and strains in the elastic range, the stress levels at which yield, fracturing, or slip occurs within the rock mass, and the postpeak stress–strain behavior of the fractured or failed

Rock Mass Characterization

179

rock (Brady and Brown, 2004). The collection of data for use in estimating some of these properties is part of the rock mass characterization process discussed in this chapter. In some problems, it is the behavior of the intact rock material discussed in Section 4.6 that will be of concern. This will be the case when considering the excavation of rock by drilling and blasting (to be discussed in Chapter 6) or when considering the stability of excavations in good quality brittle rock. In other cases, the behavior of single discontinuities or of small numbers of discontinuities may be of paramount importance. This class of problem includes the equilibrium of blocks of rock formed by the intersection of three or more discontinuities with the roof or wall of an excavation, and cases in which slip on a fault must be considered. A different class of problem is that in which the rock mass must be analyzed as an assembly of discrete blocks as discussed in Section 3.3. In this case, the normal and shear force–displacement relations at face-toface and corner-to-face block contacts are of importance in the analysis. Finally, it is sometimes necessary to consider the overall response of a jointed rock mass in which the discontinuity spacing is small on the scale of the problem domain and the rock mass can be treated as an equivalent continuum having isotropic material properties. The remainder of Section 4.7 will consider the strength and deformability of rock masses in these circumstances. 4.7.1  Hoek–Brown Empirical Strength Criterion In an attempt to provide a first-pass method of estimating the strength of jointed rock masses for use in underground excavation design, Hoek and Brown (1980) developed an empirical rock mass strength criterion based on their earlier work on the brittle fracture of rock and the mechanical behavior of discontinuous rock masses. The criterion took the strength of the intact rock as the starting point and introduced factors to reduce the strength on the basis of the spacing and characteristics of the joints within the rock mass. Initially, Hoek and Brown (1980) used the 1976 version of Bieniawski’s RMR (see Section 4.2.3) as an index of the geological characteristics considered likely to influence the mechanical properties of the rock mass. Because of a lack of suitable alternatives, the Hoek–Brown criterion was soon adopted by rock mechanics practitioners and sometimes used for purposes for which it was not originally intended and which lay outside the limits of the data and methods used in its derivation. Because of this, and as experience was acquired with its practical application, a series of changes were made and new elements were introduced into the criterion (e.g., Hoek and Brown, 1997). The current version of the criterion is that given by Hoek et al. (2002) and discussed by Hoek (2007).

180

Geotechnical Design for Sublevel Open Stoping

The generalized Hoek–Brown empirical strength criterion for jointed rock masses is given by a



È Ês ˆ ˘ s1 = s 3 + sci Ím b Á 3 ˜+ s ˙ (4.17) Î Ë sc ¯ ˚

where σ1 and σ3 are the major and minor principal stresses at peak strength σci is the UCS of the intact rock mb is a parameter that reflects the frictional strength of the rock mass s is a parameter that reflects the cohesive strength of the rock mass and depends on the rock mass quality as does the index a, which takes a value of close to 0.5 for hard, fresh rock For intact rock, s = 1.0. The values of mb, s, and a are related to the GSI of the rock mass by the following relations:

m b = m ieGSI-100 28-14D (4.18)



s = eGSI-100 9-3D (4.19)

and

a = 0.5 +

e-GSI 15 - e-20 3 (4.20) 6

where mi is a strength parameter for the intact rock (Figure 4.56) D is disturbance factor that varies with the degree of disturbance due to blast damage and stress relaxation D varies from 0 for undisturbed in situ rock masses to 1.0 for very disturbed rock masses. The rock material parameter, m i, is obtained by the statistical analysis of a set of triaxial compression tests on carefully prepared 50 mm diameter core samples of the intact rock. If it is not possible to carry out a set of triaxial tests, m i may be estimated as σci/T where T is the uniaxial tensile strength of the intact rock (Brown, 2007). Because of potential differences in the failure mode, the value of the UCS estimated from the intercept of the peak strength envelope with the σ1 axis as shown in Figure 4.55b

181

Rock Mass Characterization

Rock Type

Class

Group

Texture

Coarse a

Conglomerates

Brecciasa

Sedimentary

Clastic

Carbonates Nonclastic

Crystalline limestone (12 ± 3)

Evaporates

Medium

Fine

Sandstones (17 ± 4)

Siltstones (4 ± 2) Graywackes (18 ± 3)

Sparitic limestone (10 ± 2) Gypsum (8 ± 2)

Micritic limestone (9 ± 2) Anhydrate (12 ± 2)

Metamorphic

Organic Marble (9 ± 3)

Nonfoliated Slightly foliated Foliatedb Light

Igneous

Plutonic Dark Hypabyssal

Volcanic

Lava Pyroclastic

Hornfels Quartzites (19 ± 4) (20 ± 3) Metasandstone (19 ± 3) Migmatite Amphibolites Gneiss (29 ± 3) (26 ± 6) (28 ± 5) Phyllites Schists (7 ± 3) (12 ± 3) Granite Diorite (32 ± 3) (25 ± 5) Granodiorite (29 ± 3) Gabro Dolerite (27 ± 3) (16 ± 5) Norite (20 ± 5) Porphyries Diabase (20 ± 5) (15 ± 5) Dacite Rhyolite (25 ± 3) (25 ± 5) Andesite Basalt (25 ± 5) (25 ± 5) Agglomerate Breccia Tuff (19 ± 3) (19 ± 5) (13 ± 5)

Very Fine Claytones (7 ± 2) Shales (6 ± 2) Marls (7 ± 2) Dolomites (9 ± 3)

Chalk (7 ± 2)

Slates (7 ± 4)

Peridotite (25 ± 5)

FIGURE 4.56 Values of the constant m i for intact rock by rock group. (From Hoek, E. et al., Support of Underground Excavations in Hard Rock, Balkema, Rotterdam, the Netherlands, 1995.) a Conglomerates and breccias may present a wide range of m values, depending on the nature i of the cementing material and the degree of cementation, so they may range from values similar to sandstone to values used for fine grained sediments (even under 10). b These values are for intact rock specimens tested normal to bedding or foliation. The values of mi will be significantly different if failure occurs along a weakness plane.

182

Geotechnical Design for Sublevel Open Stoping

may differ from the mean value of UCS obtained from a series of uniaxial compression tests as discussed in Section 4.6.1. It is the value of UCS obtained by extrapolating the peak strength envelope back to σ3 = 0, represented by the symbol, σci, that should be used in the Hoek–Brown criterion (Equation 4.17). 4.7.2  Rock Mass Deformation Modulus As noted in Section 4.7.1, stress and deformation analyses of the responses of rock masses to the creation of mining excavations within them require the input of a range of parameters describing their stress–strain behavior. In the case being considered here in which a jointed rock mass may be represented as an isotropic equivalent continuum, the main parameter required is the deformation modulus, Em. Over the years, a wide range of methods of estimating Em for different purposes have been proposed in the literature. In the main, these methods use some measure of rock mass quality such as joint spacing, RQD, RMR, Q, or GSI to give empirical estimates of the rock mass modulus, Em, sometimes by reducing the modulus of the intact rock, Ei (e.g., Bieniawski, 1976; Serafim and Periera, 1983; Barton, 2002; Zhang and Einstein, 2004; Hoek and Diederichs, 2006). Figure 4.57 shows plots of a range of measured values of Em fitted by the equations based on RMR proposed by Bieniawski (1978) and by Serafim and Periera (1983), and an equation based on Qc = Q σc/100 proposed by Barton (2002). As Figure 4.57 shows, these empirical equations generally give unrealistically high estimates of rock mass modulus at high values of RMR or Q, in some cases being asymptotic to infinity as RMR approaches 100. Hoek and Diederichs (2006) evaluated a wider range of field measurements of rock mass deformation modulus and fitted them by a sigmoidal relation to overcome the problem of exponentially increasing values of Em at high values of RMR, Q, or GSI. The expression developed by Hoek and Diederichs (2006) is



È 1 - (D 2 ) ˘ (4.21) Em = Ei Í0.02 + 60 +15 D-GSI ) 11 ˙ 1+e ( Í ˙ ˚ Î

where Ei is the modulus of the intact rock GSI is the Geological Strength Index introduced in Section 4.4.6 D is the disturbance factor introduced in Section 4.7.2

183

Rock Mass Characterization

Compromise RMR = 15 log Q + 50

Deformation modulus Emass (GPa)

90 80

Emass = 2 RMR – 100

70 60

Emass = 10 Q1/3 c

50 40 30 Emass = 10

20

Case histories

(RMR – 10) 40

Bieniawski (1978) Serafim and Pereira (1983)

10 0

0

10

0.001

30 40 50 60 70 80 20 Geomechanics rock mass rating (RMR) 0.01

0.1

1.0 Q rating

10

100

90

100

1000

FIGURE 4.57 Measured values of static rock mass modulus, Em, and some empirical relations. (After Int. J. Rock Mech. Min. Sci., 39, Barton, N., Some new Q-value correlations to assist in site characterization and tunnel design, 185–221, Copyright 2002, with permission from Elsevier.)

If a laboratory-determined value of Ei is not available, a value may be estimated from Ei = (MR) σc, where MR is the modulus ratio for the rock type concerned as given in a table provided by Hoek and Diederichs (2006) and σc is the UCS of the intact rock.

4.8  Rock Stress In sublevel open stoping, knowledge of the in situ stress field is critical in order to achieve extraction sequences giving 100% recovery with minimal dilution and ore loss. In particular, the stress field data are used as an input to rock mass classification and numerical modeling, thus enabling various sized, shaped, and oriented stopes to be arranged and extracted within manageable expressions of rock mass failure. Clearly, formal engineering design of open stoping including pillars cannot be attempted without a reasonable knowledge of the stress field. Figure 4.58 shows a

184

Geotechnical Design for Sublevel Open Stoping

FIGURE 4.58 Highly stressed stope and pillar damaged by rock bursting.

highly stressed stope and pillar where excessive stress caused significant seismicity and related damage. Another typical expression of high stress is commonly found around the vertical development required in open stoping, such as raises for cutoff slots and ventilation shafts or even blastholes. Large concentrations of stress at the boundaries of subvertical raises create rock mass failures that can be used to estimate the orientation of the major principal stress (Figure 4.59). 4.8.1  Stress Tensor Stress is a mathematical concept used to represent stored strain energy within a rock mass volume. However, it is beyond the scope of this book to address with any detail the tensorial nature of stress in three dimensions. For a complete description of the fundamental principles of stress, the reader is advised to study the books by Brady and Brown (2004) and Hudson and Harrison (1997). This book will focus on stress measurements using oriented exploration core and their interpretation. A reliable and representative estimation of in situ stress is a major requirement for the optimized design of an extraction sequence of open stopes, especially at depth. Stress tensor notation can be represented as follows:

185

Rock Mass Characterization

FIGURE 4.59 Raise wall damage due to excessive horizontal stresses.



s11 sij = t21 t31

t12 s 22 t32

t13 t23 (4.22) s 33

The tensor may be transformed to a unique orientation in which the normal stresses are maximized and the shear stresses vanish. These maximized normal stresses are termed the principal stresses, denoted by σ1, σ2, and σ3 and referred to as the major, intermediate, and minor principal stresses, respectively (Brady and Brown, 2004). 4.8.2  Stress Measurements Using Oriented Core Stress measurements using oriented core are classified as destressing– restressing techniques. These techniques involve completely decoupling a volume of rock from the stress field, then reloading the rock volume back to its original stressed condition (Villaescusa et al., 2003b). The intention is to return a rock core volume to its in situ state. The method discussed here has been called the WASM AE stress measurement technique (Villaescusa et al., 2002). It is a technique that utilizes a completely decoupled volume of rock from exploration core that is reloaded to its original stress state by reference to one indirect parameter, the acoustic emission event count.

186

Geotechnical Design for Sublevel Open Stoping

Basically, the method involves a sequence of six steps: 1. An oriented sample volume, usually common oriented exploration core (termed here the main core), is isolated from a rock mass. 2. The main core is transported to a rock mechanics laboratory and resampled by a number of smaller subcores that are taken at certain orientations relative to the axis of the main core. 3. The oriented subcores are precision ground for rightness and flatness, then fitted with suitable acoustic emission sensors. 4. Each subcore is tested under monotonically increasing uniaxial load (stress). The acoustic sensors measure the event count rate attributed to the deformation, dislocation, and propagation of preexisting cracks and the initiation of new cracks, as the stress is increased. 5. The applied stress versus the count rate is approximately bilinear with the change of relationship indicated by a demonstrable increase in noise count rate at a certain stress level (Figure 4.60). This transition point is taken to indicate the largest contemporary stress experienced by the subcore in the direction of the subcore axis. 6. The stress measurements for the oriented subcores are used in conjunction with their orientations relative to the original oriented core to determine the largest contemporary stress field experienced by the main core (Figure 4.61). Provided the rock specimen has been selected from an area previously in equilibrium with gravitational loading and tectonics (Windsor et al., 2006, 2007), this is the maximum previous stress to which a particular rock mass has been subjected by its environment.

Cumulative AE events

30 25 20

Previous maximum stress

15 10 5 0

0

5

10

15

Stress (MPa) FIGURE 4.60 Typical AE cumulative events versus applied uniaxial stress.

20

25

187

Rock Mass Characterization

WASM AE stress measurements

Pole plot

N

σ1

σ2

E

W

σ3

S WA School of Mines 90 σ1 = 0.0406 × Depth + 6.1

Stress magnitude (MPa)

80

σ2 = 0.0334 × Depth + 1.7

70

σ3 = 0.0270 × Depth

60

σv = 0.0278 × Depth

50 40 30 20 10 0

0

200

400

600 800 1000 1200 1400 Vertical depth from surface (m)

1600

1800

FIGURE 4.61 Principal stress orientations and magnitudes determined using oriented core.

2000

188

Geotechnical Design for Sublevel Open Stoping

This section presents the scalar characteristics (i.e., the stress magnitudes alone) from approximately 240 WASM AE rock stress tensor determinations obtained from different geological and geodynamic regimes from different continents and compares them to results compiled in an Earth Rock Stress Tensor Database (ERSTD) (Windsor, 2009). The data comprise results from techniques that attempt to measure, without a priori assumption, the complete rock stress tensor (e.g., it does not include results obtained from the hydraulic fracturing technique). The data are presented as reported, without prejudice or censorship. The distributions of the vertical stress, the principal normal stresses, and the maximum shear stress with depth in the upper 3 km of Earth’s crust from the WASM AE data set and from the ERSTD are shown in Figures 4.61 through 4.63, respectively. Figure 4.62 indicates that both data sets are distributed about a theoretical linear relationship for vertical stress given Vertical stress (MPa) 0

0

20

40

60

80

100

120

ERSTD WASM AE

500

Depth (m)

1000

1500

2000

2500 Theoretical vertical stress (unit weight 27 kN/m3)

3000 FIGURE 4.62 Distribution of vertical stress with depth, measured by WASM AE and from the ERSTD. (From Villaescusa, E. et al., Stress measurements from oriented core—A decade of results, Presented at MassMin 2012, Sixth International Conference & Exhibition on Mass Mining, Sudbury, Ontario, Canada, June 10–14, 2012a, Paper 6842, 9pp.)

189

Rock Mass Characterization

–20 0

500

0

Principal normal stresses σ1, σ2, σ3 (MPa) 20

40

60

80

100

120

140

160

S1 ERSTD S2 ERSTD S3 ERSTD S1 WASM AE S2 WASM AE S3 WASM AE

Depth (m)

1000

1500

2000

2500

3000 FIGURE 4.63 Distributions of principal normal stresses with depth, measured by WASM AE and from the ERSTD. (From Villaescusa, E. et al., Stress measurements from oriented core—A decade of results, Presented at MassMin 2012, Sixth International Conference & Exhibition on Mass Mining, Sudbury, Ontario, Canada, June 10–14, 2012a, Paper 6842, 9pp.)

by σv = zγr where z is the overburden depth and γr is the unit weight of rock, which is set here at 27 kN/m3. The WASM AE data appear to fit better with this relation than the ERSTD (Villaescusa et al., 2012). The distribution of principal normal stresses (σ1, σ2, and σ3) with depth given in Figure 4.63 shows a low frequency of tensor measurement below 1.5 km, with scatter increasing with depth. It indicates slight nonlinearity of the WASM AE data set and greater nonlinearity of the ERSTD. Note that the ERSTD is influenced at depth by a greater frequency of deeper- and lowerstress magnitudes measured around South African mine sites. Figure 4.64 shows the distribution of the maximum shear stress from WASM AE and from the ERSTD. Both data sets show nonlinearity and considerable scatter with depth, which is thought to be linked to the variability in the shear strength of Earth’s crust and its ability to sustain shear stresses (Windsor, 2009).

190

Geotechnical Design for Sublevel Open Stoping

Maximum shear stress τmax (MPa) 0

0

10

20

30

40

50

ERSTD

500

WASM AE

Depth (m)

1000

1500

2000

2500

3000 FIGURE 4.64 Distribution of maximum shear stress with depth, measured by WASM AE and from the ERSTD. (From Villaescusa, E. et al., Stress measurements from oriented core—A decade of results, Presented at MassMin 2012, Sixth International Conference & Exhibition on Mass Mining, Sudbury, Ontario, Canada, June 10–14, 2012a, Paper 6842, 9pp.)

5 Span and Pillar Design

5.1 Background The development of sublevel open stope mining methods enhanced the mechanization and increased productivity of underground bulk mining operations. This in turn led to a need to optimize the size and shape of the open stopes in order to maximize production. Unacceptable waste dilution plagued many bulk mining operations, and traditional trial-and-error approaches to optimizing stope dimensions became economically unacceptable. Furthermore, inadequate design methodologies often resulted in failure of secondary stopes with resulting production delays, increased costs, and, in some cases, loss of ore reserves. In this chapter, modern stope and pillar design methodologies will be discussed.

5.2 Empirical Span Determination Using Rock Mass Classification Methods Rock masses represent extremely complex media in which to design and construct engineered structures. During the early design stages of a project, such as the scoping and prefeasibility stages, when little detailed information on a rock mass and its stress and hydrologic characteristics are available, the use of a rock mass classification scheme can be of benefit. At its simplest, this may involve the use of a classification scheme as a checklist to ensure that some geotechnical information has been considered. At the other extreme, one or more classification schemes can be used to build up a picture of the composition and characteristics of a rock mass to provide initial estimates of allowable spans and support requirements, and to provide estimates of its strength and deformation responses to the excavation process. Classification and its application to underground support is primarily founded in civil engineering tunnel construction (e.g., rock quality

191

192

Geotechnical Design for Sublevel Open Stoping

designation (RQD)—Deere et al., 1967; rock mass rating (RMR)—Bieniawski, 1989; tunnel quality index (Q)—Barton et al., 1974). Due to the relatively modest depth (0–500 m) of many of these case studies and the relatively high-safety factors demanded in civil works, design recommendations from these classification systems may be difficult to apply directly in an open stoping context. They can, however, provide a first or conservative estimate of allowable span and support requirements. Laubscher and Taylor (1976) and Laubscher (1993) modified RMR for use in the design of blockcaving mines. Caving operations are beyond the scope of this book, and Laubscher’s method will therefore not be discussed further. Mathews et al. (1980) and Potvin (1988) modified the Q system and applied it to open stope design. Their methodology has been modified slightly and is presented in this chapter. A problem with rock mass classifications is that, in addition to being conservative, they are likely to be missing a key parameter, for example, joint termination (see Chapter 4). Furthermore, the stress path is not really considered and this is a significant difference with respect to civil engineering, where there is less interaction among excavations compared to the complex extraction sequences utilized in the mining industry. 5.2.1  Span Determination Using Bieniawski’s RMR System The rock mass rating (RMR) system was originally developed by Bieniawski (1973). Over the years, it has been successively refined, as more case studies have been added to its database. The reader should be aware that, over time, Bieniawski has made several changes to the ratings assigned to the different parameters (Bieniawski, 1976, 1989). Figure 5.1 presents an additional modification to Bieniawski’s (1989) span versus stand-up time graph. The changes have been made to account for the very large and stable open spans that are being achieved in massive silicified skarns at medium confining stress (Figure 5.2). This is in part due to the silicification of the orebodies and host rocks, the relatively shallow depths being mined and also the favorable condition of the geological discontinuities with respect to the exposed spans. The concept of stand-up time was originally conceived by Lauffer (1958, 1960) to indicate the time period within which an excavation will remain serviceable and after which significant instability and caving would be experienced. A stope span is defined as the minimum dimension of an open stope wall. Hutchinson and Diederichs (1996) have presented the maximum stable unsupported span as a function of Bieniawski’s (1989) RMR (RMR89) value as shown in Figure 5.3. In the absence of large-scale geological discontinuities, or very high induced stress, a temporary mine opening such as a 10 m-wide drill drive in downhole bench stoping can be analyzed. If the required stand-up time is typically less than 5 years, then it can be seen

193

Span and Pillar Design

50

1 h

10 h

1 day

1 week

1 month

20

5 10 years years

10

90

80

70

R 89

RM

15

60

50

8

40

6 5 4

30 50 40

2

RMR 89

No support required

30 1

70

60

3

1

6 1 months year

Immediate collapse

30

Unsupported span (m)

Maximum unsupported stand-up time

10

100

1,000

10,000

100,000

Maximum unsupported stand-up time (h) Tunnels

U/G mines

FIGURE 5.1 Unsupported tunnel limits. (Modified from Bieniawski, Z.T., Engineering Rock Mass Classification, John Wiley, New York, 1989, 251pp. With permission.)

that for a rock mass having an RMR89 of greater than 80, the drill drive may not need systematic cablebolt reinforcement, with the exception of bolts and mesh for personnel safety. The RMR89 data shown in Figure 5.4 indicate that few unsupported spans are stable when their dimension exceeds 20 m. This is due to the majority of the data being collected in cut-and-fill operations (Pakalnis, 2002), where full operator access is required and potentially unstable spans cannot be effectively stabilized even with the implementation of cablebolting. However, recent experience in open stoping in extremely hard rock mines, where the orebody and host rocks have been altered by a strong silicification, unsupported stable spans ranging from 20 to 40 m can be safely achieved. The open stoping data (spans exceeding 20 m) in Figure 5.4 show circles representing stable spans (depths of failure less than 2 m), square symbols representing transitional spans (depths of failure ranging from 2 to 4 m), and

194

Geotechnical Design for Sublevel Open Stoping

FIGURE 5.2 Very large and stable span exceeding 25 m, Sabinas mine, Mexico. (Photo courtesy of Peñoles, Mexico City, Mexico.)

Fair

25

Good

Very good

s ear

s ear 5y

ear 1y

6m ont

hs

ay

1w

10

1m ont

eek

h

15

Hard rock mine design zone

Unsupported stand-up time

Immediate collapse

10 y

20

1d

Maximum stable unsupported span (m)

Span

5 No support required 0 40

50

60

70

80

90

Rock mass rating (RMR89)

FIGURE 5.3 Alternate representation of RMR89 stand-up time guidelines. (From Hutchinson, D.J. and Diederichs, M.S., Cablebolting in Underground Mines, Bitech Publishers, Richmond, British Columbia, Canada, 1996, 406pp. With permission.)

195

Span and Pillar Design

nal

50

Tra nsi ti o

45 Unsupported span (m)

40 35 30

Stable

Unstable

25 20 15 10 5 0

10

20

30

40

50

60

70

80

90

100

Rock mass rating (RMR89) FIGURE 5.4 Span design using the RMR89 method. (After Pakalnis, R., Empirical design methods—UBC geomechanics an update, in R. Hammah, W. Bawden, J. Curran, and M. Telesnicki, eds., Mining and Tunnelling Innovation and Opportunity, Proceedings of the 5th North American Rock Mech Symp & 17th Tunnelling Association of Canada Conference, Toronto, July 7–10, 2002, pp. 203–210, University of Toronto, Toronto, Ontario, Canada.)

triangles representing unstable spans (depths of failure exceeding 4 m). The data can also be used as a guideline for design against immediate collapse, large instabilities, or an indication where systematic cablebolting may be required. A point to notice when using the RMR89 method for span design is that the stress path effects, as well as the localized effect of large-scale structures likely to form wedges, must also be considered. Hence, for safe access, ground support is always recommended for sublevel stope access infrastructure, even in very hard rock masses. Modern sublevel open stoping mines use cavity-monitoring systems (CMS) to continually collect data and develop databases that encompass the final geometry of the stope voids. Stope performance is determined by the depth of failure, which is defined as the distance from a design surface to a resulting wall following complete stope extraction (Villaescusa, 2004). Furthermore, rock mass classification databases from drill holes (Cepuritis, 2004; Dempers et al., 2010) can be used to establish contours of RMR89 values for each stope wall (Figure 5.5). The rock mass classification data coupled with the depths of failure from the CMS and the design stope geometry can be used to establish relationships similar to that shown in Figure 5.6. The proposed limits for the stable (depth of failure  6 m) regions for stope spans exceeding 20 m are usually based upon local mine economics.

196

Geotechnical Design for Sublevel Open Stoping

85 82 2200 RL

79 76 73 70 67

2000 RL

64

RMR

8600 E

8400 E

55

1800 RL

9000 E

58

8800 E

61

FIGURE 5.5 Contoured grid model of hangingwall RMR89 values.

50 45 40 6.8 m 6.2 m

30

20.0 m 5.5 m

25

7.0 m

20

4.8 m 4.0 m 6.0 m 6.0 m 5.0 m

2.5 m

2.0 m

0.5 m

2.0 m

0

10

20

30

40

50

1.0,2.5 m 1.5,2.0 m 0.5,3.0 m 0.5 m 4.8 m 0.5 m 6.0 m 1.0 m

m) ( 6

10

0.5 m 1.0 m

1.0 m

15

sta

Stope span (m)

35

0

4.5 m

2.6 m

60

Stope, depth of failure 1.0 m

70

RMR89 FIGURE 5.6 Depths of failure for a number of stope spans and varying RMR89 values.

80

90

100

Span and Pillar Design

197

The data shown in Figure 5.6 are for stope designs in very hard, silicified rock masses extracted by conventional sublevel open stoping. The stope data shown earlier relate to medium confining stress in mining epithermal orebodies having depths of less than 500 m. A limitation is that induced stresses cannot readily be considered when calculating the RMR89 values. Hence, a designer trying to implement a similar strategy would need to ensure that stress-driven instability is not a prominent failure mode prior to implementing an approach similar to the one described here. 5.2.2  Span Determination Using the Tunnel Quality Index (Q) System Barton et al. (1974) described the application of the Q system for rock mass classification for the determination of no-support limits for various types of excavations. Some 200 original case studies were used in the original calibration of the method. Over the next 18 years, more than 2000 new empirical tunnel and large cavern designs were successfully carried out (Barton et al., 1992). Figure 5.7 shows the updated plot for ground support recommendations. The solid lines bound the limits of practical support application, with the lower line demarcating the stability limit for unsupported excavations of a given equivalent span, ES = Span/ESR, where values for excavation support ratio (ESR) are given in Table 5.1. The ESR is a factor used by Barton to allow for varying degrees of instability based on excavation service life and use. The actual span of the excavation is divided by the ESR value to obtain the equivalent span for use in Figures 5.7 and 5.8. Hutchinson and Diederichs (1996) note that the number of mining case histories leading to the recommendation of ESR = 3–5 for temporary mine openings is extremely limited and therefore recommend using a maximum ESR of 3 for these openings unless local experience justifies the use of higher values. Certain mining excavations are more critical than others from both operational and safety points of view. Figure 5.8 (after Hutchinson and Diederichs, 1996) provides guidelines for no-support limits in order of decreasing reliability, relating them to Barton’s original ESR values. Figure 5.8 is plotted against actual excavation span. Nevertheless, the direct use of Q for open span design is not well documented within the mining industry.

5.3  Stability Graph Method Sublevel open stoping has become one of the most common underground mining methods in the world due largely to its safety and efficiency. Dimensioning of sublevel intervals, strike spans, pillars, and their location is

0.001

1

2

5

10

20

50

100

ESR

0.004

0.01

0.04

1.5 m

1.7 m

0.4

1

1.3 m

5c

m

Good

B

2.3 m 2.5 m

ng

2.0 m

10 Jw SRF

i 1.6 m pac lt s Bo

4 RQD Jr Rock mass quality Q = Jn Ja

0.1

m

1m

9c

e te et e re et cr tc cr ot lts o t h h t o s s bo ls s sh olt d bo ed d ce nd ed d b rc an r c o ) fo a or an nf in ) ei cm nf ) re cm -r 5 ei m r- –9 er 2–1 - r 12 c e r b b Fi (1 be – Fi (5 Fi (9

m) 5 c lts tc >1 d bo ( Cas e ret e an otc et sh otcr d h rce of s nfo s cm cm cm rei d rib 12 25 15 r e e rc b i F nfo i Re

ing e lin

ret onc

1m

a

re eted a shotcr ing in c a p s Bolt 1.3 m 1.2 m

Unsupportable

Fair

Poor 2.1 m

C

Rock classes D

i

sh

n nu

rea

da ete

40

100

1.5

2.4

3

5

7

11

20

1000

Exc. good

400

No support required

r otc

3.0 m

A Ext. good

Spot bolting

4.0 m

Very good Bolt length in metres for ESR=1

FIGURE 5.7 Updated ground support recommendations. (After Grimstad, E. and Barton, N, Updating the Q-system for NMT, in R. Kompen, O.A. Opsahl, and K.R. Berg, eds., Proceedings of the International Symposium on Sprayed Concrete—Modern Use of Wet Mix Sprayed Concrete for Underground Support, Fagernes, Norway, October 17–21, 1993, pp. 46–66, Norwegian Concrete Association, Oslo, Norway.)

Excavation span (m)

E Very poor

ste

F Extremely poor

ing

G Exceptionably poor

nd (a

) cm 4

un Sy re st in em 4 c forc ati ed c b m sh olt o tc ing re te ,>

olt cb

ma ti

Sy

198 Geotechnical Design for Sublevel Open Stoping

199

Span and Pillar Design

TABLE 5.1 Excavation Support Ratio Type of Excavation

Number of Cases

ESR (Approx.)

2 83

3–5 1.6

25

1.3

79

1

Temporary mine openings Permanent mine openings: low-pressure water tunnels; pilot tunnels; drifts and headings for large openings Storage caverns; water treatment plants; minor road and railway tunnels; surge chambers; access tunnels, etc. Power stations; major road and railway tunnels; civil defense chambers; portals; intersections Underground nuclear power stations; railway stations; sports and public facilities; factories

2

0.8

Source: Barton, N.R., Rock mass classification and tunnel reinforcement selection using the Q-system, in L. Kirkaldie (ed.), Rock Classification Systems for Engineering Purposes: ASTM Special Technical Publication 984, ASTM International, Philadelphia, PA, 1988, pp. 59–88.

Poor

Fair

Very good

Good

Extremely good

Maximum unsupported span (m)

200 100 50

try

n

ope

alls

—w

es stop

ited

rifts

d cess

ac

10 5 2

1

4

10

im s—l

40

100

1.6 1.3 1.0

gs

enin

p nt o ane gs ack m b r n e e i op nd pal open s n st ls a Ope nne critic station u t e e nd ulag ns a efug e ha statio and r n i M sher ions t d Cru aft sta uire req Sh t r po sup No

en Non

20

1

3

e

ilur

fa iate

ed

Imm

Exceptionally good ESR 5

400

1000

Rock tunnelling quality index (Q) FIGURE 5.8 Q system; no-support span limits for underground mine openings. (After Hutchinson, D.J. and Diederichs, M.S., Cablebolting in Underground Mines, Bitech Publishers, Richmond, British Columbia, Canada, 1996, 406pp. With permission.)

200

Geotechnical Design for Sublevel Open Stoping

very important to the success of the method. An empirical method of evaluating strike length span stability was developed in Canada by Mathews et al. (1980). The method was further developed and applied by Potvin (1988), Bawden et al. (1988, 1989), Nickson (1992), and Mawdesley et al. (2001), among others. The original intent was to provide a practical design tool for Canadian mine operators. The following five objectives were set for model development (Bawden, 1993):

1. The model should be capable of predicting the overall stability of a stope in terms of operating problems. Instead of focusing on precise calculations and the identification of every single potential block fall, the model should concentrate on defining conservative stope dimensions, less conservative stope dimensions, and critical stope dimensions above which open stoping becomes impractical. 2. The model must be reliable and hence sensitive to all key geotechnical parameters affecting underground stope design. It is also important that the different conditions associated with open stope mining such as stope geometry, mining sequence, blasting, and support from fill and cablebolts are directly or indirectly accounted for. 3. The model must be easy to use by mining or geological engineers on site. The input parameters should rely mainly on observational methods rather than expensive testing, lengthy studies, and sophisticated equipment. 4. The model should be usable at any stage of mining (i.e., at the feasibility study and for short-term and long-term planning). Although the precision of any model is largely a function of the quality of the input parameters, which are better understood as mining progresses, the model should be capable of providing at least approximate answers at the feasibility study stage. 5. The model should be representative of rock mass behavior and be capable of identifying underground modes of failure. This will provide a better understanding of the ground conditions and help in selecting proper remedial solutions to ground control problems. 5.3.1  Updated Determination of the Stability Graph Parameters The stability graph method is effectively a modification of the Q (1974) rock mass classification method. The method relies on relating a stability number (N′) to a stope wall hydraulic radius by way of a number of curves, each depicting various levels of stability. For each stope wall, a stability number is defined as follows:

N¢= Q¢ABC (5.1)

201

Span and Pillar Design

where A is a stress factor B is a rock defect orientation factor C is a design surface orientation factor (Potvin, 1988) Q′ is defined following Barton et al. (1974) as Q¢=



RQD Jr (5.2) Jn Ja

where RQD, Jn, Jr, and Ja are defined as per Table 4.4 and the different classification guidelines described in Section 4.3.3. Furthermore, Figures 4.40 through 4.43 illustrate the typical variability in the individual parameters required to determine Q′. Following a similar procedure to that described in Section 4.5.5, the value and variability of Q′ can be evaluated as shown in Figure 5.9. The parameters A, B, and C are defined individually as in the following subsections. 5.3.1.1  Factor A The rock stress factor A was initially designed to replace the stress reduction factor (SRF) in the original Q (Barton et al., 1974) system (Mathews et al., 1980). Similarly to the SRF, it was defined as the ratio of uniaxial compressive strength (UCS) of intact rock to the induced compressive stress parallel to 20 18 2200 RL

16 14 12 10

2000 RL

8

0

Q

1800 RL

FIGURE 5.9 Contoured grid model of hangingwall stability number Q′.

9000 E

2

8800 E

8400 E

4

8600 E

6

202

Geotechnical Design for Sublevel Open Stoping

20

σ1 = In situ main principal stress σc = Uniaxial compressive strength of intact rock

Stress reduction factor (SRF)

10 5

Near surface 2.5–5.0

Heavy rockburst zone

2 1

High stress

0.5

Low confining stress

Medium confining stress

0.2 0.1 1

2

5

10

σc

20

50

100

200

σ1 FIGURE 5.10 Stress reduction factor. (After Hutchinson, D.J. and Diederichs, M.S., Cablebolting in Underground Mines, Bitech Publishers, Richmond, British Columbia, Canada, 1996, 406pp. With permission.)

the stope surface under consideration. However, the factor A as proposed by Mathews et al. (1980) does not specifically account for the loss of confinement, as the SRF does (Figure 5.10). Stress relaxation may have a large effect on jointed rock masses as it provides freedom of movement for individual blocks. This is taken into account by the SRF within the low confining stress zone. In addition, the original SRF factor accounts for improved stability while mining under medium confining stress conditions. Experience has shown that even a modest amount of confining pressure is likely to increase the ultimate strength around stope walls. Data from many years of numerical modeling and observations of open stoping at Mount Isa Mines (Villaescusa, 1996), as well as the relative data from the SRF, were used to review the stress factor A for input into stope stability assessment. The results in Figure 5.11 show that the original factor A (Potvin, 1988) is significantly more conservative than the SRF in the original Q (1974) method. Consideration of modern stope blasting practices and back analysis of strength/stress ratio data from open stope walls at a large number of Australian mines have been used to define a new A factor (Figure 5.12). As with the original SRF, the benefits of medium confining stress are taken into account, and it is suggested that no correction for compressive failure need be undertaken when the ratio of UCS/induced stress exceeds 5.5. The resulting variability using the new factor A throughout four stoping blocks and individual stope outlines is shown in Figure 5.13. The results

203

Span and Pillar Design

1.0 0.9

Stress factor A

0.8

Villaescusa (1996) Potvin (1988) Q-1974

0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0

0

1

2

3

4

5

6

7 8 9 UCS/stress

10

11

12

13

14

15

FIGURE 5.11 A comparison of SRFs from a number of sources.

Rock stress factor A

1.0 0.9 Sto

0.6 0.5

σmax

0.0

Medium confining stress

High stress

0.4 0.3 0.2 0.1

all

pe w

0.8 0.7

Very high stress 0

1

2

3

4

5

6

7

8

9

10

σmax from 3D numerical modeling

σ Ratio: σ c

max

FIGURE 5.12 Stress factor A and regions of stress considered.

are in accordance with observed conditions at the Kanowna Belle mine, where very little stress-related failure was experienced within stoping blocks A, B, and C. 5.3.1.2  Factor B The rock defect orientation factor B is a weighting factor based on the orientation of the discontinuity set that is considered most likely to detract from the stability of a particular stope surface (Potvin, 1988). The method requires analysis of the discontinuity data to determine the most critical discontinuity

204

Geotechnical Design for Sublevel Open Stoping

1

Block A

.9 10000 N

.8

Block B

.7 .6

9800 N

Block C

.5 .4 .3

9600 N

20400 E

9400 N

Factor A

20200 E

0

19600 E

.1

20000 E

Block D

19800 E

.2

FIGURE 5.13 Contoured grid model using new factor A for stope hangingwalls, Kanowna Belle mine, Western Australia.

likely to control stability. The determination of factor B requires the calculation of the true angle between a planar stope surface and the critical geological feature. Considering that the most critical discontinuities are subparallel to a stope surface, a few changes have been implemented to the original factor B (Potvin, 1988). Based on many observations of actual stope wall failures, it is suggested that no correction for discontinuity orientation should be undertaken when the true angle with a stope surface exceeds 65° as shown in Figure 5.14. In addition, a maximum penalty of 60% to the calculated Q′ is suggested for the effects of subparallel discontinuities. An example of the variability of factor B throughout a number of stoping blocks is shown in Figure 5.15. The solid angle α (Figure 5.14) between the poles of a stope wall (P) and a critical geological discontinuity (D) can be calculated from the dot product

P ◊D = P D cos a (5.3)

The vector P is the unit vector of the direction cosines of the normal to a stope wall (P), which are defined by

205

N Discontinuity pole Stope wall pole

α

0

10

20

30

40

50

60

70

80

ll wa

W

pe

1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0

o St

Joint orientation factor B

Span and Pillar Design

90

E

y

uit

tin

con

Dis

S

True angle between face and discontinuity (angle α between poles)

FIGURE 5.14 Influence of joint orientation—factor B. (Modified after Potvin, Y., Empirical open stope design in Canada, PhD thesis, University of British Columbia, Vancouver, British Columbia, Canada, 1988, 350pp.)

1 .9 10000 N

.8

Block B

.7 .6

9800 N

.55

Block C

.5 .45

9600 N

.4

20200 E

20000 E

9400 N

Factor B

19800 E

19600 E

0

20400 E

Block D

.3

FIGURE 5.15 Contoured grid model of new factor B for stope hangingwalls, Kanowna Belle mine.

206

Geotechnical Design for Sublevel Open Stoping



Px = cos jp sin bp



Py = cos jp cos bp (5.4)



Pz = sin jp ,

where βp and φp are the trend and plunge of the normal to the stope wall plane (see Figure 4.30). The vector D is the unit vector of the direction cosines of the normal to a critical discontinuity, which are defined by

D x = cos jd sin bd



D y = cos jd cos bd (5.5)



D z = sin jd ,

where βd and φd are the trend and plunge of the normal to a critical discontinuity, as defined in Figure 4.29. The angle α is thus given by

cos d = Px D x + Py D y + Pz D z (5.6)

5.3.1.3  Factor C The design surface orientation factor C was proposed to account for the influence of gravity on the stability of the stope surface (Potvin, 1988). The factor is based on the assumption that under the effects of gravity, a vertical stope wall is more stable than a horizontal stope back. Surfaces where sliding blocks can form or where significant overhangs occur (i.e., stope backs and hangingwalls) will have the most detrimental influence on stability. Two adjustment factors were proposed by Potvin (1988) and have been modified here to account for the back analysis of stope stability at a number of Australian mines. The effects of gravity fall and slabbing are considered in Figure 5.16. The adjustment factor has been made constant for flat stope backs having a dip of less than 20° (Bieniawski, 1989). The second adjustment factor proposed by Potvin (1988) to analyze sliding modes of failure of stope walls is shown in Figure 5.17. Assuming that the frictional resistance of a critical discontinuity exceeds the driving force, the amount of adjustment has a maximum value of 8 when the dip of a critical discontinuity is less than 30°. It is proposed here that as the dip of a critical discontinuity increases, the adjustment will decrease to a minimum value of 4. According to Potvin (1988), the potential mode of failure can be determined with a simple diagram in which the excavation and the critical joint are sketched. If a gravity vector represented by a vertical arrow drawn from

207

Span and Pillar Design

pe

Dip n sto

10 9 8 7 6 5 4 3 2 1 0

Slabbing

Ope

Gravity adjustment factor C

Gravity fall

0

10

20

30

40

50

60

70

80

90

Dip of stope wall (degrees) FIGURE 5.16 Determination of gravity effects—factor C. (Modified after Potvin, Y., Empirical open stope design in Canada, PhD thesis, University of British Columbia, Vancouver, British Columbia, Canada, 1988, 350pp.)

the approximate center of gravity of the block formed by the critical discontinuity falls directly inside the opening, the mode of failure will be gravity fall. In addition, if the gravity vector stays inside the medium without intersecting the critical discontinuity, slabbing or buckling failure can occur. Furthermore, when the gravity vector crosses the critical joint, the potential for sliding failure exists (Potvin, 1988). An example of the variability of factor C throughout a number of stoping blocks is shown in Figure 5.18. 5.3.1.4  Hydraulic Radius The hydraulic radius concept to account for the size and shape of a stope plane under analysis was introduced by Laubscher and Taylor (1976). Hydraulic radius is the quotient of the stope wall area and the stope wall perimeter, and favors long and narrow shapes over square shapes (see Figure 1.5).

208

Geotechnical Design for Sublevel Open Stoping

Ope n st ope

Discontinuity dip

Gravity adjustment factor C

Sliding 10 9 8 7 6 5 4 3 2 1 0

0

10

20

30

40

50

60

70

80

90

Dip of critical discontinuity (degrees) FIGURE 5.17 Determination of sliding effect on critical joint—factor C. (Modified after Potvin, Y., Empirical open stope design in Canada, PhD thesis, University of British Columbia, Vancouver, British Columbia, Canada, 1988, 350pp.)

Hydraulic radius is easy to assess as most stope shapes are not very complex. The methodology allows the analysis of stope surfaces wall by wall. The relationship between hydraulic radius (i.e., area/perimeter) and excavation length, given a fixed height, usually defined by the sublevel interval, is given by



HR =

(H)(L) (5.7) 2(H + L)

and



L=

2(H)(HR ) (5.8) H - 2(HR )

209

Span and Pillar Design

10 Block A

9 10000 N

8

Block B

7 6

9800 N

Block C

5 4 3

9600 N

2

Block D

20400 E

20200 E

20000 E

9400 N

Factor C

19600 E

0

19800 E

1

FIGURE 5.18 Contoured grid model of new factor C for stope hangingwalls, Kanowna Belle mine.

where HR is the hydraulic radius and H and L are the height and length of the stope wall, respectively. In order to determine the maximum allowable unsupported lengths, the height or width of the excavations needs to be first determined. For vertical walls, this generally relates to floor-to-floor dimensions for the stope surface under consideration. Consider, for example, Figure 5.19a, which shows that for a footwall, the stope down-dip span is “fixed,” as it is determined by the sublevel interval chosen. For the stope backs and end walls, the width is generally controlled by the ore or stope width (as for narrow vein, generally, stopes are purposely not mined wider than the ore width). For a hangingwall, because of the cablebolting reinforcement at every sublevel interval, the “fixed” dimension is the down-dip span between the cablebolts (for steeply dipping orebodies, this is approximately equal to the level interval spacing). Figure 5.19b shows that the rock mass exposed between the cablebolts must be inherently stable, as the cablebolts only minimize the deformation locally near the stope drives. The cablebolts are also very effective in arresting failures up-dip (see Figure 1.13). 5.3.2  Prediction of Stope Stability The calculation of the stability number (Equation 5.1) for a particular stope wall is achieved by multiplying the variables accounting for the geotechnical

210

Geotechnical Design for Sublevel Open Stoping

HR = HR back =

Area Perimeter

L*W L*H HR hangingwall = 2 * (L + W) 2 * (L + H)

Hangin gw Down-d all ip span (H )

Foo Down-d twall ip span

(H)

Maximum 2 * HR * H = allowable Back (H – 2 * HR) length (Lmax) ) Stope width (W) x a ll wa gth (L m ing ng le len a H wab llo xa a M

x Ma

(a)

) ax all th (L m w g t n o Fo ble le wa allo

(b)

FIGURE 5.19 (a) Fixed and allowable stope dimensions and (b) hangingwall failure. (b: Courtesy of Mount Isa Mines, Mount Isa, Queensland, Australia.)

parameters previously described. The initial back analysis work in Canada included a total of 175 case studies of unsupported open stope walls from 23 Canadian mines (Potvin, 1988). The initial stability graph shown in Figure 5.20 is composed of stable and caving zones, separated by a transition zone. The stope walls were divided by Potvin (1988) into three groups. Stable walls that experienced low dilution were represented by roundshaped points. Stope walls that experienced dilution and rock falls causing operational problems were classified as unstable. They are shown on the graph as square-shaped points. The triangular points represent stope walls that experienced severe instability. The solid black line shown in Figure 5.20 was calculated by Nickson (1992) to statistically account for the difference between stable and caved points. The relationship between the stability number N′ and the maximum allowable unsupported hydraulic radius (HRallowed) is given as a function of the stability number by

HR allowed = 10[0.573 +0.338 log N¢] (5.9)

Nickson’s boundary allows for larger stope dimensions than those predicted by Potvin’s unsupported transitional zone. The statistical boundary developed by Nickson can be used to predict the maximum allowable stable open

211

Span and Pillar Design

1000

Stable zone

Stability number (N')

100

HR== 10 HR

10

(0.573 + 0.338 log N')

Caved zone

1.0

0.1

0

5

10

15

20

25

Hydraulic radius (m) FIGURE 5.20 Initial stability graph calculated from 175 case histories of unsupported open stope walls. (After Potvin, Y., Empirical open stope design in Canada, PhD thesis, University of British Columbia, Vancouver, British Columbia, Canada, 1988, 350pp.)

stope surface relating to that particular stability number. For example, for a stability number of 11, a corresponding hydraulic radius of 10 is allowed and is recommended as a first estimate for stope span design. Nickson (1992) also increased the initial database for the stability graph method and eventually updated the stability graph to the form shown in Figure 5.21. This figure can be used to evaluate maximum allowable stope wall sizes for either unsupported or pattern (full coverage) cablebolted stope walls. However, Nickson (1992) clearly stated that the graph cannot be used to design cablebolted hangingwall spans where the cables are installed from localized drill drive locations (point anchored hangingwall cablebolting or rib rock; see Rauert, 1995). Stability evaluations of cablebolted stope hangingwalls must ensure that any unsupported rock mass exposed down-dip between finite cablebolting locations is inherently stable, as per Equation 5.9. Figure 5.22 shows stope stability data for unsupported and completely stable (zero depth of failure) open stope walls at the Cannington mine, Queensland (Coles, 2007). The figure compares the calculations from Potvin (1988) with the prediction using the updated parameters presented in this chapter. The figure also shows the original relationship developed by Nickson (1992).

212

Geotechnical Design for Sublevel Open Stoping

1000 500 200

HRallowed = 10[0.573 + 0.338 log N'] Stable zone

50 20

U

10 5 2 1 0.5

e

zon

Caving zone

Re

0.2 0.1

r

dt

rte

po

p nsu

on

iti ans

St ca able ble w bo ith inf lt orc re full c ed in tra fo ove rc ra nsi em ge tio en nz t on e

Stability number (N')

100

0

5

Full coverage cablebolting

10 Hydraulic radius (m)

15

20

Localized cablebolting

d Unsu pporte Unsu

pporte

d

g

ltin ebo

l cab

Ponit an ch cablebo ored (Rib-roc) lt reinfo rcemen t

age

ver

l co

Ful

Extended chart is only applicable for full coverage cablebolt reinforcement geometries Limit for unsupported design given by HRallowed = 10[0.573 + 0.338 log N']

FIGURE 5.21 Stability graph showing zones of stable ground, caving ground, and ground requiring cablebolt reinforcement. (After Nickson, S.D., Cablebolt support guidelines for underground hard rock mine operations. MASc thesis (unpublished), University of British Columbia, Vancouver, British Columbia, Canada, 1992.)

213

Span and Pillar Design

1000

Potvin (1988)

Stable zone

Modified after Potvin (1988)

Stability number (N')

100

HR

10

= 10

.338

3+0

0.57

´

log N

1 Caved zone 0.1

0

5

10

15

20

Hydraulic radius (m) FIGURE 5.22 Stability graph for unsupported, completely stable (zero depth of failure) open stope walls, Cannington mine, Queensland. (Data from Coles, D., Performance of open stopes at BHPBilliton Cannington mine, BEng thesis, Western Australian School on Mines, Curtin University of Technology, Kalgoorlie, Western Australia, Australia, 2007, 161pp.)

Regardless of the empirical methodology chosen, the final design of an open stope must always consider the geotechnical issues described earlier together with economic, scheduling, and mining constraints. Consequently, engineering judgment is always required to establish the most efficient stope wall design. 5.3.3  Use of the Stability Graph as a Design Tool Relational geotechnical databases that include information on UCS, rock mass classification data such as RQD, joint set number, joint orientation and condition, and stope wall performance such as depth of failure can be used to calibrate the stability graph for existing stoping blocks. The sample locations (X,Y,Z) for each data point in the database can be plotted in three dimensions to obtain a visual appreciation of the spatial distribution and density of the database with respect to the orebody and its immediate boundaries. It is important to note that, although the total number of samples in the database is always significant, it is critical to ensure that the relevant samples are actually located within the immediate hangingwall and footwall or the orebody in question (see Figure 4.7). Figure 5.23 shows the modeled spatial variability of the Q′ parameter at the Kanowna Belle Orebody, Kalgoorlie, Western Australia. The model predicts a reduction on rock mass quality at

214

Geotechnical Design for Sublevel Open Stoping

20 18

Block A

10000 N

16 14

Block B

12

9800 N

10 Block C

8 9600 N

6 4

20200 E

20000 E

Q

19800 E

9400 N

19600 E

0

20400 E

Block D

2

FIGURE 5.23 Contoured grid model of Q′ for stope hangingwalls, Kanowna Belle mine.

depth for the Kanowna Belle hangingwall boundary and the stope design must account for such variation in space. In addition to rock mass classification data, the anticipated maximum and minimum induced principal stresses tangential to the stope walls are also required to more accurately determine the stress conditions required to calculate factor A in the stability graph method. The induced stresses can be estimated using three-dimensional numerical modeling. For each mining step within the numerical model, the major and minor induced principal stresses across each mining surface are located and recorded, along with the threedimensional coordinates of these points (Figure 5.24). It must be remembered that the induced stresses depend on the stoping extraction sequence. An example of a longitudinal view of the induced major principal stresses on a hangingwall plane for four stoping blocks is shown in Figure 5.25. A significant increase in induced stress with depth can be seen. Very high stresses are expected in stoping block D where induced stresses up to three times higher than those experienced in block A are predicted. The stability number (N′) must be calculated independently for each stope wall. Instability will occur in surfaces where sliding blocks can form, or where significant overhangs occur. Flat joints are likely to have a significant effect on stope backs (crowns) and the stability within vertical walls will be

X

X

0.0

0.0

34 MPa

2

3 1

20.0 40.0 60.0 80.0 100.0 120.0 140.0 160.0 180.0 200.0

65 MPa

20.0 40.0 60.0 80.0 100.0 120.0 140.0 160.0 180.0 200.0 3 2 1

CT98-62

CT02-62

CP03-62

DR03-59

Induced (σ1) model

CS00-62

CP80-62

• Major induced stress (σ1) at time of stope extraction utilized (i.e., sequence dependent)

σ1 (MPa)

0

12

24

36

48

60

72

84

96

108

120

FIGURE 5.24 Estimating the induced major principal stress using the computer program MAP3D.

Y

σ1 (MPa)

Y

σ1 (MPa)

CT05-62

CP07-

Span and Pillar Design 215

216

Geotechnical Design for Sublevel Open Stoping

120 108

Block A

10000 N

96 84 72

Block B 9800 N

60 Block C

48 36

9600 N

24

20200 E

20000 E

19800 E

9400 N

19600 E

0 Signal (MPa)

20400 E

Block D

12

FIGURE 5.25 Contoured grid model of the induced major principal stress (σ1) for stope hangingwalls, Kanowna Belle mine.

controlled by the presence of subvertical to moderately dipping geological discontinuities having strikes oriented subparallel to a stope surface. The mode of failure, however, is dependent on the dip direction of the critical joint with respect to the particular stope wall. Figure 5.26 shows back analysis data from open stoping at the Olympic Dam mine, South Australia, showing different degrees of instability for the different stope walls forming the stope shapes. The stability number for a particular stope surface can be calculated for the grid models by multiplying the component terms from each of the grid models to evaluate Equation 5.1. An example of the resulting stability number is presented in Figure 5.27. The allowable hydraulic radius (HRallowed) for a given N′ value is given by Equation 5.9 (Nickson, 1992). The HRallowed results for the grid model are shown in Figure 5.28. In order to determine the maximum allowable unsupported lengths (Lmax), the height of the designed stopes needs to be first established. A decision must be taken to determine if cablebolt reinforcement effectively reduces the down-dip span as shown in Figure 5.19a. An example of a contoured grid model of interlevel down-dip height for the stope hangingwall surfaces at the Kanowna Belle mine is given in Figure 5.29.

217

Span and Pillar Design

100

Stable

90

Unstable

Failed

Occurrence (%)

80 70 60 50 40 30 20 10 0

Crown

N

E

S

W

Stope stability by wall orientation FIGURE 5.26 Stope stability by wall orientation at the Olympic Dam mine, South Australia. (From Oddie, M.E. and Pascoe, M.J., Stope performance at Olympic Dam mine, Proceedings of the 9th Underground Operators’ Conference, Perth, Western Australia, Australia, March 7–9, 2005, pp. 265–272, AusIMM, Melbourne, Victoria, Australia. With permission.)

20 Block A

18 10000 N

16

Block B

14 12

9800 N

10

Block C

8 9600 N

6 4

20200 E

20000 E

N

19800 E

9400 N

19600 E

0

20400 E

Block D

2

FIGURE 5.27 Contoured grid model of stability number N′ for stope hangingwalls, Kanowna Belle mine.

218

Geotechnical Design for Sublevel Open Stoping

15 13.5

Block A

10000 N

12 Block B

10.5 9

9800 N

7.5

Block C

6 4.5

9600 N

3

20200 E

20000 E

19800 E

9400 N

19600 E

0 Hydraulic radius

20400 E

Block D

1.5

FIGURE 5.28 Contoured grid model of stable unsupported hydraulic radius (HR allowed) for stope hangingwalls, Kanowna Belle mine.

40 36

Block A 10000 N

32

Block B

28 24

9800 N

Block C

20 16 12

9600 N

8

Block D

20400 E

20200 E

20000 E

9400 N

19600 E

0 Down dip height

19800 E

4

FIGURE 5.29 Contoured grid model of interlevel down-dip height for the hangingwall surfaces, Kanowna Belle mine.

Span and Pillar Design

219

Given a fixed sublevel interval, the stable hydraulic radius contour plots (HRallowed) enable a determination of the maximum allowable stable lengths (Equation 5.8) as shown in Figure 5.30. The plot also shows actual (mined or designed) dimensions for comparison. The close agreement for block A suggests that the modifications to the factors A, B, and C presented here are well established. 5.3.4  Design Validation The stability graph method was originally developed as an initial assessment of stability at the prefeasibility stages of projects. Currently, the method is being used worldwide as a design tool in all stages of stope dimensioning and has become an established empirical tool for dimensioning open stope walls. However, the system has a number of limitations that must be understood in order to assess its applicability in any particular geotechnical environment. Over the years, the applicability and limitations of the method for open stope design has been reviewed by several authors (Pakalnis et al., 1995; Stewart and Forsyth, 1995; Suorineni et al., 2001; Suorineni, 2012). In particular, the following observations are considered to be important:

1. The definitions of stable versus caving conditions are subjective since the depth of failure is not reported. In addition, the method does not incorporate complex failure mechanisms involving more than one family of geological discontinuities. Specifically, the method does not consider buckling in which the frequency of subparallel discontinuities may be critical. 2. Despite the use of quantifiable input values, the precise degree of inherent conservatism is not known. 3. The method reflects mining practice, which may have been influenced by factors such as legislation, local practices, and particular geological peculiarities. The method lacks sufficient precision for stope dimensioning (excessive scatter). The following factors are likely to have affected the stable/unstable boundaries identified during method development in Canada and may not necessarily be the same elsewhere: • • • • • •

Stoping style methodology Volume of overbreak or dilution levels Blasting practices Stress regime (including destressing or tensile failures) Determination of induced stress in complex stoping geometries Mine-wide determination of intact rock parameters such as UCS

9400 N

9600 N

9800 N

10000 N

(a)

Block D

Block C

Block B

Block A

(b)

20200 E

20000 E

19800 E

20200 E

20000 E

19800 E

Block D

Block C

Block B

Block A

FIGURE 5.30 Contoured grid model of (a) maximum allowable unsupported length and (b) mined and designed strike length for the hangingwall, Kanowna Belle mine.

Strike length

–0

–4

–8

–12

–16

–20

–24

–28

–32

19600

–36

20400 E

–40

20400 E

220 Geotechnical Design for Sublevel Open Stoping

221

Span and Pillar Design

• Quality control on reinforcement installation • Type of reinforcement, use of plates, etc. • Quality of rock mass characterization, detailed mapping including biases Therefore, the stability graph method may not necessarily constitute an optimum design methodology but, rather, a starting point for each particular geotechnical environment. Empirical evidence and ongoing documentation are therefore critical to the implementation of optimized stoping geometries at any particular mine site. Consequently, design validation represents a critical component in the application of the stability graph. Validation is accomplished through the use of various instrumentation strategies ranging from simple underground observations at the most basic, to minewide microseismic systems at the most complex. Geotechnical instrumentation is of critical importance to the mine design approach discussed herein. Other than for local safety considerations, instrumentation should be placed to help calibrate design models. It is essential that all instrumentation be very carefully designed and located to ensure maximum benefit and interpretability. In order to emphasize this applicability and validation point, Figure 5.31 from a published back analysis of open stopes at the Olympic Dam mine 1000

Stable Unstable

Stable region

Failed Stability number (N')

100

n

Tra

ion

l reg

na sitio

10 Unstable region 1.0

0.1

0

5

10

15

20

Hydraulic radius (m) FIGURE 5.31 Stability graph calculations for unsupported stope walls at the Olympic Dam mine. (From Oddie, M.E. and Pascoe, M.J., Stope performance at Olympic Dam mine, Proceedings of the 9th Underground Operators’ Conference, Perth, Western Australia, Australia, March 7–9, 2005, pp. 265–272, AusIMM, Melbourne, Victoria, Australia. With permission.)

222

Geotechnical Design for Sublevel Open Stoping

(Oddie and Pascoe, 2005) is presented. The resulting data show little or no correlation with the stability graph, suggesting that a local parameter, perhaps not well accounted for by the stability graph methodology, controls the stability of the Olympic Dam mine open stope walls.

5.4  Numerical Modeling of Stope Wall Stability The main objective of numerical modeling is to quantify the effects of induced stress on stope performance. This is achieved by relating different levels of induced stress to different levels of rock mass damage around a stoping void. The underlying assumption is that stress-induced failure occurs from induced stresses exceeding the local rock mass strength, thus resulting in stope wall overbreak. Unfortunately, this assumption could lead to variability in back analysis of open stope performance results because the resulting stope void geometry may not necessarily define the excavation damage zone or yield zone of the rock mass around a stope (Cepuritis et al., 2007, Figure 5.32). Material around a stope void could actually represent “Unyielded” “unremoved” rock mass outside planned void

?

Amount of “yielded” rock mass unknown with CMS data “Yielded” “unremoved” rock mass inside planned void

Unyielded/yielded boundary (EDZ) Unremoved/removed boundary Inside/outside planned void boundary

“Yielded” “removed” rock mass outside planned void

“Unyielded” “unremoved” rock mass inside planned void “Yielded” “removed” rock mass inside planned void “Yielded” “unremoved” rock mass outside planned void

FIGURE 5.32 Schematic showing resulting stope void with respect to possible yielded rock mass conditions and planned void geometry. (From Cepuritis, PM. et al., Back analysis and performance of block A long hole open stopes—Kanowna Belle Gold mine, in E. Eberhardt, D. Stead, and T. Morrison, eds., Rock Mechanics: Meeting Society’s Challenges & Demands, Proceedings of the First Canada—US Rock Mechanics Symposium, Vancouver, British Columbia, Canada, May 27–31, 2007, pp. 1431–1439, Taylor & Francis, Leiden, the Netherlands.)

Span and Pillar Design

223

“yielded” yet “unremoved” rock mass, where the local shape and span could have arched holding up yielded material. In addition, “yielding” of a rock mass cannot always be solely attributed to stress-induced rock mass damage, as other influences such as poor drill and blast practices may also contribute. Nevertheless, numerical modeling techniques can be used to identify and quantify the relative contributions of the various factors that influence stope performance, including stope geometry, development location and undercutting, rock mass characteristics, in situ and induced stresses, and the influence of large-scale geological structures. For open stoping, the choice of modeling technique includes linear elastic numerical modeling, such as the program Map3D (Mine Modelling, 2013), and nonlinear continuum or discontinuum finite element analysis, such as Abaqus (Beck and Duplancic, 2005). In particular, Abaqus is used specifically for the analysis of stoping problems where there is potential for significant plasticity and high levels of deformation with large-scale structures explicitly incorporated in the model. 5.4.1  Linear Elastic Numerical Modeling Wiles (2001) suggested that rock mass damage can be related to the relative level of linear elastic overstressing (Figure 5.33a). The critical stress levels are dependent on mine site-specific parameters and can be correlated using the observed rock mass response and the results from numerical modeling. The assumption is that below a site-specific damage threshold, the rock mass response is elastic and consequently very little damage is observable. As the level of overstressing increases, the observed damage (i.e., irrecoverable strain) should increase, leading to a zone of potential overbreak around the excavation. Increased overstressing beyond this level may cause stress-driven failures and eventually the rock mass may become unsupportable. Wiles (2001) proposed that this methodology could be incorporated into a comprehensive back analysis technique to assist in quantitative stope design (Figure 5.33b). Furthermore, the damage model assumes that the level of overstressing is a direct cause of an increase in σ1, while confinement is kept constant. In practice, the stress path experienced by a rock mass can vary (Figure 5.33c) with “excess stress” generated by any of the following: • A loss of confinement, for example, a stope wall or back (−∆σ3) • An increase in load, for example, a pillar or stope wall (+∆σ1) • A combination of both, typical of a stope block abutment failure (+∆τmax) Back analysis of stress-driven open stope damage is best undertaken for primary stopes, where a condition of minimal stress-induced damage prior to stoping can be assumed. Thus, the stress path in the immediate vicinity of

224

Geotechnical Design for Sublevel Open Stoping

Unsupportable σ-driven failure POB

σ1

Damage

(a)

Unsupportable

σ1

POB

σ-driven failure ε1

σ3-Confinement

Collapse σ1

Increasing damage

(b)

Unstable

Stable

σ1

∆σ3 ∆σ1

∆τmax

Undamaged

σ3

(c)

σ3

FIGURE 5.33 (a) Linear elastic stress damage model for monotonically increasing stresses, together with related strain damage. (After Wiles, T.D., Map3D course notes. Masters of Mining Geomechanics, Western Australian School of Mines, Mine Modelling Pty Ltd., Mount Eliza, Victoria, Australia, 2001, 124pp.) (b) Generalized damage model. (After Wiles, T.D., Map3D course notes. Masters of Mining Geomechanics, Western Australian School of Mines, Mine Modelling Pvt Ltd, Leinster, Western Australia, Australia, 2001, 124pp.) (c) Stress path overstressing. (From Cepuritis, P.M. et al., Back analysis and performance of block A long hole open stopes—Kanowna Belle Gold mine, in E. Eberhardt, D. Stead, and T. Morrison, eds., Rock Mechanics: Meeting Society’s Challenges & Demands, Proceedings of the 1st Canada—US Rock Mechanics Symposium, Vancouver, British Columbia, Canada, May 27–31, 2007, pp. 1431–1439, Taylor & Francis, Leiden, the Netherlands.)

the stopes may be attributed to the primary stope extraction sequence. The number, location, and orientation of large-scale geological discontinuities (Villaescusa and Cepuritis, 2005) must be also taken into account to facilitate the interpretation of the numerical modeling results (Cepuritis et al., 2007). Cepuritis et al. (2007) show example results of σ1 versus σ3 contoured by the depth of stope wall overbreak or the calculated depth of failure. The results were subdivided into regions based on the likely stress path experienced (Figure 5.34). For moderately jointed to massive rock masses (Figure 5.35), the onset of increased depth of failure shows good correlation with an estimated Hoek–Brown strength envelope (Cepuritis et al., 2007).

225

Span and Pillar Design

σ1

Monotonic Shear –45°

–15°

–15° Confined

∆σ1,∆σ3 Low confinement

In situ stress

–90° Unloading

–180° σ3

FIGURE 5.34 A stress path classification used in back analysis of stope wall overbreak. (From Cepuritis, P.M. et al., Back analysis and performance of block A long hole open stopes—Kanowna Belle Gold mine, in E. Eberhardt, D. Stead, and T. Morrison, eds., Rock Mechanics: Meeting Society’s Challenges & Demands, Proceedings of the 1st Canada—US Rock Mechanics Symposium, Vancouver, British Columbia, Canada, May 27–31, 2007, pp. 1431–1439, Taylor & Francis, Leiden, the Netherlands.)

More significantly, the depth of overbreak increases with overstressing, and progressively increases as the stress path changes from monotonic loading and shear, through to low confinement conditions. Increased falloff occurs under unloading conditions, particularly close to the stope-scale rock mass damage initiation criteria (Cepuritis et al., 2007). For highly fractured rock masses influenced by large-scale geological discontinuities, the overbreak generally occurs at lower stress levels, and the extent occurs over a wider range of stress conditions (see Figure 5.36, Cepuritis et al., 2007). 5.4.2  Nonlinear Numerical Modeling Nonlinear modeling of complex open stoping sequences can be undertaken using a nonlinear, general purpose, three-dimensional finite element analysis program such as the Abaqus Explicit (Beck and Duplancic, 2005). Abaqus is well suited to the analysis of mining problems where a potential exists for significant plasticity, complex extraction sequences, high levels of deformation, and large numbers of material discontinuities. Models required to represent global stoping sequences, large-scale geological discontinuities, and stope-scale structures are routinely implemented (Beck and Duplancic, 2005). Large-scale global models are constructed incorporating all stoping geometries including shafts, ramps, access development, and mine-scale geological discontinuities. Smaller, more detailed submodels are subsequently

226

Geotechnical Design for Sublevel Open Stoping

80

70

Estimated rock mass strength Monotonic loading

High confinement

σ1 (MPa)

60

50

Shear Rock mass damage initiation Stope wall Depth of failure (m)

40

30

Low confinement

< 2.5 m 2.5–3.0 m 3.0 –3.5 m

20

3.5 – 4.0 m 4.0 –4.5 m

10

–20

4.5–5.0 m

Unloading

–10

0

10

20

30

σ3 (MPa) FIGURE 5.35 Example of σ1 versus σ3 for moderately jointed to massive rock masses. (From Cepuritis, P.M. et al., Back analysis and performance of block A long hole open stopes—Kanowna Belle Gold mine, in E. Eberhardt, D. Stead, and T. Morrison, eds., Rock Mechanics: Meeting Society’s Challenges & Demands, Proceedings of the 1st Canada—US Rock Mechanics Symposium, Vancouver, British Columbia, Canada, May 27–31, 2007, pp. 1431–1439, Taylor & Francis, Leiden, the Netherlands.)

constructed in key areas, with strain outputs and tractions from the global models being used as the boundary conditions for the submodels. Modeling is specifically targeted at understanding rock mass response and the influence of stope-scale structures on stope wall performance (Cepuritis et al., 2010). Extraction sequences in a global model are implemented in approximately quarterly steps, while block-scale models are extracted in steps no larger than one stope at a time. Selected stopes are extracted and then filled sequentially. A large number of extraction steps are required to ensure that the stress path throughout an entire area of interest is captured. For the submodels, each stope can be extracted involving a number of intricate firings, usually consisting of (a) a full-height cutoff slot or approximately 10% of the final void,

227

Span and Pillar Design

Monotonic loading

80

70

σ1 (MPa)

60

50

High confinement

Subperpendicular to wall surface Sub-parallel to wall surface

Shear

40

30

Low confinement

σ1 – σ3 = 25 MPa Stope wall Depth of failure (m) 100 mm per step in the model, a 0.50 correspondence with observed falloff was determined. An example of the relationship between plastic strain during stope extraction (regardless of velocity) and the percentage of “unstable” points is shown in Figure 5.40. The data indicate that instability due solely to plastic strain only accounts for a maximum of around 25%–30% of observed instabilities. This highlights the importance of stope-scale structure, its role in instability, and its influence on the strain field itself. The criterion is a reasonable predictor of overall instability, with a peak probability of falloff of 0.15–0.2 at more than 5% plastic strain, which corresponds to extremely comminuted material, or crushed rock (Beck and Duplancic, 2005). Rock masses with this corresponding level of plastic strain would almost certainly unravel if unconfined and exposed on a stope wall. The correlations of instability with velocity and plastic strain are encouraging in terms of predictors of stope wall instability, and hence appear attractive as design tools (Cepuritis et al., 2010). Levels of instability can be predicted for a variety of stope geometries, layouts, and sequences by forward numerical analysis. Simplistically, points in the forward analysis that display large velocities are predicted to have a very high likelihood of being associated with instability. Points showing high levels of plastic strain, low

231

Span and Pillar Design

0.30

Probability of instability

0.25

0.20

0.15

0.10 Probability of instability = 0.686 (Plastic strain)0.452 2 R = 0.98

0.05

0.00 0.00

0.02

0.04

0.06

0.08

0.10

0.12

0.14

0.16

0.18

0.20

Plastic strain

FIGURE 5.40 Plastic strain values versus mining step. (From Cepuritis, P.M. et al., Back analysis of overbreak in a longhole open stope operation using non-linear elasto-plastic numerical modelling, Proceedings of the 44th US Symposium Rock Mechanics and Fifth Canada—US Rock Mechanics Symposium, Salt Lake City, UT, June 27–30, 2010, Paper ARMA 10-124, 11pp.)

levels of confinement, and that are exposed at a stope wall are expected to have a moderate chance of reporting as falloff.

5.5  Pillar Stability Analysis 5.5.1  Basic Concepts Pillar design and stability analysis is a critical component of the stope design process. Although the fundamental concepts of factor of safety as the pillar strength/average pillar stress ratio and pillar stability have been understood for some time, it is only more recently that the tools have become available to allow more quantitative analyses of pillar strength and stability to be carried out. In basic engineering mechanics terms, stability refers to the stability of equilibrium, or the ability of the overall structure, or an element of that structure (in the present case, a mine pillar), to undergo a small change in the equilibrium state of loading without producing a state of unstable

232

Geotechnical Design for Sublevel Open Stoping

equilibrium involving a sudden release of stored strain energy or large deformations (Salamon, 1970; Brady and Brown, 2004). This form of instability may lead to crushing and the total collapse of a pillar and, in some cases, its surrounds. In other cases, the peak load-carrying capacity of a pillar may be exceeded and it may show visible signs of having been overloaded, but it may retain some load-carrying capacity and continue to provide support to the mine structure without undergoing unacceptably large deformations. The analysis of pillar stability in these engineering mechanics terms is beyond the scope of this book. Here, the emphasis will be on the relationship of the average pillar stress to the pillar strength. The terms stability and instability will not always be used in the strict engineering mechanics sense, but may be used simply to indicate that the stress imposed on the pillar exceeds its strength. Early developments in empirical pillar design were dominated by contributions from room and pillar methods, particularly in coal mining. More recently, reliable hard rock empirical and numerical pillar design tools have become available and have been implemented in sublevel open stoping for the design of secondary stope geometries. In general, pillar strength and stability are controlled by a large number of factors that include structural geology, compressive strength and deformability of the rock mass, the pillar dimensions including the width/height ratio, the degree of confinement, the percentage extraction, and the quality of mining such as drilling and blasting. 5.5.2  Average Pillar Stress Using the Equivalent Area Approach The stress analysis approach to pillar design requires that the load acting on the pillar be determined using either analytical or numerical techniques. The average pillar strength must then be evaluated and the pillar strength/stress ratio can then be used to estimate pillar stability. The simplest approach to the evaluation of pillar stability uses the “equivalent pillar area” technique to estimate pillar loads. Figure 5.41 illustrates a typical square room and pillar layout used in mining horizontally bedded deposits. Assuming that the pillars shown are part of a large array of pillars and that the rock load is uniformly distributed over these pillars (Hoek and Brown, 1980), the average pillar stress, σp is given by 2



2

Ê Wo ˆ Ê Wo ˆ sp = sz Á Á1 + Wp ˜ ˜ = gz Á Á1 + Wp ˜ ˜ (5.10) Ë ¯ Ë ¯

where γ is the unit weight of the rock z is the depth below surface Wo and Wp are the widths of the opening and the pillar, respectively.

233

Span and Pillar Design

Plan area of pillar on surface Wp

Wo + Wp

Z

Wp

Wp Wo FIGURE 5.41 Load carried by a single pillar assuming total rock load to be uniformly distributed over all pillars. (After Hoek, E. and Brown, E.T., Underground Excavations in Rock, IMM, London, U.K., 1980, 527pp. With permission.)

The average pillar stresses for different pillar layouts are summarized in Figure 5.42 and, in all cases, the value of σp is given by the ratio of the weight of the rock column carried by an individual pillar to the plan area of the pillar. The analysis incorporates several significant simplifications and in practice its use is restricted to shallow flat-lying deposits of significant lateral extent. As such, it may be of limited use for most hard rock mine pillar analyses. Hence, this method must be used with caution in sublevel open stope design, as it can be very conservative. 5.5.3  Empirical Rib Pillar Stability Chart Hudyma (1988) analyzed data from rib pillars in a number of Canadian open stope mines and plotted this in terms of the Y-axis (normalized pillar load to material UCS) and X-axis (pillar width/height). The database

234

Geotechnical Design for Sublevel Open Stoping

Unit length

Wp

Wo

Rib pillars σp = γz (1 + Wo /Wp)

Rock column area

Wo + Wp

Pillar area

Irregular pillars σp = γz =

Rock column area Pillar area

FIGURE 5.42 Average vertical pillar stress in typical pillar layouts using equivalent area method—plan views. (After Hoek, E. and Brown, E.T., Underground Excavations in Rock, IMM, London, U.K., 1980, 527pp. With permission.)

incorporated a wide variety of rock types and pillar loads that were derived from three-dimensional linear elastic numerical modeling. The data showed that squat pillars under low stress were stable (lower right quadrant, Figure 5.43). Pillars become less stable as they move toward the upper left region. Hudyma divided the graph into three general zones: failed, transition, and stable. The database also included 13 case studies in which pillars were originally stable and subsequently yielded. These cases were observed to move correctly through the three zones on the graph. Hudyma also suggested that the graph could be used to predict pillar yield in open stoping design. 5.5.4  Confinement Pillar Stability Chart A pillar stability database was developed at Westmin Resources Myra Falls operations and was combined with seven existing pillar databases, four consisting of detailed information and three with limited information. Detailed databases included the Westmin Resources data, Hudyma’s database collected from 13 Canadian operations, a database from the Selbi-Phikwe mine in Botswana (Von Kimmelmann et al., 1984), and the Hedley and Grant (1972) database from the Elliot Lake district in Ontario. The three limited databases were from the Black Angel mine in Greenland (Krauland and Soder, 1987), from the Zinkgruvan mine in Sweden (Sjöberg, 1992), and from Brady (1977) from Mount Isa Mines in Australia. Each of the databases listed used some form of pillar stability classification. In order to bring these data to a common frame of reference, a simplified pillar stability classification scale was developed (Lunder, 1994; Lunder and Pakalnis, 1997). Pillar stability was classified as being stable, unstable, or failed. The classification methods used for the combined database ranged from a six-level classification quantifying various levels of pillar instability to a more limited classification identifying only stable, sloughing, or failed

235

Span and Pillar Design

Open stope rib pillar data

0.60

Stable Sloughing Failure

Pillar load / UCS

0.50

0.40

0.30

0.20

0.10

0.00

0.0

0.4

0.8

1.2

1.6

2.0

2.4

Pillar width/pillar height FIGURE 5.43 Pillar stability graph—stable, transition, and failed zones. (After Hudyma, M., Development of empirical rib pillar failure criterion for open stope mining, MASc thesis, Department of Mining and Mineral Processing, University of British Columbia, Vancouver, British Columbia, Canada, 1988.)

conditions. Figure 5.44 is a schematic illustration of the pillar stability classification method developed for use at the Myra Falls mine. Pillar classifications of 2–4 represent an unstable pillar classification for the combined database. Table 5.2 describes the criteria used at Myra Falls to make an assessment of the pillar stability classification. Figure 5.45 shows the excellent rock mass conditions for typical class 1 pillars. The average pillar stresses considered in this analysis were predominantly calculated using linear elastic numerical modeling with the exception of Hedley and Grant (1972), who used tributary area theory. Pillar strength was presented in a general form as shown in Equation 5.11. This equation is divided into two general terms, the first representing the strength of the intact pillar and the second representing the effect of pillar shape on pillar strength:

Ps = Size ¥ shape (5.11)

where Ps is the estimated pillar strength (MPa), size is a strength term that incorporates the size effect and the strength of the intact pillar material (MPa), and shape is a geometric term that incorporates the shape effect of the pillar.

236

Geotechnical Design for Sublevel Open Stoping

Opening

Opening Class 1

Class 2

Opening

Opening Class 3

Class 4

Opening Class 5

FIGURE 5.44 Schematic illustration of the pillar stability classification method developed for use at Westmin Resources Ltd. (After Lunder, P., Hard rock pillar strength estimation: An applied empirical approach, MASc thesis, University of British Columbia, Vancouver, British Columbia, Canada, 1994, 166pp.)

TABLE 5.2 Visual Assessment of Pillar Stability Pillar Stability Classification 1 2 3

4 5

Observed Pillar Conditions No sign of stress-induced fracturing Corner breaking up only Fracturing in pillar walls Fractures  5 mm,  10 mm Fractures through pillar core

Source: Lunder, P., Hard rock pillar strength estimation: An applied empirical approach, MASc thesis, University of British Columbia, Vancouver, British Columbia, Canada, 1994, 166pp.

Span and Pillar Design

237

FIGURE 5.45 Excellent rock mass conditions: Example of Class 1 pillars (MRM, Northern Territory).

Two formulae that can be used for the estimation of pillar strength were developed by Lunder (1994) including the “log-power shape effect formula” and the “confinement formula.” Both formulae are virtually identical when plotted on a stability graph. However, the difference is that the “log-power formula” is a purely empirical formula, while the “confinement formula” is a modified form of the Mohr–Coulomb failure criterion (Lunder, 1994). Both formulae use the average pillar confinement term as subsequently described. The combined database was analyzed in order to determine if any past methods could be applied successfully to the combined database. It was determined that these historical methods could not adequately represent the combined database over the full range of pillar width/height ratios (Lunder, 1994). Individual linear shape effect constants were derived for each of the databases described earlier. These values enabled the assignment of a strength factor that is used to correct (i.e., scale) the unconfined compressive strength of intact pillar material to the full-size unconfined compressive strength of the pillar. This value is the “size” term in Equation 5.11, where the full-size unconfined compressive strength of a mine pillar can be represented by ≈44% of the unconfined compressive strength of the intact pillar material (Lunder, 1994). Pillar strength has been related to the pillar width/height ratio extensively in the past. However, the strength of a rock mass is known to be a function of the applied and the confining stresses. Using two-dimensional elastic boundary element modeling, it was determined that a relationship exists between the pillar width/height ratio and a term called the “average pillar confinement” and represented by the symbol Cpav. The average pillar confinement is defined as the ratio of the average minor pillar stress (σ3) and the

238

Geotechnical Design for Sublevel Open Stoping

average major pillar stress (σ1). These values are measured at the mid-height of the pillar. Equation 5.12 is the relationship that was determined to relate the pillar width/height ratio and the “average pillar confinement.” The value of “coeff” in Equation 5.12 is dependent on the extraction ratio in the vicinity of the pillar. For typical extraction ratios in underground hard rock mines of 70–90%, a value of 0.46 for “coeff” has been determined to be acceptable with less than 10% error (Lunder, 1994): 1.4



Cpav

È Êw ˆ˘Êw ˆ = coeff Ílog Á + 0.75 ˜˙ÁË h ˜¯ (5.12) ¯˚ Î Ëh

where Cpav is the average pillar confinement coeff is the coefficient of pillar confinement and set to 0.46 w is the pillar width (m) h is the pillar height (m) A modified strength formula, “the confinement formula,” that resembles the Mohr–Coulomb strength criterion was determined by Lunder (1994) to represent the combined database with a prediction success that has a slightly higher predictability rate (87% versus 85%) than the “log-power” formula. The “confinement formula” is represented by Equation 5.13. Empirical constants representing rock mass properties have been determined for C1 and C2 to be 0.68 and 0.52, respectively. This method is presented graphically along with all of the case histories in the combined database on Figure 5.46. The fundamental difference between the “log-power formula” and the “confinement formula” is that the latter is based on the theory of the strength of a rock mass, while the “log-power formula” is a purely empirical formula for which curve-fitting parameters have been determined. Pillar strength in the “confinement formula” is driven by the mine pillar friction term “kappa,” as defined in Equation 5.14, which is a function of the applied and confining stresses on the pillar only:

Ps = (k sc ) (C1 + C2 kappa ) (5.13)

where Ps is the pillar strength (MPa) k is the pillar size factor = 0.44 σc is the unconfined compressive strength of the pillar material (in MPa for a 50 mm diameter sample) C1 and C2 are empirical rock mass constants (0.68 and 0.52, respectively) kappa is a mine pillar friction term, calculated as follows:

239

Span and Pillar Design

0.7

F.S. = 1.0

Average pillar stress/UCS

0.6 0.5

F.S. = 1.4

0.4 0.3 0.2 Stable Sloughing Failure

0.1 0.0

0.0

0.4

0.8

1.2

1.6

2.0

2.4

2.8

3.2

Pillar width/pillar height FIGURE 5.46 The confinement formula stability graph plotted with all case histories from the combined databases. (From Lunder, P., Hard rock pillar strength estimation: An applied empirical approach, MASc thesis, University of British Columbia, Vancouver, British Columbia, Canada, 1994, 166pp.)



È Ê1 - Cpav kappa = tan Ícos -1 Á Á1 + Cpav Í Ë Î

ˆ˘ ˜ ˜˙ (5.14) ¯˙ ˚

where Cpav is the average pillar confinement and is defined by Equation 5.12. The lines dividing each of the pillar stability classifications have been assigned a factor of safety. This assignment is based upon the assumption that the line dividing the unstable and failed pillars has a factor of safety of 1.0. Using this as a baseline, it was determined that the transition from unstable to stable pillar conditions would have a calculated factor of safety of 1.4 (Lunder, 1994). In order to use the design guidelines developed with confidence, the method must be calibrated to existing conditions. Calibration is accomplished through the observation of existing pillar conditions and calculated stress values. If the observed pillars do not fall in the correct region on the pillar stability plots, modification to the input parameters is required. The modification can either be to the values that are used as input for stress determination (the in situ stress values) or to the unconfined compressive strength of the intact pillar material such that the pillars used for calibration fall in the correct region on the pillar stability plots. Figure 5.47 shows Lunder’s Canadian database and over 50 points from the McArthur River

240

Geotechnical Design for Sublevel Open Stoping

10.0

Lunder Stable Unstable Failure

UCS/average pillar stress

9.0 8.0 7.0

MRM Stable Unstable Failure FS1.4 FS1.0

6.0 5.0 4.0 3.0 2.0 1.0 0.0

0

0.5

1

1.5

2

2.5

3

3.5

Pillar width/height ratio FIGURE 5.47 Pillar stability graph—Lunder and MRM data. (From Schubert, C.J. and Villaescusa, E., An approach to hard rock pillar design at the McArthur River mine, Proceedings of the AusIMM Annual Conference—The Mining Cycle, Mount Isa, Queensland, Australia, April 19–23, 1998, pp. 255–259, AusIMM, Melbourne, Victoria, Australia. With permission.)

Mine (MRM) in Australia (Schubert and Villaescusa, 1998). The MRM results confirm the generality of the method. The data suggest that for a ratio of σc/σp less than 2, the majority of the pillars are unstable, regardless of the pillar W/H ratio. Furthermore, when the σc/σp ratio is greater than 5, even slender pillars are stable. This supports the changes suggested earlier to the factor A in Figure 5.12. 5.5.5  Numerical Modeling for Pillar Design Both three-dimensional linear elastic and nonelastic numerical models can be used for pillar design. For linear elastic analysis, the three-dimensional stoping geometries can be represented in almost any required detail incorporating sequencing. Elastic models are generally run as single material models as incorporation of multiple geological materials generally has limited effect on the final stress outcome. Some models allow inclusion of a limited number of major geological discontinuities. The programs MAP3D (Wiles, 2006) and Examine3D (Rocscience Inc, 1990) are typical of the three-dimensional elastic numerical analysis software available. Output from such models is generally relatively straightforward to interpret with contours of principal stress and factor of safety often displayed (Figure 5.48).

241

Span and Pillar Design

While three-dimensional elastic models provide a reasonable representation of the stress redistribution resulting from a stoping process, many pillars are subject to varying degrees of failure, particularly at the exposed pillar faces, and resultant stress redistribution to the pillar core cannot be simulated unless nonlinear models are used to analyze the responses σ1 (MPa) 100.0 90.0 80.0 70.0 60.0 50.0 40.0 30.0 20.0 10.0 0.0

400 Level 380 Level 360 Level 340 Level 320 Level 300 Level

σ3 (MPa) 50.0 45.0 40.0 35.0 30.0 25.0 20.0 15.0 10.0 5.0 0.0

400 Level 380 Level 360 Level 340 Level 320 Level 300 Level

(a) FIGURE 5.48 Major principal stress and strength factor for Eloise Deeps mine 30 m wide pillar using the program MAP3D. (a) Longitudinal view of major and minor principal stresses.

242

Geotechnical Design for Sublevel Open Stoping

Strength Factor-A 1.00 1.10 1.20 1.30 1.40 1.50 1.60 1.70 1.80 1.90 2.00

Average pillar SF-A= 1.7

400 Level 380 Level

2.8 m

UCS = 48 = 45°

6.5 m

360 Level 340 Level 320 Level 300 Level

Strength Factor-A

1.00 1.10 1.20 1.30 1.40 1.50 1.60 1.70 1.80 1.90 2.00 UCS = 48 = 45°

N

(b) FIGURE 5.48 (continued) Major principal stress and strength factor for Eloise Deeps mine 30 m wide pillar using the program MAP3D. (b) Longitudinal and plan view of Strength Factor-A.

of yielding pillars. Consequently, one of the most significant improvements in mine design has come from a move toward calibrated, multiscale, nonlinear numerical modeling. Gross deformation simulated at a global stope sequencing scale can be used to provide the boundary conditions for a smaller, stope length scale model that incorporates more detailed material properties incorporating discrete fracture networks (Beck et al., 2010).

Span and Pillar Design

243

Massive, strain softening, dilatant analysis is often used for multiscale stope design and analysis. The greatest improvement has been the rationalization of the use of submodels, which have the ability to correctly replicate observed displacements at all length scales. An immediate consequence is the ability to use velocity and displacement as criteria for instability (see Section 5.4.2). The mechanisms of damage and deformation that affect stability at each stoping sequence can then be successfully captured. Currently, nonlinear three-dimensional modeling can be conducted using various commercially available finite element codes such as the program Abaqus. Other specialized codes such as FLAC3D (three-­dimensional finite difference) and 3DEC (three-dimensional distinct element) are also available. Each of these techniques can be very useful depending on the specific problem to be solved. For example, if the nature of the problem involves slip on major geological structures intersecting the pillar, then a distinct element program such as 3DEC may provide the appropriate analysis tool. In cases where general plasticity (crushing) failure is dominant, a continuum three-dimensional code may provide the most appropriate analysis tool. In all cases, however, the key to successful prediction of rock mass behavior is the ability to quantify rock mass failure and its behavior after failure. The selection of a realistic nonlinear constitutive model to provide the relation between the stresses and strains that can be sustained by a fractured hard rock mass is required. However, a detailed development of such a topic is beyond the scope of this book.

6 Drilling and Blasting

6.1 Introduction Drilling and blasting in sublevel open stoping involves the interaction of the rock mass, the drillhole patterns, the explosives types, and the initiation sequences. The performance is measured in terms of safety, rock fragmentation, muckpile characteristics, stability of the exposed stope walls, and damage to nearby areas and equipment (Figure 6.1). The objective of the blast design process is to determine the number, position, and length of required blastholes with respect to the available development and the stope boundaries, while taking into account the orebody shape, ground conditions, groundwater, available equipment, stope access geometry, hole size, and the explosive types. In addition, the economic objective is to achieve the desired fragmentation (with minimum damage to the exposed stope walls and stope accesses) by means of a minimal use of explosives, materials, and time. Damage to the surrounding areas such as dented ventilation fans, ripped ventilation bags, dislodged and broken service pipes, and electrical cables must be avoided. Furthermore, the consequences of coarse fragmentation range from hung-up drawpoints and excessive secondary breakage to difficult mucking, loading, and tramming, causing increased maintenance costs on trucks and loaders. The effects of undersized fragmentation are excessive fines, overloaded equipment, and milling problems.

6.2  Longhole Drilling Sublevel open stoping requires the accurate and efficient drilling of relatively long blastholes within a designed stope boundary. Depending upon the rock mass conditions and stoping geometry, ring drilling may involve upholes, downholes, one-sided rings, and full 360° rings in vertical, inclined, or horizontal planes. Drilling is achieved by percussion mechanisms and adequate feed pressure, with bit penetration resulting from localized crushing and 245

246

Geotechnical Design for Sublevel Open Stoping

In situ block size distribution Geological discontinuities Intact rock bridges + Blast energy

Fragmentation

Gas expansion Vibrations Drilling accuracy Explosive strength Confinement Stand-off distance

Muckpile shape, looseness, and muckability +

Damage

Instability, dilution, airblast

FIGURE 6.1 The drilling and blasting process in sublevel stoping.

chipping at the rock–bit interface. In addition, rotation is required to change the button position within the toe of the hole following each percussive impact of the striker bar on the drill string or bit. Finally, flushing is required to remove the rock cuttings and also to cool the drilling tools (Puhakka, 1997). Drilling for longholes in sublevel stoping involves either top-hammer or in-the-hole (ITH) drilling mechanisms. In a top-hammer configuration, the rock drill or drifter remains on the top of the drill string, requiring transfer of the impact energy from the drifter through the entire drill string to reach the bit. In ITH drilling, the impact mechanism is located directly above the bit and enters the hole as the first piece of the drill string (Hamrin, 1993). Therefore, the impact energy is transferred over a shorter distance of the drill string prior to reaching the rock–bit interface. The minimum drillhole diameter and the required drilling accuracies determine which type of drilling configuration is suitable for each application. Percussion drilling is restricted by the ability of a drill steel to transmit energy. Drill stems are likely to deteriorate when subjected to excessive energy during impact force transmission. Consequently, percussion pressure settings must be established considering penetration rates and drill steel economy. Optimal feed pressures can be determined for a particular rock type following observations of penetration rates, bit wear, and steel threadwear (Puhakka, 1997). Excessively high feed forces do not necessarily

Drilling and Blasting

247

achieve increased penetration rates (Schunnesson and Holme, 1997). One problem experienced with excessive feed forces during drilling is bending of the drill steels resulting in increased drillhole deviation. 6.2.1  Top-Hammer Drilling Top-hammer drilling relies on the transferal of percussive energy (torque and impact) to the rock–drill bit interface via the drill stem. This energy is generated by a piston in the rock drill using pneumatic or electrohydraulic means. The drill bit contains no moving parts and simply screws onto the drill rod end. The rate of bit penetration is a function of the transferred impact force, the blow frequency, rotation speed, and the flushing efficiency (Puhakka, 1997). Energy losses along the drill string increase with hole depth, thereby reducing penetration rates. The hole diameter for top-hammer production hole drilling ranges from 51 to 127 mm with the hole length limited to 50 m (using a 127 mm hole diameter) due to the weight of the drill string and storage capacity of the tube magazine (Hamrin, 1993). In most cases, however, the hole length is usually restricted to less than 35 m due to limitations in hole drilling accuracy. Top-hammer rigs have drifters that are suitable for a small range of hole diameters and a typical rig is only capable of covering a spread of 50 mm between the minimum and maximum hole diameters. In order to drill a different-sized hole, a change of drifter as well as a change of drill string and hydraulic pumps may be required. 6.2.2  In-the-Hole Drilling In this drilling method, the percussive hammer is located inside the hole directly above the bit. The drilling bit is a continuation of the shank on which the drill piston impacts directly. Consequently, little energy is lost during the drilling process and penetration rates are almost constant regardless of hole depth. ITH drilling is typically only applicable to larger-diameter blastholes due to the space required to house the in-hole striker element and the increased drill string diameter. Drilling directions are logistically limited to subhorizontal to vertical downholes due to the inherent difficulties of charging explosives into large-diameter upholes. The main advantage of ITH drilling of longholes is improved hole accuracy compared with top-hammer drilling. This is very important in sublevel open stoping where the ability to accurately drill long, large-diameter holes allows for greater distances between sublevels, thereby reducing the costs of stope development access. Commonly used hole sizes for ITH drilling range from 85 to 215 mm, with holes extending up to 60 m in length. A disadvantage of ITH drilling is that low penetration rates, compared with the top-hammer technique, are likely to be achieved. In addition, the need for a large separate compressor results in reduced equipment mobility. Nevertheless, ITH drilling is the

248

Geotechnical Design for Sublevel Open Stoping

only technique capable of drilling very longholes with satisfactory accuracy. Another advantage with ITH drilling is that all the specified diameters can be drilled using one drill rig as the ITH hammer can be exchanged for a hammer of the required diameter and the existing drill string is retained. 6.2.3  Drilling Equipment Selection Considerations of the general mine layout including any special drilling needs are required during equipment selection. The equipment must be mobile and versatile as it is likely that it will perform a number of tasks while traveling to different locations in a reasonable amount of time. Typical tasks may include drilling holes of varying lengths, multiple diameters, different dip and dump angles, and upholes or downholes. In all cases, the selection of the stiffest rod–bit combination within a drill steel is critical to minimize hole deviation. Table 6.1 shows some suitable combinations of bit and rod diameters for longhole drill strings for production drilling in sublevel open stoping. Additional capabilities that require consideration during rig selection include crawler or wheel-mounted carriers and the selection of a boom capable of drilling a full 360° ring while tilting backward and forward. Other considerations are the selection of a feed system that can provide an adequate and smooth feed force at all feed pressures (to ensure that straighter holes are drilled), selection of a suitable rod changer, drill bit type, shape and cutter configuration, drill rod types and couplers, and any additional drill string stabilizing elements such as tubes or guides. One important operational factor is the flushing velocity of the air/water/ oil required to remove rock fragments from the face of the drill bit and propel them out of the drillhole. The ITH drilling system uses large-diameter tubes TABLE 6.1 Selection of Drill String Combinations for Longhole Drilling Hole/Bit Diameter (mm) 51 64 73 76 89 102 115 127 140 165

Rod Diameter (mm)

Tube Diameter (mm)

32 38 38 45 51 – – – – –

– – – 64 76 85 89 89 115 115

249

Drilling and Blasting

resulting in small apertures between the tubes and the wall of the hole. Given that a constant volume of air is pumped down the string to operate the hammer, high air velocities with excellent flushing capabilities are achieved. On the other hand, top-hammer drilling utilizes small-diameter strings resulting in low flushing velocities due to the large aperture between the drill steel and the wall of a hole. However, if a drill string composed of drill tubes is used, the flushing capabilities of a top-hammer system can be increased. In addition, drillhole deviation can be minimized. For similar diameter holes, the initial cost of a top-hammer drill rig is usually higher than that of an ITH drill rig. However, for short-length, smalldiameter holes, the high productivity achieved with the top-hammer drill rig ensures that it remains competitive. In summary, the decision on which method will be used depends on many factors, some of which may be sitespecific. Usually the depth of the holes and required accuracies are primary considerations, with ITH drilling preferred for holes exceeding 35 m in length. Conversely, for short-length, small-diameter holes, top-hammer drilling is well suited. 6.2.4  Drilling Deviation

Drill deviation (equivalent blasthole diameters, m)

Blasthole deviation is defined as the difference between the designed path of a drillhole and its actual trajectory. The total deviation from a planned drillhole location can be attributed to three factors. These are incorrect collar positioning, drill alignment error, and ITH deviation from a planned trajectory (Figure 6.2). The extent of each of the three sources of error depends upon the rock properties and geometry of the blast, type of drilling equipment, drill bit 25 20

l ota

T

15

Bending error

10 Setup error

5 0

or

err

Collaring error 0

50 100 150 200 250 Drilled depth (equivalent blasthole diameters, m)

300

FIGURE 6.2 Drill deviation types in longhole drilling. (After Heilig, J., Blast engineering—course notes for the masters of engineering science in mining geomechanics, Western Australian School of Mines, Curtin University of Technology, Kalgoorlie, Western Australia, Australia, 1999.)

250

Geotechnical Design for Sublevel Open Stoping

and rod specifications, and drill operation parameters (Kleine et al., 1992). The first two types of error are usually random in nature, and can be minimized by adequate markup and drilling procedures. ITH deviation is the bending of the holes as they are drilled and is a function of the forces acting on the drill strings and the drill string flexibility. This third type of error can compound the effect of either or both of the previous error types leading to an aggregate total error greater than any of the three alone (Kleine et al., 1992). 6.2.4.1  Collar Positioning

2.0 1.5 1.0

North (m)

A collar position error arises from the inaccurate location of the drill rig prior to drilling. Usually, the drive centerline and ring positions are marked on the backs or walls of a drilling drive either by the mining survey department or by the drillers. The collar positions of the holes within each individual ring can be painted on the floor, walls, or back and spaced out using a tape measure. The rig is then positioned between the ring markings and drilling is undertaken over the marked collars. For modern computer-controlled drill rigs, drill collar location control can be maximized by positioning the drill rig at a designed pivot point within the drilling drive. The hole collar positions are then identified by a hole dip and dump angle as drilled from the specified pivot point. Errors in hole collaring are independent of the hole diameter, length, and drilling equipment used. The errors can be determined by comparing the actual collar locations with the planned collar locations (Figure 6.3). In this figure, the planned locations of the hole collars are represented by the intersection of the two axes, while the actual collar locations are represented by each of the points in the plot. The data suggest a smaller error in the north– south direction than in the east–west direction. In this particular case, errors in a north–south direction are minimized by marking each ring position on both of the drill drive walls. Errors in an east–west direction are incurred by the poor location of each individual collar within the ring. Therefore,

0.5 –2.0

–1.5

–1.0

–0.5

–0.5 –1.0 –1.5 –2.0

FIGURE 6.3 Collaring error due to poor drill rig positioning.

East (m) 0.5

1.0

1.5

2.0

251

Drilling and Blasting

good drilling surfaces and the accurate marking of each individual collar is required to minimize such errors. Quality control during the drillhole design process and drill setup is the simplest way to reduce collar-positioning errors. However, this is potentially the most difficult solution to implement consistently, as it depends upon the attitude and work procedures followed by the drillers. Setup errors are increased by driller boredom and compounded when the drillers are paid large meterage bonuses. Having a quality component as part of the wages has been known to reduce this type of error, as drilling inaccuracies can be considered a symptom of a “people problem,” usually caused by an underlying management problem. Design issues when using computer-controlled drill patterns have also been identified (Fleetwood, 2010). If the actual floor elevation is different from that used in the design, hole collar location errors are increased. This is typical of when the floor of the drive is loose from overblasting of the lifters during development and the floors are cleaned up prior to drilling to make for easier collaring. This change in floor elevation is not taken into account in the drillhole design which used the initial drive laser surveys for collar location designs. 6.2.4.2  Drillhole Alignment Drill alignment error arises during the siting of the drill boom such that the initial orientation of the drillholes does not match the design. A change from design in either bearing or plunge will cause drill deviation that will increase as the drill path progresses. This error can be detected either by monitoring the initial drill setup angles or by calculations from down the hole survey measurements within the initial 2.5 m from the hole collar. Little or no in-hole deviation would be expected to occur in the first 2.5 m due to bending of the drill string. The alignment error can be calculated as the solid angle between the planned bearing and plunge and the surveyed bearing and plunge (Figure 6.4). Z (Elev) Planned (P)

Drilled (D)

δ

Collar

p

d

βp

Px = cos Py = cos Pz = sin

p p

sin βp cos βp

p

Y (North) βd

X (East) FIGURE 6.4 Solid angle between a drilled path and a designed path.

Dx = cos Dy = cos Dz = sin

d d d

sin βd cos βd

252

Geotechnical Design for Sublevel Open Stoping

The solid angle δ can be calculated from the dot product of the unit vectors of the direction cosines of the planned and the drilled holes (see Equation 5.3). Extensive surveying data from a number of typical bench stoping operations in Australia indicate that average solid angles of about 2° are typical. The estimated deviation due to ring misalignment from such a solid angle is a very significant ±3.5%. Experience suggests that in addition to incorrect drill positioning, uneven drilling surfaces also contribute to this type of error. Errors in azimuth are related to errors in burden, which can be minimized with the use of ring laser alignment. Plunge misalignment relates to deviations in toe spacing, where boom kickback when collaring also contributes to the error. In general, boom stability can be improved with the use of sufficiently long “stingers” capable of reaching both the floor and the back of the drilling drives or by a “horseshoe” stabilizer on the drill boom. The use of electronic pendulums or digital tilt-meters can help to monitor the alignment of the boom while drilling (Hamrin, 1993). Conventional ring drilling alignment consists of aligning the drill boom by eye, to a pair of paint marks on each side of the drill drive defining the planes of the rings. This technique is subject to markup, and setup errors and deviation from a design plane could be high as different drillers line up the drill rig from slightly different positions. An alternative is to set up a longitudinal alignment technique in which laser beams are used to locate the drill rig parallel to the drill drive center line (Figure 6.5). A set of suitable targets are accurately surveyed into position at each end of the drilling drives, allowing rig alignment by means of the laser beams. Hole survey results indicate that the bearing alignment error can be reduced by up to 5° using this technique. Additional lasers can be added perpendicular to the spine of the rig for accurate alignment with the plane of the ring as specified by the wall markups. 6.2.4.3  In-the-Hole Deviation ITH deviation is related to the bending of drillholes and occurs when the bit deviates from a straight path as it drills through a rock mass. Bending of

Ring line Laser beam Filled stope

Drill rig Targets

FIGURE 6.5 Longitudinal drill rig alignment using lasers. (Courtesy of Mount Isa Mines, Mount Isa, Queensland, Australia.)

253

Drilling and Blasting

holes is a function of noncontrollable factors (rock properties and geological features) as well as drill operation parameters such as thrust and torque and rod and drill bit specifications (Kleine et al., 1992). Rod deviations are caused primarily by the forces acting on the drill string and also due to the drill string flexibility. The flexibility depends upon the rod stiffness, which is a function of the physical makeup and the active length of the drill string to diameter (l/d) ratio. As drill string flexibility increases and/or the annulus (area difference between rod and drill bit) increases, a greater possibility of ITH deviation arises. In the early stages of drilling, drill string flexibility (related to the ratio l/d) is low. As drilling progresses and the ratio increases, so too does the flexibility, and increased bending is likely. In addition, more flexible strings will offer less resistance to side-loading changes on the bit. This can occur when the bit drills across rock types having different strengths or stiffnesses. Each change on side loading causes the bit to drill off in a slightly different direction, thus contributing to deviation. Consequently, to minimize ITH deviation, it is important to use stiff rods to prevent flexing as well as a suitable choice of rod and drill bit combination. For a given hole length and drill bit diameter, smaller diameter rods have more space within which to flex in a drillhole. Hence, it is expected that a T45 speed rod and a 76 mm bit combination would drill straighter holes than a T45 rod using an 89 mm bit. In both cases, the flexibility is the same; however, the T45 rod–76 mm bit combination has a smaller annulus, thereby reducing in-hole deviation. One problem with a smaller annulus is that rod couplings may become entrapped if the bit wears down or rocks fall behind a coupler. Flexing of the drill string can also be reduced with the use of tube drilling technology. Hence, a combination of 64 mm tube and 76 mm bit is expected to deviate less than a T45 rod–76 mm bit combination due to the increased stiffness provided by the tubes. The ITH deviation magnitude and orientation can be calculated using a three-dimensional vector analysis. ITH deviation is analyzed by considering both the planned drillhole trajectory and the actual surveyed path as follows (see Figure 6.4). P is the unit vector in the direction of the planned hole whose direction cosines are Px, Py, and Pz as defined by Figure 6.4. Let the coordinates of a surveyed point S along a drilled hole be

S = (Xs , Ys , Zs ) (6.1)

A general vector on a planned hole is given by where t is a real parameter.

T = (tPx , tPy , tPz ) (6.2)

254

Geotechnical Design for Sublevel Open Stoping

The distance of the surveyed point S to the planned hole is the length of the vector

TS = (Xs - tPx , Ys - tPy , Zs - tPz ) (6.3)

precisely when the vector TS is perpendicular to the planned hole, that is, when the dot product of the vector TS and the unit vector P is zero:

(Xs - Pl x , Ys - Pl y , Zs - Pl z ) i (Px , Py , Pz )= 0 (6.4)

This implies that

t=

XsPx + YsPy + ZsPz (6.5) Px2 + Py2 + Pz2

Therefore, the distance d from the surveyed point S to the planned hole, which is the ITH deviation, is given by

d = ÷ (Xs - tPx )2 + (Ys - tPy )2 + (Zs - tPz )2 (6.6)

where t is given by Equation 6.5. The average orientation of the deviation for each surveyed depth along a drilled hole can be estimated by a method suggested by Priest (1985). First, each surveyed point can be represented by a unit vector centered at the origin of the system of coordinates shown in Figure 6.4. The X, Y, Z coordinates of the terminal point of the ith vector are given by

Nix = cos jd sin bd



Niy = cos jd cos bd (6.7)



Niz = sin jd

where βd and φd are the trend and plunge of the drilled hole at the point of measurement. The X, Y, Z coordinates of the terminal point of the resultant or average hole deviation are given by







rx =

N

 Nix i =1

ry =

N

 Niy (6.8) i=1

rz =

N

 Niz i =1

255

Drilling and Blasting

The trend (βave) and plunge (φave) of the average deviation are given by Ê rx ˆ bave = arctan Á ˜+ q (6.9) Ëry ¯

and

Ê rz jave = arctan Á 2 Á (rx) + (ry)2 Ë



ˆ ˜ (6.10) ˜ ¯

where the term q is an angle that, depending on the sign of rx and ry, ensures that βave lies in the proper quadrant. If (rx ≥ 0 and ry ≥ 0), then q = 0 and when (rx  0), then q = 0, otherwise, q = π. ITH deviation calculated from surveyed data from a number of drill string and rod bit combinations can be analyzed for different hole lengths in order to determine the critical depth for each rod–bit combination. Figure 6.6 presents a comparison of average deviation with depth from downhole bench stoping using a top hammer Atlas Copco Simba H221 fitted with T38 and T45 speed rods. Each hole was surveyed at depths of 0 T38–73 mm T45–76 mm T45–89 mm

Depth (m)

–5

–10

–15

–20

0.0

0.1

0.2

0.3 0.4 0.5 Drill deviation (m)

FIGURE 6.6 Example of average deviation for different drilling strings.

0.6

0.7

0.8

256

Geotechnical Design for Sublevel Open Stoping

2.5, 5, 10, and 15 m. The rod deviations from the collar to a 10 m depth ­differed only slightly for the three combinations, with an absolute deviation of about 0.2 m. However, a significant difference was found at 15 m, where the straightest drilling string was T45 rods with a 76 mm bit. An average deviation of 0.28 m was observed with a standard deviation of 0.09 m. The worst performing combination was the T45 rods in conjunction with 89 mm bits that resulted in an absolute average deviation of 0.60 m with a standard deviation of 0.26 m. In comparison, an average collar position error of 0.24 m with a standard deviation of 0.15 m was determined for this particular drilling operation. In addition to the average deviation values, individual hole deviation distribution from a design target must be considered. Figure 6.7 presents a comparison of deviation distribution at a hole depth of 10 m for each of the drill string combinations. A definitive trend for the blastholes to deviate in an east–west direction (toe spacing) was found at this site and the T45–89 mm string combination was the worst performer.

0.8

0.8

∆Northing (m)

0.4

0.6 0.4 ∆Northing (m)

0.6

T38–73 mm at 10 m

0.2 0.0 –0.2

0.2 0.0 –0.2

–0.4

–0.4

–0.6

–0.6

–0.8 –0.8 –0.6 –0.4 –0.2 0.0 0.2

0.4 0.6 0.8

T45–76 mm at 10 m

–0.8 –0.8 –0.6 –0.4 –0.2 0.0 0.2

∆Easting (m)

∆Easting (m)

0.8 0.6 ∆Northing (m)

0.4

T45–89 mm at 10 m

0.2 0.0 –0.2 –0.4 –0.6 –0.8 –0.8 –0.6 –0.4 –0.2 0.0 0.2

0.4 0.6 0.8

∆Easting (m)

FIGURE 6.7 Drill deviation at 10 m hole depth from different drill string combinations.

0.4 0.6 0.8

257

Drilling and Blasting

Deviation North (m) 1.50 1.00 Deviation East (m)

0.50 3.00

–2.00

–1.00

1.00 Average

–0.50 –1.00 –1.50

2.00

3.00

Deviation at 8.5 m depth

Deviation North (m) 1.50 1.00 Deviation East (m) 3.00

0.50 –2.00

–1.00

1.00

–0.50 –1.00 –1.50

2.00

3.00

Average Deviation at 15 m depth

Deviation North (m) 1.50 1.00 Deviation East (m) 3.00

0.50 –2.00

–1.00

–0.50 –1.00 –1.50

1.00

2.00

3.00

Average Deviation at 20 m depth

FIGURE 6.8 Drill deviation for different hole depths. (After Cameron, A. and Paley, N., Assessment of blasting to reduce damage in B704 bench stope at Mount Isa Mines, in T. Szwedzicki, G.R. Baird, and T.N. Little, eds., Proceedings of the Western Australian Conference on Mining Geomechanics, Kalgoorlie, Western Australia, Australia, June 8–10, 1992, pp. 375–383, Western Australian School of Mines , Kalgoorlie, Western Australia, Australia.)

Figure 6.8 presents a comparison of average deviation with depth from downhole bench stoping using a top hammer Atlas Copco Simba H221 fitted with T38 speed rods. Each hole was surveyed at depths of 8.5, 15, and 20 m. The results show that for the 20 m-long blastholes there was a high probability of both excessively small and large toe burdens. A definitive trend for the blastholes to deviate in an east–west direction was also found at this particular site.

258

Geotechnical Design for Sublevel Open Stoping

Studies of hole deviation have shown that greater accuracy can be achieved by adding guide rods to a drill string or by using a tube string. Results from a study entitled the “straight hole” project carried out by Atlas Copco and the LKAB Kiruna iron ore mine attempted to quantify the effects of drillhole diameter on the expected deviation. The study was based on uphole drilling (drillhole length up to 50 m) using top-hammer drills in conjunction with tube strings (Hamrin, 1993). The results shown in Figure 6.9 can be used to determine the maximum drillhole length for a given diameter, where the target deviation for 95% of the holes does not exceed half the normal ring burden. It is important to note that the “straight hole” project guidelines are only applicable if the conditions associated with modern techniques of precision drilling apply. The maximum hole depth in Figure 6.9 may be achieved by a drill rig with appropriate angle instrumentation setup and by the use of a rigid tube drill system. In addition, a minimum collaring error of ±0.10 m and a rig alignment error of only ±1.0% have been assumed. As a comparison, Table 6.2 shows deviation data collected from 89 mm-diameter holes drilled with 64-mm diameter tube strings. Average bearing and plunge misalignments of 3.9° and 1.1°, respectively, were determined from the calculations. Another factor causing deviation is the effect of gravity on the bit. A pendulum effect may be experienced in longholes when gravity forces acting in a bit cause it to cut the bottom of the hole, gradually steepening the hole (Figure 6.10). Solutions such as increasing drilling thrust or placing stabilizing devices near the bit (to rotate it into the desired orientation) have been suggested to correct this problem. A negative offset in Figure 6.10a means that the hole has deviated north. The shallow holes are closest to the design, but tend to deviate to the north, perhaps due to drill setup error. The calculated burden in Figure 6.10b shows that all the holes begin within the correct plane, but get out of plane by the toe, especially the steeper holes.

6.3  Blast Design Parameters The dimensions of a blasthole pattern must be selected to suit the rock mass conditions, the geometry of the orebody, and the limitations of the drilling equipment. Blast patterns can then be adjusted to determine an optimal design for the different stope geometries such as production rings, cutoff slot (COS) holes, fill diaphragm, and trough undercut (TUC) rings. This process is based upon accumulated knowledge from previous experience in rock masses having similar strength and jointing conditions. The factors considered are the drilling access, the blasthole diameter and length, the burden and spacing, the explosive types, and the effects of timing and sequencing. The benefits achieved when a blast design is optimized include increased excavation stability, good fragmentation with reduced mucking (loader) unit

50

89

100

150

89 102 115 127 140 155 165

Hole diameter (mm)

57 64 76

ting

las ole b

gh

Lon

bla

Top hammer range ITH drilling range

h nc Be

0.50

1.00

1.50

2.00

0

(1.3 m)

10

Example: hole diameter: 89 mm Limit for hole spread (m) Estimated in-the-hole deviation: 2.5% Maximum hole depth: 35 m

30 Maximum hole length (m)

20

ed

1.0

%

40

50

Set-up and direction 1.0% of hole depth

hol

Collaring 0.1 m

the In

n atio evi

% 2.0

%

3.0

FIGURE 6.9 Maximum hole length using precision drilling. (After Hamrin, H., Precision drilling extends the range of longhole blasting, in G. Almgren, U. Kumar and N. Vagenas, eds., Proceedings of the 2nd International Symposium on Mine Mechanization & Automation, Luleå, Sweden, June 7–10, 1993, pp. 143–151, Balkema, Rotterdam, the Netherlands.)

1.00

2.00

(2.6 m)

3.00

4.00

ng sti

u um Ac c

r sp d lat e

0%

4. ea d

Nominal burden (m)

Drilling and Blasting 259

260

Geotechnical Design for Sublevel Open Stoping

TABLE 6.2 Hole Deviation for 89 mm Holes Drilled with 64 mm Tube Strings Hole ID

Depth (m)

Total Deviation (m)

In-the-Hole Deviation (m)

Bearing Misalignment (°)

Dip Misalignment (°)

R1-HW R1-FW R1-easer R2-FW R3-FW R3-easer Average

28.1 29.0 28.5 15.0 15.0 28.7 24.0

0.64 2.07 0.61 1.04 0.59 0.18 0.86

0.32 0.52 0.36 0.12 0.05 0.16 0.26

−3.17 10.06 1.74 7.81 −0.42 −0.34 3.92

0.78 −1.36 −0.15 1.97 −2.04 0.11 1.07

2606

2606

2596 Offset (m) 900 mm) is the standard dry AN prill and fuel oil mixture known as ANFO. ANFO consists of 1.5–3 mm-diameter AN prill coated in fuel oil at an optimum mixture of 5.7% fuel by weight. The detonation efficiency of ANFO varies with the percentage of fuel, where underfueling results in a greater reduction in output energy than overfueling. The typical loaded density of ANFO ranges from 0.82 to 0.95 g/cc based on pour loading or pneumatic loading, respectively. Several different modified formulations of ANFO are available for use in a wide range of applications. These include formulations having properties of reduced density, low fume production, water resistance, buffering against thermally active or chemically reactive ground, and high-strength breakageresistant prill for pneumatic loading. For mostly dry blasting conditions requiring an even distribution of shock and heave energy in medium- to large-diameter holes, ANFO is the preferred explosive choice. Due to the popularity of ANFO as the preferred explosive choice over several decades, the strength characteristics of other explosive formulations are regularly listed with reference to the standard strength of ANFO. These properties include relative bulk strength and relative weight strength. Due to the susceptibility of ANFO to water-induced explosive degradation, blasting in wet holes or where long sleep times are required is not recommended. Additionally, pneumatic charging of ANFO in large-diameter upholes (>89 mm) can result in excessive explosive loss due to fallout and therefore is advised only for subhorizontal to vertical downholes or smallerdiameter upholes. The lack of water resistance and cohesion of ANFO were two properties that prompted the development of other explosive formulations. These desired properties were drivers for the development of fluidbased bulk watergel and emulsion explosive formulations to aid in replacing ANFO for certain applications. Of the two types of explosives, emulsions have been developed more extensively for use in modern bulk delivery application. 6.5.4  Watergels or Slurries Ammonium nitrate watergels, commonly known as “slurries,” began development in the late 1950s (Du Pont, 1977), and consist of the same three

Drilling and Blasting

283

components as ANFO (oxidizer, fuel, and sensitizer). The phases and mixing procedures of the three components differ from ANFO, leading to increased water resistance and detonation characteristics. Slurry explosives contain oxidizer salts, fuels, and sensitizers dispersed in a continuous liquid phase. The addition of gelling agents or cross-linking agents retards the separation of the three components, controls the density and viscosity of the product, and adds water resistance to the mixture. The droplet size of the oxidizer in a slurry explosive is in the order of 0.2 mm (Bampfield and Morrey, 1984). The detonation characteristics of watergels are generally more efficient than those of ANFO because of the decreased size of the particles and the increased intimacy between the components. The method of oxidizer, fuel, and sensitizer suspension in watergels causes poor gap sensitivity and high sensitivity to changes in product and ground temperatures. For these reasons, watergel explosives are not used extensively in the modern mining industry, having been largely replaced by emulsion explosives. 6.5.5 Emulsions Emulsion explosives are similar to slurries in that the active components are suspended in a continuous liquid phase and are therefore water-resistant and highly pumpable. The differences between watergels and emulsions become apparent when reviewing the mixing process of the separate phases and the common sensitizing agents used in each type of explosive. The basic formula of an AN emulsion explosive is the suspension of small droplets of AN solution in a continuous oil (fuel) matrix. The droplet size of the AN solution in the emulsion matrix is on the order of 0.001 mm or less (Bampfield and Morrey, 1984). Common sensitizing agents used in bulk emulsion explosives are glass microballoons or gas bubbles formed by a chemical reaction within the emulsion after it is delivered into the blasthole. The required charging equipment, loaded densities, desired detonation characteristics, and storage and transportation requirements of each type of sensitized product generally determine which is the most suitable for particular applications. Each of these factors is closely linked to the method of product sensitization prior to or during blasthole loading. Under current explosives regulations, unsensitized emulsion is considered to be a bulk oxidizer much like agricultural fertilizer. Once the sensitizing agent is introduced, the emulsion becomes a blasting agent and is therefore subjected to more stringent storage and transportation regulations. The method of emulsion sensitization significantly influences the physical properties and the detonation characteristics. Microballoons generally yield an emulsion product that is more resistant to dynamic shock-induced desensitization and is better suited to close-in or highly confined blasting conditions. Microballoon-sensitized emulsions can also be sheared to change the rheology for loading into larger-diameter upholes or where a lower-viscosity

284

Geotechnical Design for Sublevel Open Stoping

Unconfined velocity of detonation (m/s)

4500 4000 3500

“Gassed” emulsion (0.85 g/cc)

3000 2500 2000

ANFO (0.85 g/cc)

1500 1000 500 50

60

70

80 90 100 Blasthole diameter (mm)

110

120

130

FIGURE 6.25 Comparison of ANFO and emulsion velocity of detonation.

product is required. In general, microballoon-sensitized emulsions have a high loading density (1.2–1.35 g/cc), which is not adjustable without the manual addition of low-density additives such as polystyrene or other organic bulking agents. When compared with ANFO, emulsions typically behave more as “ideal” explosives, having higher velocities of detonation and lower sensitivities to blasthole diameter (Figure 6.25). Additionally, the energy distribution within an emulsion explosive differs dramatically from that of ANFO, having a higher percentage of shock energy and a lower percentage of heave energy. Due to the reduced gas production of emulsion, the overall output energy can be less than that of ANFO even at a significantly higher charge density. Chemically sensitized or “gassed” emulsions are sensitized through the production of gas bubbles due to a reaction between chemicals added to the mixture immediately prior to or during pumping of the product into a blasthole. The rate and degree of gassing are regulated by the amount and injection location of the gassing agent or agents, the temperature of the product, the hole diameter, and the length of the explosive column. Once the emulsion is loaded in the blasthole, the chemical reaction takes place, causing the product to increase in volume and thus reduce in density. The desired density in the hole should be checked regularly during loading by performing cup density checks using standard testing practices. A wide range of in-hole product densities are available due to the easily adjustable amount of gassing agent injected. Product density ranges from 0.8 to 1.2 g/cc are common for chemically sensitized emulsions.

Drilling and Blasting

285

The fact that sensitization occurs upon loading into blastholes makes gassed emulsions a preferred product to reduce storage and transportation restrictions. The presence of free-forming gas bubbles in comparison to glass microspheres also makes gassed emulsion more susceptible to desensitization under shock conditions and largely unsuitable for highly confined blasting conditions where product sensitivity can be a concern. Additional concerns with gassed emulsions are the control of uncharged collar lengths due to mismanaged gassing rates or gassing agent amounts, quality control of the average charged density, product waste, and the variable in-hole density profile due to differential gassing deep in the column from the weight of the explosive product. 6.5.6  Special ANFO and Emulsion Blends Some customized products have been developed for specialty blasting conditions regularly experienced in underground stope blasting. These specialty products use modified formulations of existing products such as ANFO or emulsion to achieve specialized detonation characteristics. The most widely used specialty products in underground blasting include buffered explosives for resistance to thermally or chemically reactive ground and low-density products to reduce blast-induced damage or extraneous blasting vibrations in stope walls or outside the designed stope perimeter. Buffered explosive products typically include a chemical agent to reduce the sensitivity of an explosive to high temperatures or to reduce the reaction of the explosive with sulfides in the rock mass or groundwater. Excessive heat generated either through thermally active ground or through an exothermic chemical reaction between the explosive and the rock mass can lead to premature detonation of blastholes or malfunctioning of initiation systems or charge boosters. Low-density ANFO or emulsion products typically contain a low-density bulking agent such as polystyrene or other low-­density organic materials to reduce the in-hole charged density. The reduction in density and alteration of the detonation characteristics reduce the borehole pressure and the associated damage around a blasthole. Charge densities down to 0.3 g/cc are achievable in commercial low-density underground specialty products.

6.6  Explosive Placement Before placing explosive charges, blastholes are cleaned out using compressed air to remove any water, sludge, or drill cuttings to allow hole depths to be accurately measured. An ANFO hose can be used to both clean the holes and also measure the hole length. Prior to explosive charging,

286

Geotechnical Design for Sublevel Open Stoping

Nonel

Nonel Uncharged collar

Uncharged collar

ANFO

ANFO

Primer + detonator ANFO Powergel ANFO bag + powergel (a)

Primer + detonator ANFO bag (b)

FIGURE 6.26 Charged blasthole geometries in open stoping.

breakthrough holes must be blocked near the toe. To block breakthrough holes, high-energy emulsion explosive cartridges such as powergel can be placed in a plastic bag, tied together, and dropped down the hole by a strong chord that reaches the required depth. Additional powergel bags are then cut open along their axes and dropped down the holes to block the breakthrough as they split on top of the initial plug (Figure 6.26a). Alternatively, an empty ANFO bag or other material such as a ventilation bag is simply placed at the end of the ANFO hose or a rope and lowered to position just above the breakthrough depth (Figure 6.26b). Sticks or wedges attached to a rope or inflatable air bags may also be used to block breakthrough holes. To confirm the efficacy of the breakthrough blockage, the hole is checked for breathing (air flow at the hole collar) or explosive leakage at the toe. When used for charging downholes, ANFO products are either poured or blow-loaded, depending upon the hole diameter. The charging density (q) for pour-loaded ANFO is approximately 0.80–0.85 g/cm3. Small-hole diameters are typically blow-loaded to guarantee a consistent explosive density, as small pieces of rock may block the hole or create air pockets when pourloading. Similarly, all inclined holes, regardless of their diameter, are blowloaded. The charging density (q) for blow-loaded ANFO is approximately 0.90–0.95 g/cm3 due to prill breakage and increased compaction. Figure 6.27 shows uphole charging of ANFO within a typical longitudinal bench stope. Blind (nonbreakthrough) wet holes are typically left until last when charging with ANFO. If the water cannot be removed using compressed air or pumping, or if the hole will experience excessive sleep time prior to firing, pumped or cartridge emulsion products are used instead of ANFO. In some applications, holes are charged and not fired until adjacent filling operations

Drilling and Blasting

287

FIGURE 6.27 Bench stope blasting—blow-loading of ANFO using compressed air.

are under way. Although the sleep time of ANFO depends on the rock ­temperature, as a general rule charged holes should not be left unblasted for over 2 weeks. Emulsion explosive products typically have a much longer sleep time, unless there are reactive ground conditions. 6.6.1  Powder Factor Traditionally, powder factor is an indirect measure of the explosive energy being imparted to a rock mass per unit volume or weight blasted. It is calculated by dividing the weight of the explosives by the ring volume or the tonnage that is expected to be broken. Because ring blasting is a dynamic event and each rock mass is unique, the conventional definition of powder factor has limited applications other than being an index for comparisons on a global scale. The explosive quantity within each stope ring depends upon the following factors: • • • • • •

The number of meters drilled The blasthole diameter The explosive type The method of loading the holes (pour- or blow-loaded) The number of meters charged The tonnes broken

Typical powder factors for stope-blasting applications range from approximately 0.20 to 0.30 kg of explosive per tonne of ore for ring blasting, from

288

Geotechnical Design for Sublevel Open Stoping

0.20 to 0.50 kg/tonne for bench stoping, and from 1.4 to 1.8 kg/tonne for cutoff blasting. Due to excessive confinement at the toes of blasthole rings, higher localized toe powder factors are recommended to ensure adequate breakage. This can be achieved by using high-density emulsion explosives near the blasthole toes. 6.6.2  Energy Distribution Conventional powder factor calculations only provide a number and do not indicate the distribution of explosives within a ring design (Onederra and Chitombo, 2007). This is especially critical for radial drilling, where it can be very difficult to achieve an even distribution of explosives throughout a ring. In addition, larger-diameter blastholes give a poor distribution of explosives throughout the rock mass. The ring design described in Section 6.4.3 is an attempt to achieve an even distribution of explosives, since the powder factor is controlled by the burden and toe spacing chosen for each particular hole diameter. Staggered charging in adjacent ring holes also attempts to evenly distribute the explosive energy by minimizing the explosive concentration near the blasthole collars where the hole spacing is reduced. The conventional powder factor represents an average number over the blasted volume and is unable to identify regions having excessive energy such as collars near the hole. Research work at the Julius Kruschnitt Mineral Research Centre (JKMRC), Brisbane, Australia, has developed a technique to analyze powder factors within small cell areas, rather than a total area. The JKMRC QFRAG dynamic powder factor calculation can be used to analyze explosive distribution within rings and to determine regions where poor fragmentation may occur or where excessive damage is likely. The program QFRAG uses hole initiation timing to calculate the amount of breakage from each hole, and thus the energy required to achieve it. Because of detonation scatter, several simulations are required to obtain an average of powder factors from the analyzed design. The calculations are performed for a plane parallel to the ring, at one burden distance away from the blastholes. Figure 6.28 shows explosive distributions from a ring of 140 mmdiameter blastholes, drilled on a 3.5 m burden and 6 and 7 m toe spacings, respectively, and loaded with ANFO. The conventional design powder factors for the rings shown in Figure 6.28 were 0.28 and 0.25 kg/tonne for the 6 and 7 m toe spacing rings, respectively. The simulated initiation sequence was started from the bottom right with an MS #4 delay, and 3 or 4 holes per delay number were fired to give an approximate maximum charge weight of 600 kg/delay. The hangingwall holes were fired two numbers later than adjacent holes. Additional research in the early 1990s at the JKMRC saw the development of the computer program 3 × 3WIN, which is capable of calculating a 4D powder distribution by considering the influence of initiation sequence. The JKMRC 4D powder factor distribution tessellates points on a specified

289

Drilling and Blasting

6 m toe spacing

7 m toe spacing

Powder factor kg/tonne 0.250 FIGURE 6.28 JKMRC cell powder factor distribution for two ring designs.

plane using a distance weighting calculation that includes a weighting with respect to the time a deck detonates within the ring. Assumptions associated with the location and detonation time of decks are considered to calculate the 3D powder factor distribution and weighted using a factor called cooperation time. This is the time at which explosives from two decks in different holes interact on a portion of the rock mass (AMIRA, 1993). The effects of timing on explosive cooperation within a ring are shown in Figure 6.29.

6.7  Initiation Systems In the modern explosives market, two main categories of initiation system exist for underground development and production blasting applications. These two systems are categorized by the type of delay element contained within the detonator. The two main types of modern delay element are a controlled burning-front pyrotechnic element and an electronic computer chip. 6.7.1  Pyrotechnic Delay Element Detonators Pyrotechnic delay detonators are the most commonly used initiation systems in underground blasting. The well-known electric and shock-tube initiation systems contain pyrotechnic delay elements.

290

Geotechnical Design for Sublevel Open Stoping

Scale: kg/tonne 0.100 to 0.400 0.400 to 0.700 0.700 to 1.000 >1.000 (a)

(b)

FIGURE 6.29 (a) 3D powder factor distribution and (b) 4D powder factor distribution (cooperation time 35 s) from the JKMRC program 3 × 3WIN. (From AMIRA, P93E advanced blasting technology, Julius Kruttschnitt Mineral Research Centre Final Report (1990–1993), 1993. With permission.)

Pyrotechnic delay elements are composed of a controlled length of a pyrotechnic material having a highly controlled burn rate between the initiation line (downhole leg wires or shock tube) and the match head or primary charge. The delay within the detonator is therefore controlled by a physical burn process. The downhole timing accuracy of the detonator is controlled by the quality control of the manufacture of the pyrotechnic delay material and the length of the element. The timing accuracy of the entire pyrotechnic system must also consider any variations in burn time or charge transfer time for surface delays or connection mechanisms such as detonating cord or shock tube for shock-tube systems, or sequential firing boards for electric detonators. 6.7.2 Available Timing and Sources of Timing Error for Pyrotechnic Delay Elements The standard underground series of pyrotechnic detonators include two basic timing configurations. These two systems are long-period delay intervals (LP, Table 6.6) or millisecond delay series (MS, Table 6.7). Extended delay

291

Drilling and Blasting

TABLE 6.6 Comparison of Dyno Nobel NONEL LP and Orica Exel LP Delay Intervals Dyno Nobel NONEL LP Series Detonator No. 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

Orica Exel LP

Firing Time (ms)

Detonator #

Firing Time (ms)

Detonator #

Firing Time (ms)

25 500 800 1100 1400 1700 2000 2300 2700 3100 3500 3900 4400 4900 5400 5900 6500 7200 8000

0 ¼ ½ ¾ 1 1¼ 1½ 2 2½ 3 4 5 5½ 6 7 8 9 10 11

0 100 200 300 400 500 600 800 1000 1200 1400 1600 1800 2000 2250 2500 3000 3500 4000

12 13 14 15 16 17 18 19

4500 5000 5500 6000 6500 7000 8000 9000

Sources: Dyno Nobel, NONEL® LP series, technical data sheet, Dyno Nobel Asia Pacific, Brisbane, Queensland, Australia, 2007, Available at: www.dynonobel.com; Orica, Exel™ LP: non-electric, long delay detonator assembly, technical data sheet, Orica Mining Services, Mansfield, Queensland, Australia, 2008, Available at: www.oricaminingservices.com.

systems have also been developed to allow a greater range in delays for the long duration complex blasts commonly experienced in sublevel open stope mass-blasting applications. Additional delay periods between the specified in-hole delay numbers can be achieved using hole-to-hole or ring-to-ring delayed connector elements. Standard accepted delay timing errors for pyrotechnic delay element systems is approximately ±2% due to differences between delay element batches, temperature and humidity effects on shock tube and in-hole delay elements, and nonuniform standardization for all lengths of manufactured detonator. For short blast durations or use of long hole-to-hole delays, the probability of out-of-sequence firing is minimal. The accuracy error does increase the probability of out-of-sequence firing when long in-hole delays are used or the charge-to-charge delay intervals are reduced by using MS connectors.

292

Geotechnical Design for Sublevel Open Stoping

TABLE 6.7 Interhole Delays and Detonating Times for Open Stope Blasting Delay No.

Time (ms)

Interdelay Interval (ms)

Delay No.

Time (ms)

Interdelay Interval (ms)

3 4 5 6 7 8 8+(1)a 9 9+(1) 10 10+(1) 11 11+(1) 12 12+(1) 13 13+(1) 14 14+(1) 14+(2) 14+(3) 15

75 100 125 150 175 200 225 250 275 300 325 350 375 400 425 450 475 500 525 542 565 600

25 25 25 25 25 25 25 25 25 25 25 25 25 25 25 25 25 25 25 17 23 35

15+(1) 15+(2)a 15+(3) 16 16+(1) 16+(2) 16+(3) 17 17+(1) 17+(2) 17+(3) 18 18+(1) 18+(2) 19 19+(1) 19+(2) 19+(3) 20 20+(1) 20+(2) 20+(3)

625 642 665 700 725 742 765 800 825 842 865 950 975 992 1025 1050 1067 1092 1125 1150 1167 1192

23 17 23 35 25 17 23 35 25 17 23 85 25 17 33 25 17 25 33 25 17 25

a

Delay No.

Time (ms)

Interdelay Interval (ms)

21 21+(1) 21+(2) 21+(3) 22 22+(1) 22+(2) 22+(3) 23 23+(1) 23+(2) 23+(3) 24 24+(1) 24+(2) 24+(3) 25 25+(1) 25+(2) 25+(3)a

1225 1250 1267 1292 1400 1425 1442 1465 1675 1700 1717 1740 1950 1975 1992 2015 2275 2300 2317 2340

33 25 17 25 108 25 17 23 210 25 17 23 210 25 17 23 260 25 17 23

Standard millisecond delays (ms) and 25(1), 42(2), and 65(3) ms TLDs.

6.7.3  Electronic Delay Element Detonators Electronic delay element detonators have been under development since the 1980s and were launched into commercial use in the early 2000s. The electronic delay element (microchip) in general replaces the pyrotechnic element without significantly changing the design, dimensions, or physical properties of the detonator. The accepted error in electronic delay detonators is typically ±0.1% with available delays from 0 to 20,000 ms in predetermined or 1 ms intervals (e.g., Davey Bickford, 2008; Orica, 2010; Dyno Nobel, 2011). Previous research in open pit and underground mining has investigated the impacts of accurate delay timing on muckpile fragmentation and mine productivity (e.g., Tose and Baltus, 2002; Bartley and McClure, 2003; Grobler, 2003). The results of these studies largely indicate that accurate timing can improve the uniformity of the fragmentation distribution and in many cases

Drilling and Blasting

293

FIGURE 6.30 Blasting a narrow vein uphole stope using signal tube initiation systems and detonating chord.

decrease the mean particle size within the muckpile. Additional theories on the useful application of millisecond-accurate electronic firing deal with blast vibration reduction or frequency control, tailored timing for irregular or complex blasthole patterns, and collision of stress waves to improve fragmentation in specific areas of a blast. One standard practice for blasthole initiation in open stoping is to use signal tube initiation system detonators placed down (or up) the hole as shown in Figure 6.30. The signal tube initiation systems are attached to loops of detonating cord for each ring. The detonating chord is then initiated by instantaneous electric detonators connected to a mine-wide stopeblasting circuit. When more than one ring is being detonated, each loop of cord is linked to the next by a length of cord to provide security. Two electric detonators are placed on the cord loop at each ring position. In order to minimize damage from shrapnel cutting the signal tube initiation system downlines, the electric detonators should be placed under sandbags (Figure 6.31). The signal tube initiation system delay detonators are initiated by a shock wave passing through 3 mm-diameter plastic tubing, which is crimped onto a detonator. The abrasion-resistant, flexible, and high tensile–strength plastic tube has a 1.5 mm bore that contains explosive material which transmits a shock wave at 1.9 km/s. The shock front is capable of negotiating sharp bends, kinks, and knots without rupturing the plastic tube. Hence, it cannot side initiate any explosive and will minimize air blast. The reactive material is initiated by detonating cord or electric detonators.

294

Signal tube initiation systems

M/R 3

M/R 2

M/R 1

Geotechnical Design for Sublevel Open Stoping

Signal tube initiation systems

Loop of detonating cord Loop of detonating cord

Blastholes

Electric detonators sandbagged

FIGURE 6.31 Typical multi-ring blasting hookup. (Courtesy of Mount Isa Mines, Mount Isa, Queensland, Australia.)

Detonating cord is high-tensile, waterproof material, which has a core of typically 4–10 g/m of pentaerythritol tetranitrate (PETN) enclosed in plastic tapes, natural and synthetic fibers, and an outer sleeve of plastic. The 3.9–5.1 mm-diameter cord is flexible, abrasion-resistant, and relatively insensitive to detonation due to friction, impact, and electrostatic discharge. The cords have a large velocity of detonation (VOD) ranging from 6 to 7 km/s. 6.7.4 Priming The conventional approach in ring design is to place two detonators and two boosters at the bottom of each charged hole. This provides some insurance in the event that one detonator does not initiate. Double priming is indicated in the blasting plans by placing a circle around the delay number assigned to each hole. In addition, for long charge lengths (exceeding 20 m), security boosters are placed every 20 m along the charge axis. In practice, however, the location of the boosters in a charged column is largely a function of ring geometry and the location and orientation of large-scale geological discontinuities. Additional boosters are required in broken ground with high geological discontinuity connectivity, especially where large-scale faults may allow water inflow into the explosive charges. Figure 6.32 shows a typical setup where boosters are placed on both sides of faults to ensure initiation of a charged column of explosive.

295

Drilling and Blasting

Fault

Fault

Fault

Double booster

Security booster

FIGURE 6.32 Booster location with respect to large-scale geological discontinuities. (After Rosengren, M. and Jones, S., How can we improve fragmentation in the copper mine? Unpublished Mount Isa Mines Limited Internal Report, 1992. With permission.)

Damage to walls in radial patterns toeing into walls may be increased by the location of the boosters toward the end of the holes (near the excavation boundary). Boosters provide high shock energy, which is required to initiate other explosives and, consequently, the local rock damage at that point may be higher. If the boosters are moved along the charge axis (away from the boundary), the local damage may be reduced. However, the explosive column below the new booster location would reach full VOD, thus increasing the damage at the toe. In practice, boosters of holes toeing into walls are placed 2–4 m from the bottom of the holes. 6.7.5  Sequencing and Timing The fundamental objective of blasthole delay sequencing is to provide each charge column with as many free faces as possible to break into. In ring firing, this can be achieved not only by blasting holes toward a free face in the ring burden direction, but also by providing each hole with at least one free face in the direction of the adjacent blastholes in the same ring. Figure 6.33 shows the concept of interhole and inter-ring delays in ring blasting for open stoping. Interhole delays are usually kept to a minimum to optimize blasthole interaction and enhance rock fragmentation. To achieve this, short period

296

Geotechnical Design for Sublevel Open Stoping

Inter-ring delay

Interhole delay

Inter-ring delay FIGURE 6.33 Interhole and inter-ring delays in ring blasting. (From Langdon, C. and Duniam, P., Advances in theory and application of non-electric initiation systems to 60 series extraction at the Mount Lyell copper mine, in T. Golosinki, ed., Proceedings of the Sixth Underground Operators Conference, Kalgoorlie, Western Australia, Australia, November 13–14, 1995, pp. 291–298, The AusIMM, Melbourne, Victoria, Australia. With permission.)

delays (MS signal tube initiation systems series) are normally used. Because of detonator scatter, ring-to-ring timing must be designed to avoid out-ofsequence firing, where holes in the second or third ring fire prior to holes in the face row closest to the void. A minimum delay time of 20 ms/m of burden is recommended for ringto-ring timing. However, for blasts having more than three rings detonating, the only way to eliminate the possibility of inter-ring misfires is to skip one complete number in the MS series between two holes in two adjacent rings. For example, a hole shadowed by a number 8 should be fired on a number 10. Furthermore, it is important to minimize the total blast duration within a stope firing. The longer the charged holes sit while other holes are detonating, the greater the chance of blast malfunction due to hole dislocation, shearing, sympathetic detonation, or explosive desensitization. This is particularly true in rock masses in which large-scale structures are present. The standard MS signal tube initiation system series has 28 delay numbers that can be used for open stope blasting. The practice of piggybacking using trunk line detonators (TLD) at the hole collars can extend the range of delay detonators to 55 numbers. TLDs are used to provide a delay between the detonating cord and the signal tube initiation system delays placed down the hole. For security reasons, two TLDs are used and all signal tube initiation system downlines (including security detonators) are connected to the TLDs, which in turn are then hooked up to the detonating cord trunklines. Three TLDs (having 25, 42, and 65 ms delays) can be introduced anywhere in the standard MS range as shown in Table 6.7. The use of TLDs reduces the interhole delay, thus enhancing fragmentation. However, they should not be used for inter-ring delays, as an out-ofsequence detonation may result. In addition, some TLD connections may

Drilling and Blasting

297

produce shrapnel, so they may have to be sandbagged or pushed into the holes in uphole blasting. Furthermore, to minimize damage to the blast hook-up (and avoid misfires), all downhole detonators should be burning before the first hole in the blast is fired (total flame front). When all three TLD delay numbers are used in a single blast, it is recommended to fire the initial hole using delay number 4 (100 ms) to ensure all TLDs have fired and all downhole delays have initiated before any fly-rock or shrapnel is produced. In cases where blasthole rings toe into each other, neighboring holes in the opposing rings should be fired on the same delay and each hole security primed. This is the case when main downhole rings are combined with TUC holes or when rings are fan-drilled from two drill drives, one on each side of a stope. In cases where downholes toe into a horizontal hole drilled from a sublevel below, they should be fired on the same number. All holes should be security-primed and the boosters of the downhole charges pulled up 4 m from the bottom of the holes. Modern electronic detonators (Liu et al., 2002) allow a greater degree of delay interval control by enabling the delays to be modified and programmed for repetition if needed. The pyrotechnic component of the delays is replaced by an electronic component that uses a miniature electronic timing circuit to ignite the detonation charge. During detonator manufacturing, a delay sequence number is built into each detonator. During blasting the detonators fire with a constant delay interval between consecutive numbers and it is possible to program desired time intervals to suit the rock mass conditions within a stope. The following information should be recorded on ring section charge plans prior to firing, with amendments recorded during charging: • • • •

The date the ring section was fired The amount and explosive type used in each hole The actual firing sequence used Any problems encountered while charging and firing each ring section • The results of the blast including quality of fragmentation, misfires, hole freezing, stope wall falloff, backbreak, etc. The ring charge plans should be returned to the mine planning department for stope reconciliation once the firing of a stope is complete. Figure 6.34 shows a typical firing sequence in open stoping. The following are the basic guidelines for ring blasting:

1. In any ring, the longest hole is fired first. 2. Rings fire from the bottom proceeding upward. 3. Interlocking toes fire simultaneously. 4. Footwall and hangingwall holes fire late in the sequence.

298

Geotechnical Design for Sublevel Open Stoping

13

11

9

12

12

2m 13

11

10

9

10 10

12

8

11

11

11

8

8

5 5

5

2

2 7

4

1

3

6

9

7

4

1

3

6

6

4

2

1

3

5

2m

1

Detonation sequence

FIGURE 6.34 Typical detonation sequence in open stoping. (Courtesy of Mount Isa Mines, Mount Isa, Queensland, Australia.)

6.8  Raise and Cutoff Slot Blasting Successful excavation of raises and COSs is critical in sublevel stoping, as they provide free faces and voids into which the remaining ore in a stope is blasted. Raises and slots are critical areas where significant rock mass damage can occur due to the high concentration of explosive energy utilized to ensure the formation of the initial free face or void. Breakthrough holes drilled parallel to the initial raise or LHW are implemented to create the COS as shown in Figure 6.35. 6.8.1  Longhole Winzes A LHW provides an initial void into which the COS is blasted. The fundamental principle of a LHW is similar to that of a large burn cut, where a number of large-diameter relief holes are left uncharged to provide an initial void into which the charged shothole(s) can break. The key to successful LHW firing is to have adequate initial void and adequate delay times between holes. LHWs are typically fired using long period delay detonators to allow the broken rock to fall out of the winze before the next hole is detonated. Drilling accuracy in a LHW is critical. Excessive drill deviation may cause the following:

299

Drilling and Blasting

Main rings Cut-off slot

Raise or LHW

FIGURE 6.35 Three-dimensional view of a slot within an open stope geometry.



1. Blastholes designed as void become ineffective and the winze at a particular horizon does not resemble the intended design (Figure 6.36). 2. Blastholes intersect each other, causing confusion or difficulties during charging. This may also create sympathetic detonation, desensitization, or hole dislocation, thus compromising breakage. 3. Blastholes having excessive burdens may not break out adequately and cause damage around the winze. 4. Excessive movement near large-scale geological discontinuities may block holes. 5. The final axis of the winze may be changed at some horizons. Figure 6.37 shows a standard LHW design based on the premise that drill setup is a major cause of drill deviation contributing to the issues discussed earlier. A good LHW design should minimize drilling setup positions, have

300

Geotechnical Design for Sublevel Open Stoping

Top level E

F

A B

C D

G

H F

E A G

D C

B H Breakthrough level

FIGURE 6.36 Deviation of drillholes in a LHW.

1.5 m

3.0 m

#12

#11

#6

#4

1.3 m 0.5

#8

#3 m

#9

#1 #5

#2

#10

89 mm blasthole 160 mm relief hole #2 Detonation sequence

#7

#13

FIGURE 6.37 LHW pattern—Hilton Mine, Mount Isa Mines. (Courtesy of Mount Isa Mines, Mount Isa, Queensland, Australia.)

301

Drilling and Blasting

enough large-diameter relief holes, and the charged cut holes should be positioned such that they are shielded from one another. The design shown in Figure 6.37 is set up so that the winze can be started from a number of points if required. In practice, however, more holes may have been drilled than those shown in the pattern. These are either rebores for deviated drillholes or additional easer holes around the winze. Figure 6.38 shows an alternative LHW design having 17 charged holes of 73 mm diameter and 4 relief holes of 115 mm diameter. At Mount Isa Mines, a typical 12 m-long winze similar to the one shown can be drilled in 36 h with an Atlas Copco Simba H221 using T38 rods. Figure 6.39 shows a typical LHW pattern that uses a Robbins 12MD raise hole to provide a 660 mm-diameter initial void. This minimizes the potential for drill deviation-related problems experienced with LHW. Nevertheless, significant preparation work is required to use a 12MD machine, as the floor has to be cleaned and a concrete slab poured prior to raise-boring. Following the completion of the raise, an accurate survey pickup of the hole is required, so that an appropriate number of easers can be drilled around the raise to establish the COS. Data from Mount Isa Mines showed that when a stope height was greater than 25 m, a 12MD became cost-effective in comparison to the Simba for drilling an entire COS. Although difficult to quantify, it is expected that reduced dilution may occur using a 12MD cutoff compared to a Simba COS. The maximum length for a LHW in sublevel stoping is around 40–50 m. Conventionally, the holes are charged from the top of the winze, with 3–6 m cuts fired each time starting at the bottom of the winze and moving up. If a large-scale structure is present, the winze should be fired to the structure to avoid falloff. If the winze freezes (malfunction), the holes must be washed

0.65 m

2m

0.3 m 73 mm blasthole

2m

115 mm relief hole 0.65 m

FIGURE 6.38 Typical LHW pattern, Lead Mine. (Courtesy of Mount Isa Mines, Mount Isa, Queensland, Australia.)

302

Geotechnical Design for Sublevel Open Stoping

2.1

2

m

3.00 m

Relief hole

140 mm blasthole

0.60 m

6m

0.6

3.00 m

0.60 m

FIGURE 6.39 A typical blasthole pattern used in conjunction with a 0.66 m raise. (Courtesy of Mount Isa Mines, Mount Isa, Queensland, Australia.)

and blown out using compressed air. If the holes cannot be reestablished, then redrilling may be required. A typical advance for a 3 m × 2 m blasted LHW ranges from 5 to 10 m per blast, depending upon the amount of hole deviation present. 6.8.2  Cutoff Slots The COS is the most important part of the in-stope extraction sequence, as it provides the initial void into which the subsequent rings are fired (Figure 6.40). The first ring adjacent to a COS will typically only break to the width defined by the slot void, although slightly wider rings may be gradually fired into the initial slot to “gain ground.” A large increase in orebody width over the stope length may require the inclusion of an additional COS in the widest section of the orebody. In general, the decision on the location of a COS is dependent upon orebody width, the fill type of adjacent stopes, and access constraints. In narrow vein orebodies in which bench stoping is practiced, the COS is commonly designed to extend across the full width of each stope. In order to successfully fire a COS, it is recommended to have orebody sills open to full operating width above and below the slot. Fanning of blastholes into unbroken ground to strip the slot is usually unsuccessful, as the blastholes are unlikely to pull to the full design depth. This problem becomes exaggerated with increasing angles of advance (angle of arc) away from the void (see Figure 6.41). To minimize the angle of arc that the drillholes are required to cover, a drill rig has to be capable of drilling across the entire orebody width, using as near to parallel holes as possible.

Drilling and Blasting

303

FIGURE 6.40 Longitudinal view into a COS, Mount Isa Mines. (Courtesy of Mount Isa Mines, Mount Isa, Queensland, Australia.)

LHWs and COSs have the greatest concentration of explosive energy of any area of a blast design. Consequently, the LHW position within a COS should be as close as possible to the stope footwall to minimize blast damage to the stope hangingwall. A number of cutoff design rules are suggested as follows: 1. Ideally, the COS should be positioned in the widest part of the orebody. 2. In narrow orebodies, the drill drives at the COS position should be stripped to full orebody width. In wide orebodies, the drilling drives should be connected above and below the COS (Figure 6.35). 3. Drillholes should be drilled as close as parallel to the raise or LHW as possible. When holes converge at the winze or raise area, they must be fired together. 4. The COS must be enlarged by firing holes into the winze or raise, as if the slot were a narrow orebody. This is usually done using a “dice five” pattern with two lead holes firing first into the LHW, followed by an easer hole.

304

Geotechnical Design for Sublevel Open Stoping

Ground opened by cut-off slot firing

10L

11E

FIGURE 6.41 Example of fanned holes in a COS not pulling to full depth. (Courtesy of Mount Isa Mines, Mount Isa, Queensland, Australia.)

At Mount Isa Mines, some of the COSs are drilled using vertical holes of 140 mm diameter and up to 50 m in length. The holes are drilled on a “dice five” pattern, with 1.5–1.8 m burden and 3.2–3.6 m spacing (Figure 6.42). The holes are usually drilled to breakthrough on the sublevel below, which allows for drill accuracy checks and draining of water and drill cuttings. Holes are drilled parallel to the cutoff raise and, if applicable, the hangingwall and footwall of the orebody. It is important that the collars and breakthroughs are recorded by surveyors prior to designing the remaining stope blasthole patterns or slot charge plans, as these may be significantly different from the initial designed coordinates. Figure 6.43 shows the position of the COS with respect to the main rings within two stoping geometries. Figure 6.44 is a long section view showing the location of the slot with respect to the main rings within the stope. Stringing of COS holes during drilling ensures that the designed number of holes have been drilled. This also helps to establish drilling accuracy. When

305

Drilling and Blasting

Dia. (mm) Burden

Area

Spacing

TPMD

PF

Length

Upholes

70

1.8 m

1.2 m

4.6

0.67

20 m

Downholes

140

1.8 m

3.6 m

12.4

0.74

40 m

FIGURE 6.42 Plan view of COS showing raise and blasthole positions. (Courtesy of Mount Isa Mines, Mount Isa, Queensland, Australia.)

Filled stope

Expansion rings

Expansion rings 1–4

40 m

Expansion rings

Cut-off slot

Cut-off slot

Ring 5

Ring 8 Ring 9 Ring 10

Ring 6 Mass blast

Mass blast

Ring 7

Ring 7 Ring 8 Ring 9 Ring 10 Ring 11 Ring 12

40 m

FIGURE 6.43 Plan view showing COS with respect to the main ring geometries. (Courtesy of Mount Isa Mines, Mount Isa, Queensland, Australia.)

cutoff raises are raise-bored after the COS has been drilled, holes around the raise must be strung to ensure an adequate number of holes are available to break into the raise. However, the practice of raise-boring after cutoff drilling is not recommended. COS blasting sequences depend upon the size of the raise used. For example, a 60 m COS lift having a 1.8 m-diameter raise may be fired in two 30 m lifts. A slot with a 0.66 m-diameter raise can be blasted in 12–20 m lifts. Conventionally, when a LHW has been fired to approximately half way, blasting of the COS blasthole toes can start. A recommended lead-lag

306

4

Expansion

7

4

7

8

6

5

4

3

1 Cut-off slot

7

Massblast

Geotechnical Design for Sublevel Open Stoping

3

1

2

1 Blasting sequence within stope FIGURE 6.44 Typical long section for a COS. (Courtesy of Mount Isa Mines, Mount Isa, Queensland, Australia.)

25

23 24

21

22 5

22

7 25

23

21

3

1 2

8 4

6 22

FIGURE 6.45 Blasting a LHW and cutoff collars to breakthrough.

between the LHW advance and the COS advance is approximately 10 m. When firing out a COS, up to 6 holes at a time are blasted when firing toes, while up to 12 holes are blasted when firing collars (Figure 6.45). In some cases, all slot toes are fired, with an arched geometry from the LHW to the slot extents. If for some reason, cutoff toes are not fired prior to completion of the LHW, up to 6 complete COS holes are fired at a time into the LHW void. Both LHW and COS holes should be double-primed to ensure that they detonate. In addition, when firing a COS and main rings or cleaner rings, a maximum of 3 rings are recommended to ensure adequate void for the rings.

307

Drilling and Blasting

L4

1

Fa

ul

t

Uphole LHW

O392 Filled

1.7

%

2.0

%

20B

Bas

em

ent

Downhole LHW

2450 RL

21E FIGURE 6.46 COS incorporating downhole and uphole LHWs. (Courtesy of Mount Isa Mines, Mount Isa, Queensland, Australia.)

In some stoping geometries, uphole winzes may be required in conjunction with uphole COS charging (Figure 6.46). Uphole winzes require a single firing, and therefore their length usually does not exceed 15 m. If an uphole slot is to be wider than the LHW, then additional stripping holes should be fired with the LHW. If the LHW does not pull to the required depth, it is likely that a recovery LHW will have to be drilled or the entire slot will have to be redrilled. Figure 6.47 shows a 2 m × 3 m uphole LHW geometry used at Mount Isa Mines. The relief holes in the figure are drilled 1.5 m deeper than the charged holes.

6.9  Trough Undercut Blasting TUCs are designed in a similar manner to the main stope rings. However, TUCs are usually drilled using 70–89 mm-diameter upholes inclined at 70° and limited to a length of 15 m to allow conventional blow-loading of explosives. TUCs are shaped to promote the best rill angle at the stope drawpoints

308

Geotechnical Design for Sublevel Open Stoping

3.0 m

0.4 m

0.4 m

0.5 m

0.25 m

1.0 m

89 mm blasthole

0.8 m

150 mm relief hole

0.25 m

0.5 m

0.7 m

2.0 m FIGURE 6.47 Typical uphole LHW geometry. (Courtesy of Mount Isa Mines, Mount Isa, Queensland, Australia.)

while protecting the rock mass at the brows as much as possible. Inclined (dumped) TUC rings are required, so that charging of holes at the edge of a stope is not undertaken. TUC rings are designed to be small rings which minimize the charging time and exposure of charge-up personnel to the slot void or stope rill. Figure 6.48 shows a 70 mm-diameter blasthole TUC, with the first row of holes inclined at 50°, followed by inclinations of 60°, 70°, and 80°, with the remainder of the rings at 90°. The collars for the first row of holes are located 4 m away from the COS. A horizontal distance of 2.5 m is used for the remainder of the rings. Consecutive TUC rings are designed using staggered patterns, and the holes must interlock with the toes of the downholes from the sublevel above. A nominal toe overlap of 2 m is recommended.

6.10  Rock Diaphragm Blasting The role of a diaphragm is to protect an adjacent fill mass (weak or uncemented) from blast damage from production blasting and help to minimize fill failures. COSs should be designed to the width of the main ring firings, and cleaner rings adjacent to the diaphragm can be used to optimize recovery

309

Drilling and Blasting

70 mm blasthole diameter ANFO loading 3m 2m

50° 70°

(a)

Toe spacing

(b)

2.5 m

3m

2.35

m

2m

Burden

FIGURE 6.48 (a) Cross section and (b) longitudinal section views of a TUC. (Courtesy of Mount Isa Mines, Mount Isa, Queensland, Australia.)

and minimize blast damage to the diaphragm (see Figure 2.7). Cleaner rings are drilled and blasted from diaphragm ring recesses located at the edges of the stopes. Hole deviation can become a large problem while drilling cleaner rings, as the holes are drilled from limited locations and are of excessive length (Figure 6.49). Some of the holes in cleaner rings and diaphragm rings are difficult to clean out and prepare for charging, as the holes can be damaged from blasting the stope COS or main rings. In addition, insufficient information on the exact location of the fill boundaries can lead to inadequate diaphragm design thickness, further contributing to fill mass instability. Development of the cavity-monitoring system has allowed final stope geometries to be determined, although localized rock falls can occur prior to fill completion.

6.11  Mass Blasting Mass blasts consist of multiple blasthole rings and can exceed 100,000 tonnes per firing. The sequencing rules for mass blasting must take into account blast vibration constraints, rock mass damage from overconfinement, delay variability, geometrical constraints, and major geological structures as follows:



1. The longest hole in a ring should be fired first to maximize the initial void created at the start of the firing. Other holes are then sequentially stripped into the void created by the first hole. 2. Holes toeing into each other should be fired on the same number.

310

Geotechnical Design for Sublevel Open Stoping

1500 E

1450 E

2950 Meters drilled 346.4 Meters charged 188.4 Tonnes broken 7929.6 Tonnes/m drilled 22.8 Kg ANFO /ring 1957.2 /tonne 0.24 /m drilled 5.6 38.4 +3°

42.8 –10°

12

27

31 .1 35 0° 0.8 3 6° –3 –3

9

13

18

.0 40 ° 5 –2

12D

19

2 –4 7.5 3°

45.7 ° –15 3 45. ° –20

8

23

40.3 –4°

12/L

2900

FIGURE 6.49 Cross section showing a typical cleaner ring charge geometry. (Courtesy of Mount Isa Mines, Mount Isa, Queensland, Australia.)



3. A minimum of 30% expansion void is required prior to mass blasting to accommodate broken material swell. 4. A minimum delay time of 15–20 ms/m of burden is suggested between successive rings, while changes of firing direction within a mass blast require 100–200 ms delays. 5. Perimeter holes as well as holes parallel to stope crowns should be fired a few numbers after adjacent holes within a ring to allow adequate relief. This reduces confinement of explosive gases at the stope boundaries and minimizes the likelihood of overbreak. The delay between stope boundary holes and adjacent holes should follow the same rule as per burden, that is, 15–20 ms/m of toe spacing. Alternatively, one complete number in the detonation sequence should be skipped between adjacent holes.

Drilling and Blasting

311

In order to increase the number of delays available for mass blasting, a combination of trunk line delays and down-the-hole delays can be used as discussed in Section 6.7.2. It is extremely important that all surface delays be activated before the first hole in the first ring detonates to avoid cutoff. Before a mass blast is undertaken, all reentry, inspection, ventilation, and shift change procedures should be detailed. The advantages of mass blasting include the following (Guilfoyle and Bradford, 1982):









1. Safer work conditions arise when charger and ring-firer crews are not continually required to work near freshly blasted stope boundaries. 2. Improved rock fragmentation results from the shearing action of interacting detonation charges and in-flight rock collisions. 3. A better utilization of resources is possible due to a concentrated and semicontinuous charging operation which proceeds simultaneously on a number of sublevels. 4. Because fewer individual firings are required, the problem of postblast falloff is reduced. 5. Large-scale structural discontinuities (such as faults, shears, etc.) can be included within a single firing, thus minimizing ground movement and the potential loss of blastholes. 6. Stable conditions can be maintained during slotting, initial ring expansion, and charging of the mass blast, after which no need exists for personnel to reenter the area. 7. Following a mass blast, passive support to the stope walls (rock or fill masses) is provided by the broken ore. Up to three-quarters of the stope may be filled by the broken ore following blasting. 8. Fewer individual blasts are likely to minimize damage to services and other scheduled activities around the stopes. 9. The large broken ore tonnages from mass blasts allow uninterrupted production at high extraction rates from the stope drawpoints.

The disadvantages of mass blasting may include the following:

1. Multiple-lift mass blasts are typically initiated from multiple access drives. Should a cut-off occur, it may be difficult to gain reentry to those areas. 2. Mass blasts often create a large change of geometry likely to redistribute significant stress around the stope boundaries. Stress changes may induce rock noise and damage and a reentry period to the stoping area may be required, thus delaying production. 3. Any malfunction of the initiation system or explosive early in the firing sequence can “freeze” the entire mass firing.

312



Geotechnical Design for Sublevel Open Stoping

4. Mass firing on top of broken stocks within a stope can lead to excessive ore compaction at the draw point. 5. Inadequate delays in main ring firings or between main firing levels can lead to rapid overpressurization of the development drives from ore block displacement, causing damage to the ventilation system.

6.11.1  Control of Ground Vibration In addition to a correct and complete detonation sequence of all the holes involved in a mass blast, minimization of damage to adjacent structures (such as shafts, pillars, etc.) from excessive vibration is an important objective. Overpressure from blasting may also cause significant damage to ventilation systems. Consequently, the initiation sequences must be designed with a charge weight per delay evenly distributed throughout the mass blast duration. The objective is to prevent periods of high explosive concentration within the blast. Often, the quantity of explosive detonating within a specified time interval is limited to 1000 kg. The optimum delay interval between successive detonating charges to minimize wave interaction has been suggested by Heilig (1999) to be half of the duration of vibration from a single blasthole charge. For most underground rock types (for a distance up to 200 m from the blast) this value has been determined by Heilig (1999) to be approximately 20 ms. The effects of charge weight per delay distribution throughout a blast can also be determined by blast monitoring to ensure that the number of charges initiating per a 20 ms period are minimized. In mine sites where a town or city is nearby, the standard practice is to monitor the surface vibration from all stope blasts exceeding 100,000 tonnes. Monitoring experience suggests that the surface vibration values obtained from surface monitoring are likely to change from place to place. It is possible that due to large-scale geological discontinuities or the effects of mining voids or fill masses, some sites may experience higher levels than those monitored at shorter distances. Figure 6.50 shows the monitored peak particle velocity (PPV) from long-term surface blast monitoring at the Mount Isa Mines lease boundary (approximately 1 km from the blast). At Mount Isa Mines, the vibrations induced by stope blasting are generally acceptable to the community. The historical level of complaints is very low and no damage to property has ever been linked to the large-scale underground blasting activities. Consequently, on the basis of the long-term data collected at Mount Isa Mines, a suitable criterion which can be realistically achieved in sublevel stoping is suggested as follows:

1. The surface PPV of 10 mm/s may be exceeded by up to 10% of the total number of daily blasts. 2. The level cannot exceed 20 mm/s at any time, including during mass blasts.

313

30

Mass blasts

25 20 15 10 5 450,000

400,000

350,000

300,000

250,000

200,000

150,000

100,000

50,000

0 0

Peak particle velocity (mm/s)

Drilling and Blasting

Tonnes per blast FIGURE 6.50 Surface vibration levels from mass blasts at Mount Isa Mines. (Courtesy of Mount Isa Mines, Mount Isa, Queensland, Australia.)

7 Rock Reinforcement and Support

7.1 Introduction The objective of ground support is to maintain excavations safe and open for their intended purpose and lifespan (Villaescusa, 1999a). In an open stoping context, the effectiveness of a ground support strategy is important for two main reasons: safety to personnel and equipment within the stope development, and achieving the most economical extraction of ore with minimal dilution from the final stope walls. The type of ground support required in a particular stope location is dependent on several factors, including the available rock mass strength, the geometry of the excavation, the stresses present in the rock, the blasting practices, and the weathering process (see Section 1.4). Two stabilization techniques can be used to improve and maintain the load-bearing capacity of a rock mass near the boundaries of an underground excavation (Windsor and Thompson, 1992): • Rock reinforcement Reinforcement is considered to be exclusively systems of components installed in boreholes drilled in a rock mass, for example, cementencapsulated threaded bar, friction stabilizers, and cable bolts. The reinforcing elements are an integral part of a reinforced rock mass. • Rock support Support is considered to be exclusively systems of components that are located on the exposed faces of excavations, for example, mesh, straps, shotcrete, and steel arches. The supporting members are external to the rock and respond to significant inward movement of the rock mass surrounding an excavation. The reinforcing elements provide effective stabilization by helping a rock mass to support itself (Hoek and Brown, 1980). This is achieved by preventing unraveling and enhancing the self-interlocking properties of a rock mass. A reinforcement pattern strengthens the exposed rock mass around an excavation by preventing the detachment of loose blocks and by increasing the 315

316

Geotechnical Design for Sublevel Open Stoping

shear strength of the geological discontinuities intersected by the reinforcing elements. This results in a reinforced zone that helps to redistribute stresses around the excavations and also minimizes dilation of preexisting geological discontinuities. Careful blasting and correct scaling reduce the amount of loose rock that has to be supported, thus enhancing the self-stabilizing behavior of a rock mass. In sublevel open stoping mines, the primary form of excavation stabilization is provided by the reinforcement pattern installed within the various stope development excavations. Rock support, such as that provided by mesh and shotcrete, is required to provide surface restraint within a reinforcement pattern at the excavation boundaries. The reinforcement controls the overall excavation stability through keying, arching, or composite beam reinforcing actions (Windsor and Thompson, 1992), while mesh or shotcrete supports the small loose pieces of rock that can potentially detach within a bolting pattern (Figure 7.1). Ground support can be considered to consist of combinations of reinforcement and support systems. It is normal practice to design the reinforcement to act with the support to form a ground support scheme (Windsor and Thompson, 1992). That is, the support is restrained by a plate held in place by the reinforcement system. If this interaction at the collar of the reinforcement system fails, then the ground support scheme will not be effective in retaining the unstable rock. Another important aspect of the ground support design is its overall response to the amount of rock mass deformation and the rate at which this occurs.

FIGURE 7.1 Support and reinforcement of a highly stressed blocky rock mass.

317

Rock Reinforcement and Support

7.2 Terminology A classification to describe the forms, functions, basic mechanics, and behavior of the different commercially available rock support and reinforcement systems was developed by Thompson and Windsor (1992). The method classifies the existing reinforcement systems by dividing them into three basic categories in order to explain the basic mechanisms of load transfer between the reinforcing elements and a rock mass. A description and comparison of devices within a particular category or between separate categories is facilitated by the method. The categories are shown in Figure 7.2 and are described as continuous mechanical coupled (CMC), continuous friction coupled (CFC), and discrete mechanical and friction coupled (DMFC). Some typical reinforcing devices are grouped according to this classification in Table 7.1.

Type

Longitudinal view of reinforcement element Unstable surface region

Stable interior region

Unstable surface region

Stable interior region

CMC

CFC

DMFC

Unstable surface region

Stable interior region

FIGURE 7.2 Classification of reinforcement action. (After Thompson, A.G. and Windsor, C.R., A classification system for reinforcement and its use in design, in T. Szwedzicki, G.R. Baird, and T.N. Little, eds., Proceedings of the Western Australian Conference on Mining Geomechanics, Kalgoorlie, Western Australia, Australia, June 8–10, 1992, pp. 115–125, Western Australian School of Mines, Kalgoorlie, Western Australia, Australia.)

318

Geotechnical Design for Sublevel Open Stoping

TABLE 7.1 Classification of Typical Reinforcement Devices Type

Description

CMC

Full-column cement-/resin-grouted bars (grouted CT bolt, deformed bar, threaded bar, and fully grouted Posimix) Cement-grouted cables (plain strand and modified geometry) Friction stabilizers (split-set bolt, friction bolt, and Swellex) Mechanical anchors (ungrouted CT and HGB bolts, expansion shell, and slot and wedge) Single cement/resin cartridge anchors (paddle bolt, deformed bar, and debonded Posimix)

CFC DMFC

Source: Thompson, A.G. and Windsor, C.R., A classification system for reinforcement and its use in design, in T. Szwedzicki, G.R. Baird, and T.N. Little, eds., Proceedings of the Western Australian Conference on Mining Geomechanics, Kalgoorlie, June 8–10, 1992, pp. 115–125, Western Australian School of Mines, Kalgoorlie, Western Australia, Australia.

7.2.1  Continuous Mechanical Coupled A CMC reinforcing element relies on a fixing agent, usually a cement- or resin-based grout, which fills the annulus between the element and the borehole wall. The main function of the grout is to provide a mechanism for load transfer between the rock mass and the reinforcing element. The reinforcing elements are usually manufactured with variable crosssectional shapes in order to increase the element-to-grout bond strength. A mechanical key is effectively created by the geometrical interference between the element and the grout along the entire reinforcement length. The element is defined as continuously coupled to the rock mass by way of interlock with the grouting agent (Thompson and Windsor, 1992). 7.2.2  Continuous Friction Coupled A CFC reinforcing element is installed in direct contact with the rock mass. The mechanism of load transfer is a function of the frictional forces developed between the reinforcing element and the borehole wall. The load transfer is limited by the radial prestress set up during the initial element insertion. The bond strength is a function of the element diameter, the borehole diameter, and any geometrical irregularities occurring at the borehole wall. The radial stress can be related to a force along the length of the reinforcing element and is achieved by deforming the cross-sectional area of the element to suit the borehole. This can be achieved by either contracting an oversized element section into an undersized borehole (friction stabilizer) or by expanding an undersized element section into an oversized borehole

319

Rock Reinforcement and Support

(Swellex bolt). A modification of this reinforcing action can be achieved by cement grouting of the split-set bolts as described by Villaescusa and Wright (1997). 7.2.3  Discrete Mechanical and Friction Coupled A DMFC device transfers load at two discrete points, namely, the borehole collar and the anchor point, which is located at some depth into the borehole. The length of the element between the two discrete points (plate and anchor) is actually decoupled from the rock mass. The load transfer is then limited to a relatively short anchor length. Load transfer at the anchor point can be achieved by either mechanical (grouted anchor) or frictional means (expansion shell). The strength of an expansion shell may be limited by the strength of the rock at the borehole wall, and these devices are best suited to hard rock applications (Villaescusa and Wright, 1999). Grouted anchors may be used in soft rock masses, where a high load transfer can be achieved over a short length, provided that gloving by the resin cartridge does not occur (Villaescusa et al., 2008). 7.2.4  Load Transfer Concept The load transfer concept is one of the most fundamental concepts required to completely understand the behavior of a reinforcing element. The concept shown in Figure 7.3 can be used to understand the stabilizing action of all reinforcing devices and their effect on excavation stability. The concept Frictional resistance and mechanical interlock within stable (interior) region

l ica y log nuit o Ge onti c dis

Embedment length within stable region

Unstable region Movement

FIGURE 7.3 Load transfer and embedment length concepts.

Frictional resistance and mechanical interlock within unstable (wedge) region (complemented by plate)

320

Geotechnical Design for Sublevel Open Stoping

can be explained by three basic individual components (Windsor and Thompson, 1993):

1. Rock movement at the excavation boundary, which causes load transfer from the unstable rock, wedge, or slab to the reinforcing element 2. Transfer of load via the reinforcing element from the unstable portion to a stable interior region within the rock mass 3. Transfer of the reinforcing element load to the rock mass in the stable zone Failure of a rock block or a layer of rock being stabilized may be associated with any one of the three separate components of load transfer because of insufficient steel capacity (rupture of the reinforcing element) or inadequate bond strength (slippage). 7.2.5  Embedment Length Concept Embedment length is the length of a reinforcing element on either side of an active geological discontinuity defining a potentially unstable wedge or block such as that shown in Figure 7.4. The critical embedment length is the minimum length of reinforcement required to mobilize the full reinforcing capacity of the system. Short embedment lengths within an unstable region can be compensated for by the fact that a properly matched face plate provides enough surface restraint to mobilize the system capacity. Short embedment lengths within

FIGURE 7.4 Slippage within a stable region due to insufficient embedment length.

321

Rock Reinforcement and Support

the stable region are more critical, especially when a reinforcement element is installed at an unfavorable angle with respect to the free surface. 7.2.6  Reinforcement Performance Indicators A number of parameters may be used to characterize the performance of different reinforcement systems. In the absence of being able simply to simulate axial and shear loading of reinforcement, reinforcement performance is generally characterized by the force–displacement response of a reinforcement system subjected to axial loading. Figure 7.5 shows a generic force–­displacement response with annotations of a number of reinforcement system performance indicators. The performance indicators may be grouped as follows (Thompson et al., 2012):

Force

• Force capacities Fmax Maximum force. Fres Residual force at maximum displacement. • Displacement capacities δp Displacement at maximum force. δmax Maximum displacement. • Stiffnesses Kti Initial tangent stiffness.

Peak

Fmax

Residual

Fres 1

Kti

Ksp

1 1

δp

Ksr

Displacement

δmax

FIGURE 7.5 Force–displacement response for a generic reinforcement system subjected to axial loading. (With kind permission from Springer Science + Business Media: Geotech. Geol. Eng., Ground support terminology and classification: An update, 30, 2012, 553, Thompson, A.G., Villaescusa, E., and Windsor, C.R.)

322

Geotechnical Design for Sublevel Open Stoping

Ksp Secant stiffness at maximum force. Ksr Secant stiffness at maximum displacement. • Energy absorption capacity Energy absorption capacity is equivalent to the area between the force–displacement curve and the displacement axis and is relevant to the performance of reinforcement subjected to dynamic loading. Ep Energy absorption to peak force. Er Energy absorption at maximum displacement. Other parameters may need to be considered if the reinforcement system is loaded predominantly in shear. For example, it is known that strand is more flexible when loaded in shear than a solid bar and can therefore sustain higher shear displacements. The ability of a reinforcement system to sustain shear displacements is improved by de-coupling of the element from the grout as it allows for axial displacement of the element to be distributed over a longer length of the element near the discontinuity.

7.3  Ground Support Design Ground support design in most stoping operations is based on previous experience and evolves over a number of years. In many instances, there may be nothing technically wrong with the designs, and the performance can be considered to be acceptable. However, rock mass conditions usually change with the progress of a mine (e.g., stresses increase as the depth of mining increases and when the global extraction increases), and accordingly, ground support performance may change and become unacceptable. Also, the experiential ground support measures may not be optimal. That is, the installed reinforcement and support capacities may not satisfy the rock mass demand. A formal ground support design procedure (Thompson et al., 2012) attempts to

1. Identify the rock mass demand 2. Select reinforcement and support systems with appropriate characteristic responses 3. Specify their arrangement

The generic procedure consists of several distinct steps (Thompson et al., 2012):

1. Identify a mechanism of failure 2. Estimate the areal support demand

323

Rock Reinforcement and Support



3. Estimate the reinforcement length, force, and displacement demand 4. Estimate the energy demand 5. Select appropriate reinforcement and support systems 6. Propose arrangement of reinforcement and support systems and evaluate 7. Specify the complete ground support scheme

This procedure may need to be applied to several different observed mechanisms of failure. In most instances, it is not possible to perform formal designs because the rock mass variables that define demand cannot be quantified with any degree of confidence. However, the rock mass demand can usually be defined qualitatively in terms of low, medium, high, very high, and extremely high reaction pressure, surface displacement at failure, and energy demands per meter square (Table 7.2). These qualitative descriptions of rock mass demand can then be satisfied by reinforcement systems that can be classified using corresponding ratings (see Figure 7.40). The design process is more complicated when the rock mass experiences seismicity and the ground support is subjected to dynamic loading. For dynamic ground support design, it is necessary to consider the expected nature of seismic events associated with slip on major structures or unstable propagation of rock mass failure and their proximity to excavations where reinforcement and support will be installed. Ideally, the design event must be based on the history of seismic events at a particular mine and their correlation with other major influencing factors such as large faults and the stress concentrations (induced by mining) relative to the rock mass strength (Kaiser et al., 1996). This procedure assumes that the design event is remote from the surfaces of an excavation. However, it is also possible that the event source is in the immediate vicinity of an excavation wall. In this case, the mechanism of failure will result in a different form of dynamic loading of the ground support. It is worth noting that very high values of PPV have been measured without associated rock failure and ejection (Fleetwood, 2010). TABLE 7.2 Typical Rock Mass Demand for Ground Support Design Demand Category Low Medium High Very high Extremely high

Reaction Pressure (kPa)

Surface Displacement (mm)

Energy (kJ/m2)

400

300

35

Source: Modified from Thompson, A.G. et al., Geotech. Geol. Eng., 30(3), 553, 2012.

324

Geotechnical Design for Sublevel Open Stoping

7.3.1  Location of Failure due to Overstressing The analysis of stresses around underground excavations in rock can be accomplished using a number of different numerical analysis techniques. These can range from simple linear elastic analyses performed in two dimensions to complex three-dimensional nonlinear analyses. For complex geometries, two-dimensional analyses cannot be expected to provide meaningful guidance on the locations of failures. On the other hand, the latter types of analyses can be expected to provide the most detail and understanding of the changes in rock stresses as excavations are formed and extraction progresses. However, due to their complexity, they require significant resources to be expended in terms of testing to obtain realistic material properties and multiple back analyses to calibrate the models with documented on-site observations and experience (Pardo and Villaescusa, 2012). An intermediate approach is to use linear-elastic analyses in three dimensions (e.g., Wiles et al., 2004). This approach is able to model complex threedimensional models of excavations and sequences and to identify regions of high-stress concentrations and volumes of rock where it might be expected that the rock mass strength is exceeded (Figure 7.6). Again, however, it is essential that the model is calibrated with documented experience. The limitation of this approach is that the redistribution of stresses following progressive rock mass failure cannot be determined. Nevertheless, the most important outcome from the analyses is to identify areas that can be expected to experience ground stability problems due to excessive stress. Another methodology is that reported by Beck and Duplancic (2005). The basis of this method is the three-dimensional nonlinear modeling computer program Abaqus that can be used to predict the ground reaction curves at distinct locations for different extraction sequences. The energy release associated with the ground reaction curve at the particular location can also be predicted (Beck et al., 2010). 7.3.2  Depth of Failure: Stress or Strain Controlled The depth of rock mass failure around excavations can be estimated by calculating the strength factors for the rock mass near excavation surfaces. The aim of the analysis for intact rock failure is to determine the depth of failure to provide estimates for both reinforcement length demand and reinforcement and support capacity demand. At this point, it is worth noting that it may be possible to minimize or eliminate intact rock failure by modifying the excavation from a flat back to an arched profile. Analyses have shown that the stresses in the rock in the backs and shoulders of rectangular excavations are higher than when the excavation incorporates an arched profile (Figure 7.7). Martin et al. (1997) have found that the benefits of an arched back profile apply to both low- and high-stress environments. In both cases, the

325

Rock Reinforcement and Support

Strength factor-A 0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00

Broken Ground—2 m

UCS = 124 = –37°

Span—8 m

Anchoring zone

Reinforced rock mass

Cracked zone Broken–damaged zone

Span FIGURE 7.6 Modeled damage zones for rock reinforcement design. (After Wiles, T. et al., Rock reinforcement design for overstressed rock using three dimensional numerical modeling, in E. Villaescusa and Y. Potvin, eds., Ground Support in Mining & Underground Construction, Proceedings of the 5th International Symposium on Ground Support, Perth, September 28–30, 2004, pp. 483–489, Balkema, Leiden, the Netherlands.)

arch profile reduces the volume of failed rock that needs to be supported. However, in intermediate stress environments, a flat back profile was found to improve roof stability by forcing and restricting stress-induced failure to the confined regions of the shoulders. To predict the likely volume of failure, a particular site would require estimates of the in situ stresses and in situ rock strength and stiffness. For simple excavation shapes, graphical methods based on simple closed-form analytical solutions based on elasticity theory such as those presented in other text books (e.g., Hoek and Brown, 1980) could be used. Alternatively, the use of stress analysis programs will allow the stress distribution around the actual excavation shapes to be analyzed. Such an analysis could incorporate an appropriate failure criterion (e.g., Wiles et al., 2004—see Chapter 5). This failure criterion allows the depth of failure to be estimated.

326

Geotechnical Design for Sublevel Open Stoping

(a)

(b)

(c)

(d)

FIGURE 7.7 Excavation profiles for mine development in hard rock. (a) Square, (b) shanty back, (c) oval, and (d) semicircular (flat floor).

The depth of failure coupled with an estimate of a bulking factor allows an estimate to be made for the expansion of the rock surface. The estimates of failure depth and volume provide the rock mass demands that need to be satisfied by the reinforcement systems in terms of their length, force, and displacement capacities and the support systems in terms of their force and displacement capacities. The approach used by Beck and Duplancic (2005) is to conduct a nonlinear stress analysis (using Abaqus) and then to define the depth and volume of failure based on the calculated plastic strains. 7.3.3  Depth of Failure: Structurally Controlled In structured rock masses, it is possible to estimate ranges of blocks sizes formed from combinations of discontinuities with different orientations, persistence, and spacing. Rock mass characterization techniques are requ­ ired to determine the likely sizes and shapes of the unstable blocks to be supported by suitable reinforcement schemes. Depending upon the characteristics of the reinforcing scheme chosen, a suitable embedment length that ensures full capacity of the system must be designed and installed.

Rock Reinforcement and Support

327

FIGURE 7.8 Large potentially unstable wedge reinforced with cable bolting.

A similar reasoning applies for cable bolt reinforcement of large unstable wedges (Figure 7.8). Over the years, a number of procedures for examining the stability of single and multiple blocks of rock have been developed. Readers are referred to Thompson (2002) for a reasonably recent, comprehensive review of these methods. Single reinforced block stability analyses may be performed with the Rocscience program Unwedge or modules within the SAFEX package developed by Thompson (2002). A probabilistic design method developed by Windsor (1999) is also incorporated into the SAFEX package. This method uses the variability of discontinuity set orientations, persistence, and spacing combined with the excavation geometry to predict the range of possible block shapes and sizes. This information is then used to predict reinforcement lengths and the ground support pressure that needs to be provided. The ITASCA three-dimensional distinct element program 3DEC can be used to model the stability of block assemblies. The program allows for the analysis of the kinematics of the interactions between blocks and can be used to model failure mechanisms, stress redistributions, and the effects of reinforcement. An alternative approach, again incorporated into the SAFEX package, allows for the modeling of progressive unraveling in jointed rock mass assemblies and the analysis of reinforced block stability. This approach is described in detail by Thompson (2002).

328

Geotechnical Design for Sublevel Open Stoping

7.3.4  Ground Reaction Curve Concept The concept of displacement demand and appropriate reinforcement are best considered in terms of a ground reaction curve. Figure 7.9 shows a typical ground reaction curve (Windsor and Thompson, 1998), which is the relationship between radial stress and radial displacement at the boundary of an excavation. The radial direction is normal to the excavation surface. The stress reduces from its value before excavation. For a stable excavation, the radial stress will reduce to zero at a certain displacement. For unstable excavation surfaces, a restraining force (from support and/or reinforcement) is required to maintain the rock mass stability and excavation shape. Experience has shown that an equilibrium condition may be attained by limiting displacements so that the rock assists in maintaining stability. Large displacements are accompanied by rock mass loosening and may lead to larger stabilizing force requirements as the volume of failure expands. Nonlinear numerical modeling methods can be used to quantify the ground reaction curve for a given rock mass and excavation shape. It is known that the displacement demand will be a function of the stress regime and the mechanical and rheological properties of the rock. For example,

Characteristic force

Reinforced rock system response Excitation characteristic

Reinforced system response

Rock system response

Mechanistic characteristic Characteristic displacement FIGURE 7.9 Ground reaction curve showing the reduction of force with increased displacement. (From Windsor, C.R. and Thompson, A.G., Reinforcement systems—Mechanics, design and performance testing, in J. Orozco and J. Schmitter, eds., Proceedings of the Third North American Rock Mechanics Symposium, Cancún, June 3–5, Int. J. Rock Mech. Min. Sci., 35, 4–5, Paper 076, 1998, 9pp.)

Rock Reinforcement and Support

329

failure in rocks that behave in a ductile manner is accompanied by significant postyield creep displacements. On the other hand, brittle rock failure may be accompanied by a high-energy ejection of material at small displacements. The different types of rock mass behavior require support and reinforcement schemes with distinctly different characteristics. Beck et al. (2010) have used the Abaqus program to predict ground reaction curves. If the curve can be predicted, then it is possible to design ground support with an appropriate force–displacement response and capacities so that the rock mass pressure, displacement, and energy demands are satisfied. 7.3.5  Ground Support for Massive Rock and Low Stress Massive rock masses are characterized by a limited number of discontinuity sets with limited persistence and wide spacing between members of the sets. As shown in Figure 7.10, the impersistent discontinuities do not intersect to form distinct blocks of rock. Excavations formed in massive rock by drilling and blasting can be expected to have some blast damage and localized surface instabilities that can be scaled down. However, clean profiles can result from controlled drilling and blasting. This type of rock should not require surface support or internal reinforcement at the time of forming the excavation. However, allowance should be made for changes in the stress conditions that might occur as a result of future mining. 7.3.6  Ground Support for Massive Rock and Moderate Stress The creation of excavations in massive rock in a moderate premining stress field may cause localized stress concentrations at distinct locations around the excavation boundary. Figure 7.11 shows how stresses may cause failure at one shoulder and the toe of the wall on the other side of the excavation. These failures are induced by tensile cracking oriented normal to the minor

FIGURE 7.10 Massive rock with widely spaced discontinuities with limited persistence.

330

Geotechnical Design for Sublevel Open Stoping

Major principal stress

FIGURE 7.11 Localized crushing and spalling around an excavation.

principal stress. Postfailure control of the shoulder can be achieved with either mesh or shotcrete and pattern bolting for restraint of the support. It is also probable that failure at the toe of the wall may undercut the overlying rock and propagate upward. Support restrained by reinforcement should be used to maintain the stability of the toe. 7.3.7  Ground Support for Massive Rock and High Stress High in situ or induced stress regimes may exceed the strength of the intact rock and the rock mass. The failure modes may be similar to those shown for moderate stress but may occur more violently due to the energy stored in the rock mass. Also, in a highly stressed region of rock, sudden slip on major discontinuities in the vicinity of the excavation is more likely with the associated release of energy in the form of pressure and shear waves that excite the rock near the boundaries of excavations. These pressure and shear waves cause changes in the local stresses and vibrations that may be sufficient to initiate rock failure, loosening, and ejection as shown in Figure 7.12. In these circumstances where all excavation surfaces are likely to be affected, support using mesh-reinforced (embedded) shotcrete (Morton et al., 2009b) restrained by reinforcement is suggested. The shotcrete is in close contact with the rock and will provide immediate response to any rock mass movements preceding the seismic event remote from the excavation. As indicated previously, small increases in the minor principal stress normal to the excavation surfaces can increase the rock mass strength and inhibit fracture propagation. However, should the rock fail violently, the shotcrete may not have sufficient toughness to absorb the energy

331

Rock Reinforcement and Support

Removed by crushing or ejection

Vpp

Vs

Vp

FIGURE 7.12 Stress-induced violent failure with rock ejection.

associated with failure, and it may itself crack locally so that the localized displacements may not be able to be tolerated, even if the shotcrete is reinforced with fibers (Morton et al., 2009b). A layer of mesh restrained by bolts is flexible and is therefore able to sustain large rock mass movements and retain the failed shotcrete (Figure 7.13). The types of bolts that are used for shotcrete and mesh restraint may need to be specially designed to allow for the bulking of the rock associated with the rock mass failure. For example, the reinforcement element may need to be simply debonded from the rock near the collar. If the potential displacements are larger than the elongation of the element, specially designed anchors that slip may be required. In both instances, the stiffness of response to rock mass movement is reduced and can result in rock mass loosening. If this is a concern, then a combination of stiff and flexible reinforcement systems may be more appropriate.

(a)

(b)

FIGURE 7.13 Large deformation allowed by mesh-reinforced shotcrete. (a) Welded-wire and (b) woven-wire reinforcement.

332

Geotechnical Design for Sublevel Open Stoping

7.3.8  Ground Support for Layered Rock and Low Stress Stratified rock masses are characterized by continuous, approximately parallel planes with cross-jointing. In subhorizontally layered rock with crossjointing, the walls will be stable, but horizontal stresses are required to keep the vertical joints closed and create vertical frictional resistance to downward displacement in the roof. Consequently, if the stresses are low, the frictional resistance is insufficient to prevent the rock between vertical joints from falling. Progressive collapse can occur until a stable arch is formed as shown in Figure 7.14. The need for support will depend on the spacing between the vertical joints; that is, if the spacing is small, then mesh or shotcrete will be required to span between the restraint provided by the reinforcement. Reinforcement should be installed to intersect the vertical joints at an oblique angle to improve the shear resistance. Otherwise, reinforcement installed vertically may need to be longer to penetrate beyond the potential height of the stable arch. When the layering is dipping relative to an excavation, several different failure modes are possible as shown in Figure 7.15. In the absence of crosscutting joints, cantilever beams are formed in the roof of the excavation. Tensile stresses will form at the top of the cantilever, and cracks will form near the abutment or shoulder. These cracks will penetrate the full depth of the layer, and slabs of rock will fall into the excavation. This mode of failure can be prevented by the installation of reinforcement to restrain the free end of the cantilever. In the left wall, a toppling mode of failure may occur, especially if blast damage (which is frequently observed) undercuts the toe of the wall. This type of failure may be controlled by the installation for reinforcement angled upward to intersect the dipping layers. Mesh support may be required if the layers are thin. If possible, the lowest row of reinforcement should be installed horizontally to cross beyond the line of intersection with the floor of the drive. Failure by sliding may occur in

Rock fall

FIGURE 7.14 Arch formation in layered rock masses.

Rock Reinforcement and Support

333

FIGURE 7.15 Flexural toppling and sliding in layered rock masses.

the right wall. This mode of failure may be controlled by the installation of horizontal reinforcement that intersects the layers and improves the shear strength. Reinforcement should not be installed parallel to the layers. Again, mesh may be required for thin layers. 7.3.9  Ground Support for Layered Rock and Moderate Stress For moderate stresses in layered rock, the failure mode may involve a sequence of sagging, followed by buckling and eventual cracking, and failure as shown in Figure 7.16. The initial bending (sagging) is initiated by gravity forces. Following sagging, the induced stresses result in an increase in bending moments at both the center of the span and at the abutments. These bending moments result in tensile stresses at the lower surface of the rock beam at the center of the span and at the upper surface near the abutments. These tensile stresses result in crack propagation, and eventually two distinct segments of the beam may form. This mechanism

Major principal stress FIGURE 7.16 Buckling and cracking failure in bedded rock.

334

Geotechnical Design for Sublevel Open Stoping

of behavior is commonly referred to as the Voussoir beam, and its stability is controlled by the strength and stiffness of the rock and the horizontal stiffness at the shoulders of the drive (Diederichs and Kaiser, 1999). This mechanism is most effectively controlled by attempting to improve the shear resistance between the individual layers to form a thicker beam with improved resistance to bending and buckling. The reinforcement system resistance to shear for this mechanism is more important than the tensile strength. 7.3.10  Ground Support for Layered Rock and High Stress Figure 7.17 shows a layered rock mass in which the excavation is formed in a lower stiffness rock than the layers above and below. Following extraction, the less stiff rock will attempt to dilate more horizontally than the stiffer rock. This differential dilation will result in shear stresses being developed between the layers. This shear stress may result in vertical cracking in the less stiff layers and tensile cracking and shear failures in the stiffer layers near the shoulders and toes of the walls as shown. The dilation of the walls in the less stiff rock can be controlled (but not prevented) by the installation of horizontal reinforcement. The shear failures and potential overbreak from the roof can be controlled by reinforcement angled into the shoulders and vertically in the center of the span. 7.3.11  Ground Support for Jointed Rock and Low Stress Jointed rock masses are characterized by the frequent occurrence of rock discontinuities with variable persistence and spacing. The stability of blocks in jointed rock is controlled by the forces acting on the blocks and the shear strengths of the joints that form the faces. In many cases, at the time the

FIGURE 7.17 Tensile splitting, shearing, and sliding in bedded rock.

Rock Reinforcement and Support

335

FIGURE 7.18 Discrete large blocks falling or sliding from a rock mass.

excavation is formed, blocks are not fully formed; that is, the faces of the blocks have intact rock bridges. The rock bridges may be strong enough to maintain the stability of the blocks at this time. However, the changes in stresses caused by the excavation may result in the preexisting discontinuities propagating through the rock bridges to create fully formed blocks. After this time, the block stability is controlled by the orientations of the faces and the shear strengths of the fully formed faces. In a low-stress environment, the normal stresses acting across the joint faces are low and therefore the frictional shear resistances are also low. The shear stresses resisting sliding or falling are insufficient to prevent the failure modes shown in Figure 7.18. In this figure, the discontinuities are widely spaced and could be controlled by reinforcement installed normal to the excavation faces. An estimate of the maximum block size is required to enable an appropriate length of reinforcement element to be selected so that it penetrates beyond the unstable block into a rock mass region that is stable. Alternatively, if the discontinuities are closely spaced as shown in Figure 7.19, then surface support is also required to prevent unraveling and progressive large-scale collapse. In a low-stress environment, mesh is sufficient to retain the volume of failed rock. However, as mesh does not provide immediate restraint to loosening, the volume of failure may be larger and deeper than if shotcrete is used to provide immediate response to rock mass loosening. This observation is an important factor when considering the reinforcement length demand.

336

Geotechnical Design for Sublevel Open Stoping

FIGURE 7.19 Unraveling and progressive collapse of small blocks.

7.3.12  Ground Support for Jointed Rock and Moderate Stress A jointed rock mass in a moderate stress field may behave similarly to a massive rock. That is, the normal stresses acting across the joint surfaces may result in the shear strengths being greater than the shear stresses acting. Whether this is the case or not depends on the orientations of the joint surfaces relative to the orientations of the stresses. If the joints do not slide, then tensile cracking may occur as shown in Figure 7.20. These tensile cracks may coalesce and interact with the preexisting joints to form blocks that slide and rotate and result in a general dilation of the rock mass and deformation of the excavation boundaries. Both support and reinforcement are required to control this type of rock mass behavior. Shotcrete can be used for support together with reinforcement used for both restraint of the support and improved shear strength of the joints.

FIGURE 7.20 Tensile cracking, crushing, sliding, and dilation.

Rock Reinforcement and Support

337

7.3.13  Ground Support for Jointed Rock and High Stress As with moderate stress, a jointed rock mass may behave as a massive rock. This again depends on the orientations of the discontinuities relative to the stresses. Consequently, the failure modes shown in Figure 7.21 are similar to those shown in Figure 7.12. However, the support and reinforcement requirements are different. Failure at the right shoulder would result in a loss of horizontal confinement across the back, and the blocks would then fail due to gravity loading. Failure at the toe of the wall would result in undercutting of the blocks above and failure. Experience indicates that mesh-reinforced (embedded) shotcrete would be required to provide support with immediate (by the shotcrete) and postfailure (by the mesh) response to rock mass deformations. Reinforcement would be used for both restraint of the mesh-reinforced shotcrete and to stabilize rock discontinuities close to the excavation. As indicated previously, the surfaces may be stable immediately following the creation of the excavation. However, in highly stressed rock, cracks may gradually form and propagate with time. The area of cracks per volume of rock may eventually exceed some critical value at which time the rock will fail violently with fragments of rock ejected as shown in Figure 7.22. The time after the formation of the excavation at which this phenomenon occurs may range from seconds to weeks. These events are therefore a major hazard in a mine, as ejection may occur before appropriate ground support is installed or may occur from the face of an excavation during drilling and charging operations. Again, both mesh-reinforced shotcrete and reinforcement are required in a rock mass susceptible to this type of failure mode. As with the other rock types and conditions described earlier, an excavation may be stable for the stresses acting locally. However, in a highly stressed rock mass, there is a possibility of slip on discrete major structural

FIGURE 7.21 Crushing and spalling under high stress.

338

Geotechnical Design for Sublevel Open Stoping

FIGURE 7.22 Ejection of material due to stresses exceeding the strength of the rock at the boundary of an excavation.

Vp

Vs

Vpp

Incident and reflected seismic waves FIGURE 7.23 Detachment and ejection of a discrete block due to seismic waves from an event remote from the excavation.

features some distance from the particular location of the excavation. As indicated previously, these sudden, unpredictable slips can release energy in the form of pressure and shear waves that eventually reach the excavation (Figure 7.23). These waves are reflected at the excavation boundary, and stress changes result that may be sufficient to cause crack propagation, failure, and massive ejection of rock. 7.3.14  Design by Precedent Rules Precedent rules can be applied in conjunction with the concepts detailed in the previous sections. Brief details and discussion of the more common systems in use are given in the following sections. It is left to the reader to

339

Rock Reinforcement and Support

examine the sources of information and to assess whether the particular rule or system can be applied in any given circumstance. Precedent rules, based on back analysis of reinforcement that was used and found to be effective in civil engineering structures, were developed in the late 1950s (Lang, 1961). Note that these rules do not take account of the rock mass quality and the stress regimes in which the excavations were formed. For reinforcement length and bolt spacing, the following rules have been found to apply:



L min = The largest of 2s, 2b,

or

L min = The largest of 2s, 2b,

B (B < 6 m) (7.1) 2

B H (B > 18 m),† (H > 18 m) (7.2) 4 5

L smax = The smallest of  or 1.5b (7.3) 2

where L is the bolt length s is the bolt spacing b is the mean block width B and H are the excavation width and height, respectively Various other suggestions for reinforcement length have been proposed as follows (Rabcewicz, 1955; Pender et al., 1963; Benson et al., 1971; Cording et al., 1971):

L = 0.3B (7.4)



L = 1.829 + 0.0131B 2 , ≥ 3b (7.5)



L = 0.25 to 0.30B (7.6)



L = 0.35B (7.7)



L = 0.1 to 0.5H (7.8)

Choquet and Hadjigeorgiou (1993) presented a summary of the length estimates from various sources (e.g., Coates and Cochrane, 1970; Farmer and Shelton, 1980; USACE, 1980; Laubscher, 1984). It is found that the predictions of reinforcement lengths given by most of the expressions are reasonably consistent and that more complicated expressions are not required. It is also important to note that the actual required length of a reinforcement system

340

Geotechnical Design for Sublevel Open Stoping

will depend on the force demand and the load transfer mechanism. For example, a CFC system (e.g., split set) will need to be longer than a DFMC device (e.g., expansion shell anchor) to achieve the same force capacity. For excavation crown pressure demand, Cording et al. (1971) suggested that

Pc = nBg(kPa) (7.9)

where B is the crown span (m) γ is the unit weight of the crown rock (30 kN/m3) n is a constant ranging from 0.1 to 0.3, that is, for a span of 6 m; this formula predicts a crown pressure Pc in the range from ∼20 to ∼60 kPa For walls, the pressure demand, Pw, is given by

Pw = mBg(kPa) (7.10)

where B is the wall span (m) γ is the unit weight of the wall rock (30 kN/m3) m is a constant ranging from 0.05 to 0.15 This equation indicates that, statically, the pressure demand for excavation walls is about 50% of that for backs/roofs/crowns. 7.3.15  Design by Rock Mass Classification The Q system (Barton et al., 1974; Grimstad and Barton, 1993) is probably the most widely used rock mass classification system. However, it should be used with caution, particularly in regard to some of the design expressions that have been developed. It is worthwhile noting that the database was originally developed from case studies of civil engineering tunnels at shallow depths. The chart shown in Figure 5.7 is used to estimate ground support based on the Q value and the span or height of an excavation surface. The Q system incorporates relationships to estimate minimum reinforcement length. For example, rock bolt lengths are estimated using



L =2+

0.15 B (7.11) ESR

where B is the width or height of an excavation surface ESR is the excavation support ratio (see Table 5.1)

341

Rock Reinforcement and Support

The value of ESR depends on the intended function of the excavation and ranges from 0.8 for public infrastructure excavations to more than 1.5 for mine excavations. As an example, for rock bolts in a 5 m by 5 m permanent development heading (ESR = 1.6), L = 2.5 m. This is in agreement with the precedent mining practice. Note, however, that for individual blocks or stress-driven failure, longer bolt lengths may be required. Two formulae have been proposed as part of the Q system to calculate the excavation roof/back/crown pressure demand:



Proof =

200 Jn (kPa) for Jn < 6 (0 to 2 sets) (7.12) 3JrQ1/3

and

Proof =

200 (kPa) for Jn > 6 (more than 2 joint sets) (7.13) JrQ1/3

In most rock masses, the Jn value will be greater than 6 and therefore Table 7.3 shows the variation of Proof predicted using Equation 7.13 for rock masses ranging in quality from very poor to good. Note that the SRF value can range from 0.5 to 20 (see Figure 5.10) with corresponding large changes in the Q value and predicted values of Proof. A pressure demand of 312 kPa can be satisfied by twin strand cable bolts (500 kN capacity) on 1.25 m by 1.25 m pattern. However, in poor quality rock, this reinforcement would need to be complemented by a shotcrete layer to retain the small block sizes. Similar calculations for other reinforcement systems can be made to satisfy the other pressure demands given in Table 7.3. As the stress levels increase and the energy release accompanying failure increases, it can be concluded that there is a need for increased force and displacement capacities in both reinforcement and support. TABLE 7.3 Examples of Roof Support Pressure as a Function of Q Value Rock Quality Parameter RQD Jn Jr Ja Jw SRF Q Proof

Very Poor

Poor

Fair

Good

25 12 1 4 1 2 0.26 312

50 12 1 2 1 2 1.04 197

75 9 1.5 1 1 2 6.25 73

95 9 2 1 1 2 10.6 46

342

Geotechnical Design for Sublevel Open Stoping

The choice of an appropriate stiffness is an inherently difficult task when based simply on a classification such as the Q system. A higher-stiffness element can arrest rock movement with less displacement. However, the penalty is a higher force generated in the element. On the other hand, a low-stiffness element allows for greater displacement but may not absorb the released energy before the rock mass has significantly loosened to a point where serviceability requirements mean that failure has effectively occurred. One way of estimating displacement demand may be simply to assume that the stress change (from pre-mining to post-mining) occurs over a length (L) observed for a particular mine site, and the rock mass deformation modulus (Em) may be estimated from one of the several expressions available in the literature (e.g., Serafim and Pereira, 1983; Hoek and Brown, 1997; Zhang and Einstein, 2004; Hoek and Diederichs, 2006):





Em = 10 Em =

RMR-10 40

sc 10 100

(GPa) (for sc > 100 MPa) (7.14)

GSI-10 40

(GPa) (for sc < 100 MPa) (7.15)

where RMR is defined by Bieniawski (1976) GSI is the Geological Strength Index introduced by Hoek (1994) (see also, Hoek et al., 1995) The most up-to-date of these expressions is probably that due to Hoek and Diederichs (2006) and given by Equation 4.21. The displacement, δ, is then given by

d=

Ds (7.16) L Em

For example, if the average stress decrease is 40 MPa in a rock mass with E = 50 GPa over a depth of 2 m, δ = 1.6 mm. On the basis of experience, this displacement is considered to be unrealistically low for a highly stressed rock mass where loosening could be expected to accompany destressing. An alternative approach is to use the plastic strain obtained from nonlinear stress analysis. If the plastic strain is assumed to be about 5% over a 2 m depth, then the associated excavation wall displacement is about 100 mm. Another approach might be to consider the depth of failure and the bulking associated with rock mass failure and dilation. For example, if the depth of failure is observed to be approximately 1.5 m and the volume increase associated with failure is assumed to be say 20%, then an excavation wall would move about 300 mm. This magnitude of boundary displacement is considered to be more reasonable.

343

Rock Reinforcement and Support

7.3.16  Reinforcement Layout Several assumptions are implicit in the approach detailed in the preceding sections: 1. Reinforcement is justified (in terms of both safety and production requirements) and economically viable. 2. The reinforcement can be installed evenly spaced within the excavation surface associated with the failure volume. 3. The reinforcement will actually pass beyond the failure volume. All these assumptions may usually be satisfied within most stope development excavations. The average square spacing (s meters) can be determined from



s=

C (7.17) p

where C is the reinforcement design capacity (kN), not necessarily the maximum force capacity p is the pressure demand (kPa) For rectangular patterns



rs =

C (7.18) p

where r is the spacing within a ring s is the strike spacing of rings 7.3.17  Energy Release Conceptually, rock fails violently when the unloading stiffness of the surrounding rock mass is softer than the unloading stiffness of the volume of failing rock (Jaeger and Cook, 1976; Brady and Brown, 2004). It may be possible to precondition the rock mass so that these conditions do not occur. That is, the intact rock needs to be damaged prior to the formation of the excavation so that these conditions do not occur. Preconditioning of the rock mass has been used successfully at many mines (e.g., Board and Fairhurst, 1983; Chacon et al., 2004).

344

Geotechnical Design for Sublevel Open Stoping

South African and Canadian workers have provided a number of examples of the range of velocities, typical masses, and kinetic energies that have been measured or estimated for dynamic failure. For example, it has been suggested that the kinetic energy is generally in the range 20–30 kJ/m2 with a maximum velocity of 1.5–2 m/s and a displacement demand of about 150 mm. Other authors have suggested that kinetic energy may be up to 25 kJ/m2 with velocities of ejection of 2–3 m/s. Ortlepp (1992) has inferred that block velocities after dynamic failure may be considerably higher than these values, having measured an ejection velocity of about 7.5 m/s after a displacement of about 50 mm. The data provided earlier can be used to design ground support schemes that have the necessary energy and displacement capacities to survive violent rock mass failures. It is worth noting that the energy dissipation depends on both the ability of the ground support to deform and the system force capacity. Displacement is particularly important. For example, although a reinforcement system may have large displacement capacities, it may cause the rock mass to disintegrate to the point where the support system may not be able to hold the broken rock. Systems that absorb large amounts of energy, but allow large deformations are not really suitable for excavation stability. The objective should be moderate, say 100–200 mm, reinforcement displacement that is compatible with stable surface support systems (mesh and shotcrete) at the boundaries of excavations. 7.3.18  Rock Mass Demand The required force–displacement response and capacities of reinforcement should ideally be matched to the rock mass demand. This rock mass demand may be applied directly from the rock mass or through the support that is retained by the reinforcement. In almost every case, this rock mass demand is very difficult to quantify (the possible exception to this is the reinforcement of a discrete fully formed block). On the other hand, the demand may change with time for some rock masses. For example, a stiff response may be required in the short term to minimize rock mass loosening, while in the longer term, the reinforcement system may be required to absorb large displacements as the block size reduces and the rock mass creeps (Figure 7.24). In this case, a single reinforcement system may not be able to provide both the short- and long-term properties required to satisfy the rock mass demand. This also applies to areas that may be susceptible to sudden failure of the rock mass due to overstressing where the requirement of the reinforcement system to absorb energy may be incompatible with the short-term requirement to provide a stiff response to static rock mass movement and the ability to sustain the displacements associated with rock mass bulking. Support demand is even more difficult to predict due to the fact that the rock mass characteristics may change with mining and time. For example, a

Rock Reinforcement and Support

345

FIGURE 7.24 Observed damage near the boundary of an excavation in hard rock under very high stress.

massive rock mass may change to a broken rock mass following failure due to overstressing (Figure 7.24). Initially, there is apparently no demand for surface support (or even for reinforcement). Following failure, there is a definite need for surface support to retain the broken rock and the need for the support to be restrained by the associated reinforcement.

7.4  Rock Bolting of Open Stope Development Drives A selection of typical reinforcing elements will be discussed in accordance with the classification of reinforcement presented in Section 7.2. An understanding of the different elements is important, as no single reinforcing scheme is likely to match the range of observed ground behavior at a particular mine site. This is because of the likely range of failure mechanisms that can be experienced throughout a stope extraction process. Reinforcement systems may be broadly characterized as rock bolts or cable bolts according to their length. Rock bolts are generally less than 3 m long while cable bolts are longer than about 5 m. The mechanical properties vary widely as do the installation requirements. In general, rock bolts may be classified as one-pass or two-pass systems. It has been found that one-pass systems are preferred in many mines. However, the installation procedure may not be compatible with the requirement to also restrain mesh, and the mechanical properties may not be appropriate for

346

Geotechnical Design for Sublevel Open Stoping

the expected rock mass demand (in terms of one or more of force capacity, displacement capacity, or energy absorption). 7.4.1  Continuous Mechanical Coupled Rock Bolts CMC rock bolts rely on a grouting element that fills the annulus between the element and the borehole wall. The strength of the system is a function of the nominal element capacity, the grout strength, and the active embedment length. The coupling agent can be either a cement- or a resin-based grout. 7.4.1.1  Cement-Encapsulated Threaded Bar A typical cement-encapsulated threaded bar consists of a 2–3 m long, 20–25 mm diameter corrugated bar that is grouted along its entire length. The bolts are usually manufactured with a variable cross-sectional shape to provide effective geometrical interference between the grout and the bolt surface. The geometrical interference creates a mechanical interlock that extends over the entire length of the element. Figure 7.25 shows a cross section through a typical bolt and its components. The critical embedment length for a typical water cement ratio of 0.35 is approximately 0.5 m. A dense grout mix increases the bond strength, both in the bolt-to-grout contact and in the grout-to-rock contact. Each bolt provides long-term reinforcement exceeding 15 tons/m of embedment (Figure 7.26). However, this depends upon the strength of the grout mix, with the main cause of failure observed being slippage (shear failure) at the bolt–grout interface (Figure 7.27). Cement-encapsulated rock bolts can be used for long-term reinforcement in areas where stress-related damage is expected, or where weathering effects over a long period of time would make an ungrouted point-anchored rock bolt unreliable. Experience also suggests that the system may be too stiff to be used in rock masses likely to undergo large deformations or sudden movement. Cracking of the grout across a geological discontinuity may cause corrosive damage to the rock bolts, due to water being able to reach the steel bar, and sometimes the bar is galvanized prior to installation. In addition, this rock bolt may be susceptible to blast damage (flying rock hitting the exposed fine thread at the plate end) when installed very close to an active face. A coarse threaded bar can be used to overcome this problem. However, coarse threads do not allow active restraint to be maintained and can fail under dynamic loading (Player, 2012). The reasons for this are, first, the short free lengths between the internal and external fixtures, which means that small axial displacements result in larger strains (and accordingly larger force changes) than with a longer free length, and second, a coarse thread has a large helix angle, which means that a nut can rotate more easily than on a fine thread (e.g., standard metric thread) that has a smaller helix angle.

347

Rock Reinforcement and Support

Hemispherical plate Nut welded with hemispherical washer

150 mm M20 thread Right hand

33–45 mm hole (grouted)

FIGURE 7.25 Typical components of a cement-encapsulated threaded bar. (Courtesy of Mount Isa Mines, Mount Isa, Queensland, Australia.)

7.4.1.2  Resin-Encapsulated Threaded Bar In cases where immediate reinforcement is required, resin-encapsulated threaded bars can be used (Kaiser et al., 1996; Mikula, 2004; Varden, 2005). In order to install rock bolts that are coupled with resin along their entire lengths, it is necessary to insert multiple resin cartridges with sufficient volume of resin to fill the annulus between the rock bolt and the borehole. The typical rock bolts being used in underground hard rock mines have been modified from the rock bolts used in coal mining industry. The modifications have been necessary due to the need to drill larger hole diameters with the type of equipment used in hard rock metalliferous mines. The modification is mainly in the form of paddles or the use of a spring welded onto the end section of the rock bolts. Figure 7.28 shows the anchor sections for a 24 mm diameter Posimix bolt with a spiral arrangement and a 27 mm diameter Secura bolt showing a paddle arrangement. The Posimix spiral is 3 mm in diameter and has a length of 500 mm. The Posimix system is designed

348

Geotechnical Design for Sublevel Open Stoping

1 m (double) embedment length 0.35 W/C–7 days strength

250

Load (kN)

200 150 100 50 0

0

10

20 Displacement (mm)

40

30

FIGURE 7.26 Typical load–displacement response for cement-encapsulated threaded bar.

Local grout failure

Shear failure

Local crushing Shear loading Joint opening

Joint opening

Local crushing

Shear failure

Axial loading FIGURE 7.27 Failure by slippage at the bolt–grout interface. (Courtesy of Mount Isa Mines, Mount Isa, Queensland, Australia.)

Rock Reinforcement and Support

349

(a)

(b) FIGURE 7.28 (a) Posimix and (b) Secura bolts showing spiral and paddle mixing arrangements, respectively.

to push the resin cartridge plastic to the end of the hole. Additionally, the system allows the rock bolt to be centrally located in the hole allowing even distribution and mixing of the resin. The Secura paddles are 29.2 mm wide and are sheared into the end of the bolt for the purpose of mixing resin. The introduction of mechanized resin-anchored bolting using jumbos has been difficult to implement economically due to the high cost of resin transport and storage: Depending upon the local weather, underground temperatures, and humidities, this may require the use of refrigerated trucks and surface and underground storage facilities. Other problems include speed of bolt installation, including the ability to install mesh on a single pass, poor matching of bolt diameter to jumbo-drilled hole diameters, as well as operator skills. In the case of resin-encapsulated rock bolts, experience from in situ pull testing shows that high transfer loads can be achieved over short embedment lengths. However, cartridge resin systems may suffer from either underspinning or overspinning. Underspinning results in poor mixing and low resin grout strength, often at the critical anchor end of the hole. In some cases, the resin will never set. Overspinning during installation can result in shearing of the partially cured resin. This results in a reduced bonded area and lower load transfer. In addition, gloving of the rock bolts by the plastic packaging may occur completely eliminating load transfer along the bolt axis (Mould et al., 2004; Villaescusa et al., 2008). The performance and ultimate capacity of a reinforcement scheme can be affected by substandard installation practices. However, in CMC schemes, faulty installations are difficult to detect, given that the only visible parts of an installed element are the plate, nut, and a short length of the bolt indicating the orientation of installation with respect to an excavation wall. Thus, for a fully resin-encapsulated threaded rebar, it is very difficult to determine

350

Geotechnical Design for Sublevel Open Stoping

FIGURE 7.29 Bolt overcoring showing negligible resin migration within large shear zones.

the actual bonded length (bolt encapsulation) along the entire axis of the bolt. In addition, because the full steel capacity may be mobilized with very short embedment lengths of good quality resin, pull testing of exposed collar lengths within a fully encapsulated element is almost meaningless. Pull testing provides only an indication of resin effectiveness at the collar or at the first (unknown) location along a rock bolt axis where the resin is working effectively. It provides only a definite indication of poor installation in cases where the entire length of resin-encapsulated reinforcement fails at well below its designed capacity. Examination of the entire length of a fully encapsulated rock bolt in situ can be achieved by overcoring of the reinforcement element (Hassell and Villaescusa, 2005; Villaescusa et al., 2008). Rock bolt overcoring not only allows the recovery of the element, but also provides a clear view of the surrounding rock mass and a better understanding of the rock bolt system/rock mass interaction (Figure 7.29). It provides a range of information, including location and frequency of geological discontinuities, overall rock mass conditions, bolt encapsulation, load transfer along the bolt axis, and corrosion effects. Overcoring in broken ground or shear zones shows that very little resin migration occurs in jumbo-installed resin-encapsulated bolts. The resin simply fills the annulus between the bolt and the borehole. Because of its viscosity, the resin is unable to penetrate the rock mass fissures and voids. In comparison, significant cement migration has been observed during overcoring of cement-encapsulated bolts in poor ground conditions (Figure 7.30). The degree of rock mass interlocking using cement grout is superior to that achieved by resin grouting or friction stabilizers. Interlocking around an underground excavation has been suggested as an important mechanism to allow the rock mass to be self-supporting (Windsor and Thompson, 1993). Figure 7.31 shows overcoring results for 27 mm Secura bolts installed in basalt at the Bullant Mine near Kalgoorlie, Western Australia, using

351

Rock Reinforcement and Support

(a)

(b)

FIGURE 7.30 Broken rock mass interlocking (a) friction rock stabilizers and (b) cement-encapsulated rock bolts. 250 Collar region

Middle region

Toe region

Relative residual load (kN)

200

150

100

50

0

Embedment length (300 mm) Secura M27–35 mm hole

Secura M27–33 mm hole

300 mm embedment length FIGURE 7.31 Load transfer variability along bolt axis for resin-encapsulated Secura bolts. (From Varden, R., A methodology for selection of resin-grouted bolts, MEngSc thesis, Western Australian School of Mines, Curtin University of Technology, Kalgoorlie, Western Australia, Australia, 2005, 113pp.)

33 and 35 mm holes. Similar embedment lengths (300 mm) were tested. The results show that similar strengths were found for the collar and toe regions, with increased strengths for the middle regions where resin mixing appears to be more effective. The residual loads were measured at 15 mm displacement.

352

Geotechnical Design for Sublevel Open Stoping

7.4.2  Continuous Friction Coupled Rock Bolts CFC elements rely on the load transfer resulting from friction between the reinforcement element and a borehole wall. The actual strength per meter of embedment length of a CFC element is limited by the radial prestress setup during installation. 7.4.2.1  Split-Tube Friction Rock Stabilizers Thin-walled 47 mm diameter galvanized friction stabilizers are extensively used as reinforcement for stope development access. Such reinforcement elements generally have a nominal wall thickness of 3 mm and are mechanically installed using jumbos. This type of rock bolt consists of a hollow rolled tube having a slot along its entire length, which is driven into a drilled hole of smaller diameter. It relies on friction between the tube and the rock to provide reinforcement (Figure 7.32). Friction bolts are simple and quick to install, while standing up to blast vibrations relatively well. However, they have a very low initial bond strength per meter of embedment length. A capacity of approximately 4–5 tons/m of embedment has been established for 46–47 mm diameter elements (Figure 7.33). This may be insufficient to guarantee effective reinforcement of wedges, blocks, and slabs potentially formed within the immediate backs of excavations. The initial bond strength is developed during bolt insertion, where the drillhole tolerance with respect to bolt diameter is small and is likely to control the available frictional forces along the bolt length. In soft ground, the driving time to completely install a bolt is sometimes reduced indicating an even lower initial bond strength per meter of embedment length. Despite their low bond strength limitation and their susceptibility to corrosion (Hassell and Villaescusa, 2005), split-set bolts are used extensively

FIGURE 7.32 Schematic of the installation process for friction rock stabilizers.

353

Rock Reinforcement and Support

Spilt set bolts (SS46) Ungrouted strength

14 12

Load (tonnes)

10 8 6 4

Thalanga mine Stawell mine Hilton mine 1991 Hilton mine 1996

2 0

0.0

0.5

1.0 1.5 2.0 2.5 3.0 Embedment length (m)

3.5

4.0

Ungrouted SS46 1 m of embedment length

8 7

Load (tonnes)

6 5 4 3 2 1 0

0

2 4 Deformation (mm)

6

8

FIGURE 7.33 Load transfer for fully coupled friction bolts. (From Villaescusa, E. and Wright, J., Permanent excavation reinforcement using cement-grouted split set bolts, Proceedings of the AusIMM, No. 1, 1997, pp. 65–69. With permission.)

354

Geotechnical Design for Sublevel Open Stoping

throughout the mining industry even for permanent back reinforcement in blocky ground. This is because of the advantages that the system has to offer. These can be listed as

1. Immediate reinforcement to the face where damage from development blasting is minimal. 2. Low-cost mechanized bolt and mesh installation with a minimum of components. 3. Excavations can be meshed at a later date by installing a short friction bolt (having a smaller diameter) inside a previously installed friction bolt element. 4. Rock bolts can be installed into partially collapsed holes, providing reinforcement in poor ground conditions and reducing the number of holes that require redrilling. 5. In some cases, corrosion resistance can be minimized with the use of galvanized or stainless steel elements. A disadvantage is that the load transfer for a split tube friction rock stabilizer is usually limited to values that are usually insufficient to mobilize the force capacity of the element. This is particularly so if the borehole is oversized. It is worth noting that an undersized borehole may cause yield of the steel cross section. If shear occurs across the borehole in which a split tube rock bolt has been installed, then sliding in the toe region may be prevented, and the rock mass movement may be sufficient to cause the welded ring to be sheared off with a loss of the plate at the collar. Also, it is important to note that split tube bolts are susceptible to corrosion damage. Figure 7.34 shows a number of overcored friction bolts ranging in age from 1 to 5 years. Laboratory pull testing was devised to investigate the loss of frictional capacity due to corrosion over various embedment lengths in the range of 250–500 mm (Hassell and Villaescusa, 2005). Figure 7.35 clearly shows a loss of load-bearing capacity due to corrosion; the moderately corroded elements generally have twice the load-bearing capacity of their highly and severely corroded counterparts. 7.4.3  Discrete Mechanical or Friction Coupled Rock Bolts Discrete mechanical or friction coupled (DMFC) elements are point anchored and rely on load transfer over a relatively short interval of their total length. Chemical grouting is used to provide a frictional coupling element that in most cases is less than 500 mm in length. Mechanical coupling is provided by expansion-type anchorages that are shorter than 200 mm in length. External fittings such as face plates are an essential component of a DMFC system.

Rock Reinforcement and Support

355

FIGURE 7.34 Overcored friction bolt elements, ready for pull testing. (From Hassell, R.C., Corrosion of rock reinforcement in underground excavations, PhD thesis, Western Australian School of Mines, Curtin University of Technology, Kalgoorlie, Western Australia, Australia, 2007, 277pp.)

7.4.3.1  Expansion Shell Rock Bolts These discrete frictional coupled elements consist of 16–25 mm diameter steel bars (of varying lengths) that are installed with point anchor expansion shells in conjunction with face plates (Figure 7.36). The tension to the bolts is provided by tightening a nut hemispherical washer and a plate against the rock on the exposed ends of the bolts. Mechanically anchored bolts are capable of providing very reliable anchorage in hard rock applications where the rock mass has a high uniaxial compressive strength. One of the main disadvantages of mechanically anchored rock bolts is that if the anchor slips or the rock breaks around the plate, the capacity of the bolt drops to zero and the rock around the bolt can fail. In some cases, short threaded lengths (at the plate end) make the tightening of the plate against the rock very difficult to achieve, especially in uneven rock faces. The standard point anchor systems can be susceptible to corrosion and may not be effective in heavily broken rock masses in which an anchor point cannot be secured.

356

Geotechnical Design for Sublevel Open Stoping

60

Corrosion rate Severe High Moderate Light

Load (kN)

50 40

FB5a

FB1

FBx

30 20

FB6

FB8 FB3

10 0

0

2

4

6 Displacement (mm)

8

10

12

FIGURE 7.35 Galvanized 47 mm diameter friction bolts—400–500 mm embedment lengths. (From Hassell, R.C., Corrosion of rock reinforcement in underground excavations, PhD thesis, Western Australian School of Mines, Curtin University of Technology, Kalgoorlie, Western Australia, Australia, 2007, 277pp.)

Expansion shell

Bolt

Hemispherical plate Hardened washer Nut FIGURE 7.36 Components of an expansion shell rock bolt. (Courtesy of Mount Isa Mines, Mount Isa, Queensland, Australia.)

357

Rock Reinforcement and Support

Kanowna Belle mine CT bolts 2.4 m long

Load (tonnes)

25

25

20

20

15

15

10

10 FW conglomerate FW conglomerate FW conglomerate

5 0

0

5

10

15 20 25 30 35 Displacement (mm)

Cannington mine CT bolts 3 m long

Schist Pegmatite FW zinc

5

40

0

0

10

20 30 40 Displacement (mm)

50

60

FIGURE 7.37 Load–displacement responses for expansion shell–anchored bolts. (From Villaescusa, E. and Wright, J., Reinforcement of underground excavations using the CT bolt, in E. Villaescusa, C.R. Windsor, and A.G. Thompson, eds., Rock Support and Reinforcement Practice in Mining, Proceedings of the International Symposium on Ground Support, Kalgoorlie, Western Australia, Australia, 15–17 March, 1999, pp. 109–115. Rotterdam, the Netherlands, A.A. Balkema.)

The initial bolt installation can be mechanized to provide an immediate reinforcement force of approximately 10–15 tonnes. However, because of the short internal coupling, the actual point anchor strength is limited by the strength of the rock around the borehole (Figure 7.37). Point-anchored bolts tend to slip progressively due to blast vibrations when installed very close to an active face. An initial tension of approximately 7 tonnes is often used to reduce subsequent loosening due to blast vibrations. 7.4.4  Rock Bolts with Yielding Mechanisms The term yielding has been introduced and accepted by others as the appropriate terminology for rock bolt systems that have high energy dissipation capacities. Unfortunately, this term does not distinguish between systems that involve true material yield of the element or sliding movement at the anchored section of a rock bolt system. These high energy dissipation systems can be classified as follows:

1. Those involving mainly anchor slip relative to an internal fixture at a force less than the yield strength of the element. 2. Those involving mainly material yield in a decoupled length between discrete fixed anchors. 3. Those involving a combination of anchor slip and element yield. In the interests of clarity and concentrating on principles rather than specific products, it is worthwhile to review the rock bolts that have been developed

358

Geotechnical Design for Sublevel Open Stoping

to address problems associated with dynamic loading and the large displacement and energy-dissipation capacities required to maintain excavation stability. This is an area of contemporary interest and development (Player et al., 2004, 2009; Thompson et al., 2004). An early attempt to improve load transfer for strand-based cable bolts, while providing increased elongation between anchors, was reported by Schmuck (1979). A similar system with decoupling of the strand between fixed anchors was reported by Matthews et al. (1983) and demonstrated to be effective in maintaining the stability of highly stressed open stope crown pillars. The decoupling was achieved either by plastic sleeves or, more simply, by coating cable bolt strand with plastic paint. A recent development, the D-Bolt (Li, 2010), can be considered to have evolved from these earlier ideas of using the element elongation to dissipate energy between discrete fixed anchors. Conway et al. (1975) tested a mechanical anchor that allowed for sliding of a standard rock bolt through a fixed smooth bore die and reported that this system was developed in South Africa by Ortlepp and Read (1970). Thus, the Garford Solid Dynamic Bolt (Varden et al., 2008) and Roofex (Neugebauer, 2008) developed during the last decade can be considered to be commercial products that have evolved from these much earlier ideas. Another example of using element sliding relative to the internal fixture was the cone bolt developed at the CSIR in South Africa (Jager, 1992). The cone bolt is believed to be the first bolt designed to use a sliding mechanism to dissipate energy. The bolt consists of a plain bar with an expanded cross section at the toe end and a thread, nut, washer, and plate at the collar as shown in Figure 7.38. The expanded cross section of the bar is designed to provide resistance to pull out that is controlled by the strength and stiffness of the cement grout that encapsulates the bolt within a borehole. The shaft of the bolt is coated with saponified wax so that there is little or no resistance to movement of the bolt relative to the cement grout. The initial prototypes

Decoupled (waxed) length Cement grout Cone anchor

Cement grout

Smooth shaft with wax coating

Thread, nut and spherical base washer

Steel domed plate

FIGURE 7.38 Cone bolt anchor designed to pull through cement grout and increase displacement and energy absorption capacities.

359

Rock Reinforcement and Support

were manufactured from 16 mm diameter bar, and the majority of testing was performed on these bolts. Subsequent to the final development of the original cone bolt, demand for higher-capacity elements resulted in a version based on 22 mm diameter plain bar. It is believed that only limited testing has been performed for this bolt (Player, 2012). The design of the cement grout should be such that the anchor ploughs (pulls) through the grout column at a force less than the yield strength of the bolt. Both static and dynamic tests have shown that this is not the case with strong grouts, and much of the elongation is stretch of the bar (Player, 2012). It is therefore critical that both • The cement grout properties are designed to ensure that the cone pulls through the grout at a force less than the yield capacity of the bar • The equipment and procedures used for mixing and placing the cement grout paste in a borehole result in consistent strength and stiffness of the hardened cement grout Figure 7.39 shows a number of dynamic testing results obtained by Player (2012). The energy dissipation ranged from 10 to 60 kJ, with a 25 kJ dissipation achieved at approximately 150 mm of displacement (Player, 2012). More recently, a modified cone bolt was developed in Canada. This bolt, like the original cone bolt, has an expanded end but is designed to be 300

Dynamic force (kN)

250 200 150 100 50 0

0

50

100

150 200 250 Deformation (mm)

300

350

FIGURE 7.39 Performance of 22 mm diameter cone bolts. (From Player, J.R., Dynamic testing of rock reinforcement systems, PhD thesis, Western Australian School of Mines, Curtin University of Technology, Kalgoorlie, Western Australia, Australia, 2012, 501pp.)

360

Geotechnical Design for Sublevel Open Stoping

encapsulated with resin grout from a two-component cartridge that is mixed during installation. The breaking of the cartridge and mixing of the resin are aided by a flat paddle attached to the expanded end. The reported results (e.g., Simser et al., 2002; Gaudreau et al., 2004) show that this bolt performs either by gross anchor displacement or element extension, but sometimes by a combination of both mechanisms. The fact that the bolt eventually breaks suggests that the ploughing effect eventually ceases. An important consideration for any high energy-dissipation strategy is to test the full reinforcement system, including anchors, bolts, and plate/­ hemispherical nut assemblies together. Systems that dissipate large amounts of energy, but allow large deformations are not suitable. The objective should be to limit the reinforcement displacement, such that it is compatible with stable surface support systems (mesh and shotcrete) at the boundaries of excavations. In order to enable dynamic rock reinforcement design, the rock mass demand in terms of the ranges of displacement and energy presented in Table 7.2 has been combined with the WA School of Mines reinforcement dynamic capacity database (Player, 2012). The suggested design chart for rock reinforcement under dynamic loading is shown in Figure 7.40. For each rock mass demand category (Table 7.2), the corresponding ranges of displacement and energy were used to define a region (shown as a box) that has been labeled low, medium, high, and very high energy demand. For each region, the acceptable bolts should have similar displacement compatibility, while providing higher energy dissipation. That is, for each demand region, the recommended appropriate reinforcement would plot within the green (design) region. At this time, research on complete ground support schemes that include compatible support and reinforcement systems in terms of displacement compatibility is ongoing. Nevertheless, displacement at failure exceeding 300 mm is deemed very significant, given the typical bulking factors that follow dynamic rock mass failure at an excavation boundary (Figure 7.41).

7.5  Cable Bolting of Open Stope Walls Cable bolt reinforcement is used to stabilize large single blocks or wedges formed in the backs and walls of stope development infrastructure. In addition, cable bolts provide effective reinforcement of stope walls where normal rock bolts would prove geometrically inadequate due to their short embedment lengths. For stope wall reinforcement, the cable bolts are usually installed from drilling drives internal to a stope void. The main objective is to stabilize the rock mass around a stope before the stope is extracted. As an alternative to installing cable bolts from stope drill drives, special

Energy dissipated (kJ) 0

10

20

30

40

50

0

Low 100

Medium

High

Very high

200

300 400 Deformation at failure (mm)

Very significant damage to surface support 500

600

700

FIGURE 7.40 Design of rock reinforcement under dynamic loading. (Data from Player, J.R., Dynamic testing of rock reinforcement systems, PhD thesis, Western Australian School of Mines, Curtin University of Technology, Kalgoorlie, Western Australia, Australia, 2012, 501pp.)

2.4 m 550 MPa 20 mm threaded bar—T20 2.4 m 550 MPa 20 mm threaded bar—T20—no plate 2.4 m 550 MPa 20 mm threaded bar—T20 2.4 m 550 MPa 20 mm threaded bar—Secura T20—resin 2.4 m 550 MPa 23 mm threaded bar—Secura R27—resin 2.4 m 550 MPa 25 mm threaded bar—JTECH—resin-SE 3.0 m 550 MPa 20 mm threaded bar—T20—1.6 m centrally decoupled mine nut 3.0 m 550 MPa 20 mm threaded bar—T20—1.6 m centrally decoupled integrated nut/washer 2.4 m 550 MPa 20 mm threaded bar—T20—1.0 m centrally decoupled Posimix bolt—resin 3.0 m 280 MPa 22 mm threaded bar—Saferock—four buffer 3.0 m 280 MPa 22 mm threaded bar—Saferock—two buffer 2.2 m 280 MPa 22 mm threaded bar—Saferock—HC (weak grout) 2.4 m 580 MPa 22 mm Garford solid yielding bolt version 1 2.4 m 580 MPa 22 mm Garford solid yielding bolt version 2 2.4 m 580 MPa 22 mm Garford solid yielding bolt version 2—resin 2.4 m 580 MPa 22 mm Garford solid yielding bolt version 2—resin 2.4 m 400 MPa 22 mm cone bolt >40 MPa grout 2.4 m 400 MPa 22 mm cone bolt >40 MPa LE grout 2.4 m 400 MPa 22 mm cone bolt >40 MPa HE grout 2.4 m 400 MPa 22 mm cone bolt 25 MPa grout 3.0 m Roofex 12.5 mm—cement grout 3.0 m 450 MPa D-Bolt 22 mm—cement grout 3.0 m Yield-Lok 17.2 mm—775 mm yield length—cement grout 2.6 m Cable bolt-A 15.2 mm –plain strand—2.0 m toe anchor rupture 2.6 m Cable bolt-A 15.2 mm –plain strand—1.5 m toe anchor toe slid 2.6 m Cable bolt-A 15.2 mm—plain strand—0.6 m collar slid 3.4 m Cable bolt-A 15.2 mm—plain strand—1.7 m centrally debonded 3.4 m Garford yielding cable bolt - Version 2 3.0 m Cable bolt-C 15.2 mm—plain strand—two buffer LC 3.0 m Cable bolt-C 15.2 mm—plain strand—four buffer LC 3.0 m Cable bolt-C 15.2 mm—plain strand—damaged wire 2.4 m 47 mm split tube bolt—1.8 m average toe anchor 2.2 m Inflatable bolt—1.5 m average toe anchor

Failure by rupture High-impact testing

ion

Reinforcement types

R

c for Re in

s de nt em e

eg ign r

m oc k

d an em ass d

60

Rock Reinforcement and Support 361

362

Geotechnical Design for Sublevel Open Stoping

FIGURE 7.41 Example of extremely high rock mass demand, where reinforcement failure was followed by mesh loading, rock mass bulking, and load transfer to other bolts.

drives can be developed around a stoping block solely for cable bolt installation. To decrease cost and increase the reinforcing effectiveness, such horizontal drives are usually located at the same vertical horizon as the drilling sublevels and 10–15 m away from a planned stope wall location (Figure 7.42). However, special cable bolting drives are not normally used in most open stoping operations.

6 m long bulbed cable bolts

22B1 S Panel CMS section

FIGURE 7.42 Cable bolt reinforcement and resulting stope crown. (From Villaescusa, E. et al., An integrated approach to the extraction of the Rio Grande Silver/Lead/Zinc orebodies at Mount Isa, in Singhal et al., ed., Proceedings of the Fourth International Symposium on Mine Planning & Equipment Selection, Balkema, Calgary, Alberta, Canada, October 31–November 3, 1995, pp. 277–283.

Rock Reinforcement and Support

363

7.5.1  Cable Bolt Reinforcement Mechanisms The cable bolt reinforcement system is made up of four components (Windsor, 2004): • Rock mass • Element (strands) • Internal fixture (cement grout) • External fixture (plate and barrel and wedge anchor) Stope wall responses can be measured during stoping to develop a better understanding of cable bolt/rock mass interaction (Bywater and Fuller, 1983; Greenelsh, 1985; Hutchinson and Diederichs, 1996). Additionally, assessment of cable bolt reinforcement effectiveness can be based on visual interpretation of stope wall photographs (see Figure 1.13) and the survey of the resulting stope voids (see Chapter 9). Oddie and Pascoe (2005) have reported results for stope crowns at the Olympic Dam mine, where significant reductions in the resulting depth of failure were achieved with the use of cable bolting (Figure 7.43). The main mechanisms that apply during cable bolt reinforcement are as follows: • Application of compression to improve resistance against shear and tension across preexisting geological discontinuities. • Creation of a composite beam of several layers of strata (when the cables are installed in bedded rock). The stability can be improved if individual bands can be grouped together to form a much stronger composite beam. Cable bolting can be used to minimize bedding slip along strike and dip adjacent to the stope walls. • Anchoring unstable zones to stable or solid ground, while providing large retention capabilities. • Minimization of large excavation deformations, arising in part from rock mass relaxation at the mid-stope spans. For open stoping, the stabilization process requires the implementation of surface support and rock bolts to create a strong membrane along the walls of the drilling drives. Cable bolt rings are spaced every 2–3 m, and rock bolts can be installed between rings. The reinforced skin is tied into better-quality rock further into the rock mass by the longer cable bolts (Figure 7.44). The reinforcement length is typically taken as the depth of unstable rock around a stope plus a specified length for anchorage. In practice, the length of a typical cable bolt length for stope wall reinforcement ranges from 6 to 10 m. Cable bolt spacing is designed to provide a static capacity equal to the dead

0

10

20

30

40

50

60

70

80

0

10

30 40 20 Stope width (m)

50

HR = 4

HR = 6

HR = 8

HR = 10

HR = 12

HR = 14

HR = 16 Maximum depth of failure (m) 0–5 5–10 10–15 >15

0

10

20

30

40

50

60

70

80

90

100

0

10

20 30 40 Stope width (m)

Cable bolted

50

HR = 4

HR = 6

HR = 8

HR = 10

HR = 12

HR = 14

HR = 16

FIGURE 7.43 Cable bolt reinforcement and stope crown performance at Olympic Dam mine. (From Oddie, M.E. and Pascoe, M.J., Stope performance at Olympic Dam Mine, Proceedings of the Ninth Underground Operators’ Conference, Perth, Western Australia, Australia, March 7–9, 2005, pp. 265–272, The AusIMM, Melbourne, Victoria, Australia. With permission.)

Stope length (m)

90

Unreinforced

Stope length (m)

100

364 Geotechnical Design for Sublevel Open Stoping

365

Rock Reinforcement and Support

Anchorage zone Cab

Unsupported

Rock reinforced zone

le bo

lt

Cable bolts Mesh

Rock

bolt

Initial Final shotcrete shotcrete Detailed view layer layer

FIGURE 7.44 Deep cable bolt anchorage of stope walls.

weight of the failed material. For twin strand cable bolts, the spacing is typically 1.5–2 m within each ring. The underlying design philosophy is to increase the density of cable bolt reinforcement within the exposed stope walls (Figure 7.45) in an attempt to stabilize a surface band along the walls of the drill drives (Rauert, 1995). The overall result is to minimize the deformation of the final exposed stope walls.

Zone of intense cable bolting

Unsupported span

FIGURE 7.45 Zone of intense cable bolting at a stope drill drive. (Courtesy of Mount Isa Mines, Mount Isa, Queensland, Australia.)

366

Geotechnical Design for Sublevel Open Stoping

7.5.2  Cable Bolt Types The cable bolts utilized in sublevel open stoping consists of a seven-wire, stress-relieved, high-tensile steel strand with plain (round) wires. Six wires are laid helically around a slightly larger diameter central (king) wire. The regular 15.2 mm diameter strand can be produced to provide a number of grades that provide differing yield and ultimate load capacities. Standard single strands have a minimum yield force capacity of 213 kN and a minimum breaking force of 250 kN. Single or twin strand cables may be used for stope and bench hangingwall reinforcement, while twin strand cables are used for permanent back reinforcement. Figure 7.46 shows some of the typical cable bolting geometries used in the mining industry (Windsor and Thompson, 1993). 7.5.2.1  Plain Strand Cable Bolts Plain strand cable bolts may or may not have a high rate of load transfer (measured in terms of force per unit embedment length). This will depend on the cleanliness of the strand prior to grouting and the quality of the cement grout (Villaescusa et al., 1992, Figure 7.47). These cable bolts may also suffer from a significant reduction in the rate of load transfer if the borehole confinement (stress) reduces (Hyett et al., 1995). Consequently, plain cables installed in areas where the rock mass deteriorates due to the mining process may fail by slippage without developing any significant loads before failure. However, plain cables are very effective in supporting stope walls (Bywater and Fuller, 1983; Villaescusa et al., 1992). 7.5.2.2  Modified Strand Cable Bolts Two types of strands that have been modified to cause a variation in cross section along their length are known as birdcaged and bulbed strands. Longitudinal section Single plain strand

Cross section Twin birdcaged

Twin plain strand and spacers Birdcaged—7 wire

Twin bulbed

Bulbed FIGURE 7.46 Typical cable bolting geometries. (From Windsor, C.R. and Thompson, A.G., Rock reinforcement—Technology, testing, design and evaluation, in J.A. Hudson et al., eds., Comprehensive Rock Engineering, vol. 4, 1993, pp. 451–484, Oxford, U.K., Pergamon.)

367

Rock Reinforcement and Support

300

250

Load (kN)

200

150 Double (1 m) embedment test grout w/c ratio 0.30 0.35 0.40 0.45 0.50 0.55

100

50

0

0

5

10

15

20 25 30 35 Displacement (mm)

40

45

50

55

FIGURE 7.47 Influence of cement grout on the load transfer of single strand plain cables.

Both types of strands result in more effective load transfer between the strand and the cement grout. The more effective load transfer is reflected by the need for a shorter embedment length, in which to transfer the strand capacity, and higher values for the force–displacement response stiffness. Figure 7.48 shows a schematic example of a twin bulbed cable geometry used for development excavation reinforcement in which the installed bulb density is 4/m. This bulb density provides a stiff reinforcement likely to minimize the movement of the reinforced blocks in the back of an excavation (Figure 7.49). The optimal bulb diameter ranges from 29 to 31 mm, thereby facilitating the use of thick cement grouts that can effectively penetrate the bulbs.

Bulb diameter: 29–31 mm, overall cable length: 6 m ± 5 mm, tail length 0.5 m ± 5 mm,

0.5 m Tail to be plated

Bulb density 4 bulbs/m

FIGURE 7.48 Schematic of twin strand bulbed cable used for back reinforcement in hard rock.

368

Geotechnical Design for Sublevel Open Stoping

600

500

Load (kN)

400

300 Double (1 m) embedment test 0.45 grout w/c ratio

200

2 b/m + plain 2 b/m + 1 b/m

100

0

2 b/m + 2 b/m

10

0

20 30 Displacement (mm)

40

50

FIGURE 7.49 Laboratory performance of twin bulbed strand.

7.5.2.3  Debonded Plain Strand Cable Bolts A debonded plain strand cable bolt requires 0.6–1.5 m of bulbed strand at the toe of the hole to establish an acceptable anchorage capacity (Figure 7.50). The response of the anchor will be relatively stiff. However, the overall response will not be stiff due to the extension of the free length between the anchor and the collar. Therefore, to provide stiff, near-surface restraint

Fixed anchor length

Borehole

Decoupled length

Cement grout

FIGURE 7.50 Debonded plain strand cable.

De-coupling sleeve

Coupled length

Strand

Barrel and wedge anchor

Steel plate

Rock Reinforcement and Support

369

to minimize rock mass loosening will require the installation of additional stiff rock bolts. One possible advantage of decoupled strand is that it can cope better with shear displacement across the axis of the borehole than fully coupled strand and solid bar. 7.5.2.4  Cable Bolt Plates A plain strand cable bolt generally requires a plate to be effective in retaining rock. Plates are required when it is not possible to ensure sufficient load transfer near the excavations, especially when large-scale structures are present (Figure 7.51). Bulbed strand should also be plated where possible, but is more likely to be effective where it is not possible to get access to the strand (i.e., cable bolts installed in stope hangingwalls prior to stoping). The use of barrel and wedge anchors to restrain plates, straps, and mesh in cable bolt reinforcing applications commenced in the early 1980s in Australian mines (Thompson, 2004). Recent developments in cable bolt design have meant an increased reliance on anchors being serviceable for long periods of time, especially for applications where the strand is decoupled from the cement grout. A barrel and wedge anchor is essential when a length of strand is decoupled at the collar. In twin decoupled strand cable bolts, it is necessary to have barrel and wedge anchors on both strands to achieve the full capacity of the system (Figure 7.52). It is also necessary to have a clear understanding and appropriate procedures to ensure that anchors are installed correctly and perform according to specifications (Thompson, 2004; Hassell et al., 2006).

FIGURE 7.51 Plain strand cables installed with no plates unable to retain unstable blocks in a stope crown.

370

Geotechnical Design for Sublevel Open Stoping

FIGURE 7.52 Barrel and wedge anchor on both cable bolt strands.

7.6  Cable Bolt Corrosion Corrosion is one of the major factors determining which reinforcement type can be used as the permanent support. Corrosion reduces the capacity and life expectancy of ground support, creating a number of safety concerns and operational difficulties in underground mining (Villaescusa et al., 2008, Figure 7.53). Furthermore, corrosion has been found to be partly responsible for 29% of all rock bolt failures and 25% of all cable bolt failures during rock falls within the Australian mining industry (Potvin et al., 2001). 7.6.1  Corrosivity of Cable Bolt Strands Cement-grouted cable bolts are capable of high load transfer capacity and resistance to corrosion damage. This resistance is provided by the protective alkaline environment of the cement grout and the physical barrier it provides from the surrounding environment. However, experience has shown that corrosion begins once the cement grout barrier is removed. This occurs by cracking of the grout column due to ground movement, blast damage, or in sections where the element is exposed from inadequate installation.

Rock Reinforcement and Support

371

FIGURE 7.53 Severely corroded cement-grouted cable bolt.

In an effort to better understand the response of cement-grouted strand to corrosion attack following cracking of the grout column and infiltration of groundwater, a number of laboratory experiments, including data collection at eight Australian mines, have been reported by Hassell (2007). The research concluded that at least a 2 mm crack width is needed before significant corrosion occurs. Variables such as pH, temperature, total dissolved solids (TDS), dissolved oxygen, flow rate, and time were analyzed. A very good direct linear relationship was found between the dissolved oxygen and the measured corrosion rates (Figure 7.54). Dissolved oxygen content was found to be directly related to the temperature and the salinity of the water. Thus, with one parameter, three controlling variables can be taken into account (Hassell, 2007). The good correlation between TDS and corrosion rate is partly due to the temperatures being similar and having a comparable effect on the corrosion rate. In general, a reduction in the rate of corrosion over time was observed. This is due to the corrosion products partly inhibiting further corrosion. This rate becomes constant after a certain period of time, depending on the environmental conditions. The rate of groundwater flow affects the corrosion rate by two processes. Increases in the flow rate simultaneously increase the rate at which dissolved oxygen is brought in contact with the steel surface. This provides more available oxygen for the electrochemical process, and thus higher rates of corrosion occur. Higher flow rates also increase the level of physical erosion of the corrosion products and reduce the thickness of the partially protective cover increasing the corrosion rate.

372

Geotechnical Design for Sublevel Open Stoping

Corrosion rate (mm/year)

1.4

Corrosion environment CR = 0.5528 (DO) – 0.9267 Mine D R2 = 0.988 Mine G 1.2 Mine C Mine F Mine H Mine A 1.0 0.8 0.6 0.4 0.2 0.0

0

1

2 3 Dissolved oxygen (mg/L)

4

5

FIGURE 7.54 Dissolved oxygen versus corrosion rates for a number of Australian mines. (From Hassell, R.C., Corrosion of rock reinforcement in underground excavations, PhD thesis, Western Australian School of Mines, Curtin University of Technology, Kalgoorlie, Western Australia, Australia, 2007, 277pp.)

Table 7.4 shows the corrosivity classification for groundwater-affected hard rock conditions found in Australian underground mines as proposed by Hassell (2007). The classification considers two factors in determining the corrosivity of the groundwater: dissolved oxygen content as measured in situ from a dissolved oxygen probe and the groundwater flow conditions as illustrated in Figure 7.55. Uniform corrosion rates for HA300 grade steel can then be estimated for different environments. The classification provides a range of possible corrosion rates for a specific dissolved oxygen content and groundwater flow. As the groundwater condition is obtained from qualitative observation rather than quantitative assessment, this variation in values is necessary. Projection of the corrosion rates for measurements of dissolved oxygen less than 1.5 and greater than 4.5 is uncertain due to insufficient data. The given corrosion rates are for uniform corrosion only. However, it is appropriate to assume that pitting corrosion will increase with higher rates of uniform corrosion. The classification does not take into account the rock mass quality. It is assumed that if the classification is to be applicable, the reinforcement will intersect water-bearing discontinuities. In addition, the rock mass damage from stress redistribution is

373

Rock Reinforcement and Support

TABLE 7.4 Maximum Corrosion Rates for HA300 Steel in GroundwaterAffected Australian Hard Rock Mining Environments Strong flow—large continuous water flow from a large fault or many fractures Dissolved oxygen (mg/L) 1–2 2–3 3–4 4—5 Corrosion rate (mm/year)