Guidelines for Reducing the Time and Cost of TurbineGenerator Maintenance Overhauls and Inspections Volume 1: General Practices R I A L
LICE
N
SED
M AT E
Place Image Here
Effective December 6, 2006, this report has been made publicly available in accordance with Section 734.3(b)(3) and published in accordance with Section 734.7 of the U.S. Export Administration Regulations. As a result of this publication, this report is subject to only copyright protection and does not require any license agreement from EPRI. This notice supersedes the export control restrictions and any proprietary licensed material notices embedded in the document prior to publication.
Guidelines for Reducing the Time and Cost of Turbine-Generator Maintenance Overhauls and Inspections Volume 1: General Practices 1014730 Final Report, March 2007
EPRI Project Manager A. Grunsky
ELECTRIC POWER RESEARCH INSTITUTE 3420 Hillview Avenue, Palo Alto, California 94304-1338 ▪ PO Box 10412, Palo Alto, California 94303-0813 ▪ USA 800.313.3774 ▪ 650.855.2121 ▪
[email protected] ▪ www.epri.com
DISCLAIMER OF WARRANTIES AND LIMITATION OF LIABILITIES THIS DOCUMENT WAS PREPARED BY THE ORGANIZATION(S) NAMED BELOW AS AN ACCOUNT OF WORK SPONSORED OR COSPONSORED BY THE ELECTRIC POWER RESEARCH INSTITUTE, INC. (EPRI). NEITHER EPRI, ANY MEMBER OF EPRI, ANY COSPONSOR, THE ORGANIZATION(S) BELOW, NOR ANY PERSON ACTING ON BEHALF OF ANY OF THEM: (A) MAKES ANY WARRANTY OR REPRESENTATION WHATSOEVER, EXPRESS OR IMPLIED, (I) WITH RESPECT TO THE USE OF ANY INFORMATION, APPARATUS, METHOD, PROCESS, OR SIMILAR ITEM DISCLOSED IN THIS DOCUMENT, INCLUDING MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE, OR (II) THAT SUCH USE DOES NOT INFRINGE ON OR INTERFERE WITH PRIVATELY OWNED RIGHTS, INCLUDING ANY PARTY'S INTELLECTUAL PROPERTY, OR (III) THAT THIS DOCUMENT IS SUITABLE TO ANY PARTICULAR USER'S CIRCUMSTANCE; OR (B) ASSUMES RESPONSIBILITY FOR ANY DAMAGES OR OTHER LIABILITY WHATSOEVER (INCLUDING ANY CONSEQUENTIAL DAMAGES, EVEN IF EPRI OR ANY EPRI REPRESENTATIVE HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH DAMAGES) RESULTING FROM YOUR SELECTION OR USE OF THIS DOCUMENT OR ANY INFORMATION, APPARATUS, METHOD, PROCESS, OR SIMILAR ITEM DISCLOSED IN THIS DOCUMENT. ORGANIZATION(S) THAT PREPARED THIS DOCUMENT Electric Power Research Institute (EPRI)
NOTE For further information about EPRI, call the EPRI Customer Assistance Center at 800.313.3774 or e-mail
[email protected]. Electric Power Research Institute, EPRI, and TOGETHER…SHAPING THE FUTURE OF ELECTRICITY are registered service marks of the Electric Power Research Institute, Inc. Copyright © 2007 Electric Power Research Institute, Inc. All rights reserved.
CITATIONS This report was prepared by: Electric Power Research Institute (EPRI) 1300 W. T. Harris Blvd. Charlotte, NC 28262 This report describes research sponsored by EPRI. The report is a corporate document that should be cited in the literature in the following manner: Guidelines for Reducing the Time and Cost of Turbine-Generator Maintenance Overhauls and Inspections, Volume 1: General Practices. EPRI, Palo Alto, CA: 2007. 1014730.
iii
REPORT SUMMARY
Up to 70% of the outages planned for conventional steam power plants involve work on the turbine. The challenge for the engineer is to improve performance and extend reliability, while eliminating unproductive activities from the maintenance outage schedule. This report provides general guidelines for planning and performing maintenance on steam turbines during outages. Background As a focus of innovative approaches and techniques, maintenance of aging steam turbines has assumed increased importance. In 2003, coal-fired steam plants were an average of 32 years old, and oil- or gas-fueled plants were an average of 36 years old. Many old steam plants, particularly those that are coal fired and well maintained, can be positioned to succeed in the current deregulated environment. To support this goal, EPRI is developing a series of engineering guidelines, repair procedures, and support technologies. This report is part of that effort. It contains guidelines to assist the turbine engineer in reducing the time and cost associated with maintenance overhauls and inspections of turbine-generator systems. Planning and management practices are described that are common to both nuclear and fossil units. Objective • To provide general guidelines for planning and performing a steam turbine maintenance outage Approach Under the direction of a Technical Advisory Group, the project team prepared a comprehensive guideline and series of procedures to address the sequence of activities involved with planning and performing a steam turbine maintenance outage. This information is available to members in a four-CD set, to which information is added annually. Results This first volume of Guidelines for Reducing the Time and Cost of Turbine-Generator Maintenance Overhauls and Inspections presents general practices for each of the fundamental maintenance activities usually associated with an outage: •
Operational turbine-generator condition assessment
•
Pre-outage planning and bidding
•
Unit shutdown procedures
•
Foreign material exclusion
•
Disassembly and recording clearances v
•
Turbine-generator condition assessment
•
Oil flushing
•
Rotor alignment and balancing
•
Pre-startup checks
•
Post-outage activities
Volume 2 of this report provides a series of detailed repair, replacement, and inspection procedures to guide the pre-bid, inspection, disassembly, and repair of critical turbine-generator components. Volume 3 provides balancing and alignment procedures for turbines, generators, and exciters as well as an alignment primer and a balancing primer. Volume 4 provides turbine blade/bucket, HP, IP, LP, and generator rotor and stator procurement specifications; generator rotor and stator rewind specification; a turbine-generator major overhaul procurement specification; and a turbine insulation specification. Volume 5 consists of a domestic and an international turbine-generator engineering database containing unit-specific information. Volumes 6 and 7 provide blade/disk design audit and inspection procedures for HP, IP, and LP blades/disks. Disk 4 of this set contains TGAlign V2.1 in both English units and SI units and their program user manuals. TGAlign is an automated tool for turbine-generator bearing alignment. EPRI Perspective This guideline represents a significant collection of technical information related to maintenance practices associated with an outage. The information in this report, collected by the project’s Technical Advisory Group which is made up of utility members, provides an important compilation of information and procedures to be used by maintenance staff while they prepare for and complete turbine-generator outages. As the current engineering and craft work force continues to age and retire, taking their experience and knowledge with them, this document will be of assistance in transferring the skill and knowledge of the current staff to new employees. Keywords Steam turbines Maintenance Outages
vi
ABSTRACT Maintenance of aging steam turbines has assumed increased importance as a focus for reducing costs associated with scheduled overhauls and inspections. Under the direction of a Technical Advisory Group, EPRI has prepared a comprehensive guideline and series of procedures to address the sequence of activities that are involved with the planning and performance of a maintenance outage. Volume 1 consists of a comprehensive guide for operational turbinegenerator condition assessment and general practices for each of the fundamental maintenance activities generally associated with an outage.
vii
ACKNOWLEDGMENTS In 2004, the EPRI Guidelines for Reducing the Time and Cost of Turbine-Generator Maintenance Overhauls and Inspections, Volume 1: General Practices was produced by Generation Program 65.0 (Steam Turbine-Generator and Balance of Plant), Nuclear Program Nuclear Steam Turbine-Generator Initiative (NSTI), and Technical Advisory Group (TAG) members. In 2005, sections 5.17.4.1–5.17.4.3 were contributed by Geoff Klempner and Isidor Kerszenbaum, Operation and Maintenance of Large Turbo-Generators. © 2004 The Institute of Electical and Electronics Engineers, Inc. The TAG members who assisted in the production of this report are: Name
Utility
Tom Alley
Duke Power Company
Bob Bjune
South Texas Project Electric Generating Station
Randy Bunt
Southern Nuclear
Mitch Burress
Tennessee Valley Authority
Greg Carlin
Nova Scotia Power
Russell Chetwynd
Southern California Edison
John Cizek
Nebraska Public Power District
David Crawley
Southern Company
Rick Dayton
Progress Energy
Chris Essex
Detroit Edison
Bob Garver
First Energy
Tom Kordick
Ameren
Bill McGinnis
Reliant Energy
Scott McQueen
Reliant Energy ix
Chuck Mendenhall
Salt River Project
Tom Murray
Salt River Project
Don Osborne
Duke Power Company
Ken Palmer
Pacific Gas and Electric Company
Ralph Pederson
Nuclear Management Co.
Tim Scholl
Tennessee Valley Authority
Philip Schuchter
First Energy
Dave Sharbaugh
First Energy
Ken Tillich
Northern Indiana Public Service Co.
Generation Program 65.0, NSTI, and the TAG were supported in their efforts to develop this guide by: Turbine Technology International, Inc. 2024 W. Henrietta Road Rochester, NY 14623 Principal Investigators R. Dewey M. Pollard Sequoia Consulting Group, Inc. 9042 Legends Lake Lane Knoxville, TN 37922 Principal Investigator M. Tulay
x
CONTENTS
1 TURBINE-GENERATOR CONDITION ASSESSMENT – IN SERVICE ............................... 1-1 1.1
Overview ................................................................................................................... 1-1
1.2
Procedure and Objectives ......................................................................................... 1-5
1.3
Planning a Steam Turbine-Generator Condition Assessment.................................... 1-6
1.4
Documentation of Condition Assessment .................................................................. 1-7
1.5
Condition Assessment Procedure ............................................................................. 1-7
1.5.1
Turbine-Generator History, Upgrades, and Major Forced Outage Events ......... 1-9
1.5.2
Turbine Vibration .............................................................................................1-14
1.5.3
Bearing Metal and Oil Temperatures ...............................................................1-18
1.5.4
Thermal Performance ......................................................................................1-19
1.5.5
Unit Start and Load Data .................................................................................1-25
1.5.6
Unit System Steam/Water Purity .....................................................................1-26
1.5.7
Lubricating Oil and EHC Fluid Testing .............................................................1-27
1.5.8
Pump Testing ..................................................................................................1-30
1.5.9
Turbine Steam Valve Test Results...................................................................1-31
1.5.10
Overspeed and Trip Checks.........................................................................1-34
1.5.11
Instrument Surveys ......................................................................................1-36
1.5.12
Generator Electrical Operating Data.............................................................1-38
1.5.13
Auxiliary Systems Data.................................................................................1-43
1.5.14
Component Visual Inspections .....................................................................1-46
1.5.15
Out-of-Limit Conditions and Upsets..............................................................1-47
1.5.16
Review/Update Turbine Generator Maintenance Plans ................................1-48
1.6
Evaluating Situations and Making Recommendations ..............................................1-48
1.7
Condition Assessment Example...............................................................................1-48
1.8
Summary Remarks ..................................................................................................1-52
xi
2 PRE-OUTAGE PLANNING AND BIDDING ......................................................................... 2-1 2.1
Identifying and Establishing Engineering Responsibilities ......................................... 2-3
2.1.1
Engineering Responsibilities in a Major Outage Work Scope............................ 2-4
2.1.1.1
Pre-Outage Activities.................................................................................. 2-4
2.1.1.2
Outage Activities........................................................................................2-11
2.1.1.3
Post-Outage Activities ...............................................................................2-14
Post-Outage Meeting...........................................................................................2-15 2.1.2
Methods to Estimate Engineering Resources and Work Force Required.........2-16
2.1.3
Tasks Deferred to Reduce the Scope and the Potential Implications ...............2-17
2.1.4
Tools Available and Input Needed to Define Tasks for a Scope of Work .........2-17
2.2
Pre-Bidding and Procuring Parts or Services (When Scope Is Defined) ...................2-18
2.2.1
Stationary Repairs - Diaphragms, Packing Rings, and Sealing Strips..............2-19
2.2.2
Blade/Bucket Replacement or Repairs ............................................................2-28
2.2.3
Bearing and Shaft Seal Repairs.......................................................................2-31
2.2.4
Generator Repairs ...........................................................................................2-33
2.2.5
Valve Part Replacement and Repair................................................................2-36
2.2.6
Parts Stores Review ........................................................................................2-37
2.2.7
Miscellaneous Turbine-Generator Exciter Parts, Bolts, Nuts, and Other Parts................................................................................................................2-41
2.3
Identifying and Procuring Specialized Support .........................................................2-41
2.3.1
Lead Times to Arrange for Different Types of Support. ....................................2-42
2.3.2
Web Searches: Key Words or Identifiers to Produce Supplier Lists .................2-44
2.4
Scaffolding Requirements ........................................................................................2-44
2.4.1
Customization of a Scaffolding Plan ................................................................2-46
2.4.2
Ways to Reduce Scaffolding Erection Time .....................................................2-47
2.5
Safety Procedures....................................................................................................2-47
2.5.1 2.6
Environmental Planning ...........................................................................................2-50
2.6.1
EHC Fluid ........................................................................................................2-51
2.6.2
Waste Products to Be Considered ...................................................................2-53
2.7
xii
Plan for Insulation/Asbestos Identification, Handling, and Disposal .................2-48
Crane Availability .....................................................................................................2-53
2.7.1
Crane Maintenance to Be Performed in Advance ............................................2-54
2.7.2
Types of Cranes ..............................................................................................2-54
2.7.3
Crane Use Schedule .......................................................................................2-55
2.8
Turbine Deck Lay-Down Planning ............................................................................2-55
2.8.1
Basic Elements for Any Deck Lay-Down Plan..................................................2-56
2.8.2
Basic Items or Issues to Be Reviewed.............................................................2-58
2.8.3
Items to Amend in a Customized Plan .............................................................2-59
2.9
Special Tools, Equipment, and Facilities ..................................................................2-59
2.9.1
Storage and Work Space Provisions for Cleaning and/or NDE ........................2-61
2.9.2
Provisions for Cleaning and Inspecting Different Turbine Parts .......................2-61
2.9.3
Items or Issues Specified as Part of the Work Order for Vendors ....................2-63
2.10
Machine Disassembly Plan..................................................................................2-64
2.10.1
Basic Elements in the Machine Disassembly Plan........................................2-64
2.10.2
Issues or Items Reviewed ............................................................................2-64
2.10.3
Identifying Contingency Plans for Unexpected Work ....................................2-66
2.11
Foreign Material Exclusion ..................................................................................2-66
2.11.1
Organizational Responsibilities for Turbine-Generator Contracts .................2-66
2.11.2
Areas of the Turbine-Generator to Protect ...................................................2-70
2.11.3
Measures to Take for Each Critical Area ......................................................2-71
2.11.4
Implementation of FME Plans for Turbine-Generator Work ..........................2-73
2.11.5
Performance of Work Inside the Turbine-Generator FMEA ..........................2-74
2.11.6
Retrieval of Foreign Objects.........................................................................2-79
2.11.7
Video Inspection of Shells and Steam Lines ................................................2-79
2.12
Training ...............................................................................................................2-80
2.12.1
Training Formats..........................................................................................2-80
2.12.2
Recommended Training Topics ...................................................................2-81
2.12.3
Training Options...........................................................................................2-83
2.13
Rigging, Special Tools, Parts, and Expendable Materials....................................2-83
3 UNIT SHUTDOWN............................................................................................................... 3-1 3.1
Pre-Outage Testing................................................................................................... 3-1
3.2
Generic Steps for Shutdown...................................................................................... 3-3
3.3
Critical Engineering Concerns ................................................................................... 3-4
3.4
Parameters to Monitor............................................................................................... 3-5
3.5
Opportunities to Reduce Shutdown Time .................................................................. 3-7
3.6
Practices That Have Been Used to Reduce Shutdown Time..................................... 3-7
3.6.1
Overspeed Trip Testing .................................................................................... 3-9
3.6.2
Electrical Trips vs. Mechanical Trips................................................................. 3-9
xiii
3.6.3
Boiler/Reactor Feed Pump Turbine Controls ...................................................3-10
3.7
Removal of Covers and Crossover Piping ................................................................3-10
3.8
Valve Disassembly...................................................................................................3-11
3.9
Practices to Accelerate Cooling................................................................................3-12
3.10 Operations Performed During Turning Gear Operations...........................................3-12 3.11 Lubrication Oil Blanking............................................................................................3-15 3.12 Removal of Insulation...............................................................................................3-16 3.13 Lagging Removal .....................................................................................................3-16 4 DISASSEMBLY AND RECORDING CLEARANCES .......................................................... 4-1 4.1
Planning Lay-Down Areas......................................................................................... 4-1
4.1.1
Material Handling Methods and Considerations................................................ 4-3
4.1.2
Component Disassembly Requirements........................................................... 4-3
4.1.3
Component Work Scopes and Work Centers ................................................... 4-3
4.1.4
Component Weights and Floor Loading............................................................ 4-7
4.1.5
Tooling/Support Locations ................................................................................ 4-9
4.1.6
Power/Air/Water Requirements .......................................................................4-10
4.1.7
Personnel Needs (Restrooms, Eating Facilities) ..............................................4-10
4.2
Features of the Basic Rigging Plan ..........................................................................4-11
4.2.1
Rigging/Lifting Drawings for Major Components ..............................................4-11
4.2.2
Rigging Devices, Lifting Bars, Wire Rope, Synthetic Slings, and Shackles ......4-14
4.2.3
Practical Methods for Efficient Handling of Certain Components .....................4-16
4.2.4
Special Turbine Tools ......................................................................................4-17
4.3
Scheduling Overhead Crane Time ...........................................................................4-18
4.4
Moving Without the Overhead Crane .......................................................................4-18
4.5
Special Storage Considerations ...............................................................................4-18
4.5.1
Racks for Diaphragms .....................................................................................4-19
4.5.2
Valve Stands, Rotor Stands, Mandrels, Try Bars, and Stub Shafts ..................4-21
4.5.3
Shell Racks, Supports, and Cribbing ...............................................................4-23
4.6
xiv
Bolt Removal Practices and Techniques ..................................................................4-25
4.6.1
Identifying the Necessary Personnel for Unbolting the Turbine-Generator.......4-27
4.6.2
Available Tools Needed...................................................................................4-28
4.6.3
Useful Tools and Techniques for Different Applications ...................................4-30
4.6.4
Bolt Removal Sequence ..................................................................................4-30
4.6.5
Required Inventory of Bolts .............................................................................4-33
4.7
Taking Axial and Radial Clearances and Their Use..................................................4-34
4.8
Required Rotor Radial Position and Coupling Alignment Checks .............................4-38
4.9
Checks to Assess Spare Rotor Compatibility ...........................................................4-40
5 TURBINE-GENERATOR CONDITION ASSESSMENT ....................................................... 5-1 5.1
Cleaning Without Disassembly.................................................................................. 5-2
5.2
Recommended Inspection and Testing Techniques .................................................. 5-3
5.2.1
Proof Test......................................................................................................... 5-8
5.2.2
Megger Test ..................................................................................................... 5-9
5.2.3
Doble Test .......................................................................................................5-10
5.2.4
Other Tests......................................................................................................5-10
5.3
In Situ Inspection......................................................................................................5-12
5.3.1
Economic Incentives Imposed by Deregulation................................................5-12
5.3.2
Machine Access ..............................................................................................5-13
5.3.3
Video Probe Systems ......................................................................................5-14
5.3.4
Utility Experiences ...........................................................................................5-15
5.4
Accelerating Different Types of Inspections..............................................................5-16
5.4.1
Defect Sizing and Implications of Results ........................................................5-17
5.5
Cleaning Coated Versus Non-Coated Parts .............................................................5-19
5.6
Coating-Removal Techniques ..................................................................................5-20
5.7
Sampling and Analyzing Deposits ............................................................................5-20
5.8
NDE of Turbine-Generators and Collecting Boresonic Data .....................................5-21
5.8.1
Turbine-Generator Nondestructive Evaluation Techniques..............................5-21
5.8.2
Collecting Boresonic Data ...............................................................................5-22
5.8.3
EPRI-Supported Rotor Boresonic Inspection ...................................................5-24
5.8.4
Boresonic System Evaluation Procedures .......................................................5-25
5.8.5
Inspection of Boreless Rotors ..........................................................................5-26
5.8.6
Inspection of Steam Turbine Disk Blade Attachments......................................5-26
5.8.7
Inspection of Nonmagnetic Generator Retaining Rings....................................5-28
5.9
Inspection of Shrunk-On Components......................................................................5-28
5.10 Bearings – Journal and Thrust Types.......................................................................5-29 5.11 Stationary Components............................................................................................5-34 5.12 Buckets/Blades ........................................................................................................5-40 5.13 Rotors ......................................................................................................................5-43 5.13.1
Causes of Rotor Bowing...............................................................................5-43
xv
5.13.1.1
Severe Rubbing.......................................................................................5-43
5.13.1.2
Bows Caused by Water Induction............................................................5-46
5.13.1.3
Bows Caused by Metallurgical Problems.................................................5-46
5.13.1.4
Corrective Actions ...................................................................................5-46
5.13.2
Other External Rotor Problems.....................................................................5-47
5.14 Shaft Seals ..............................................................................................................5-48 5.15 Valves .....................................................................................................................5-51 5.15.1
Stop Valves ..................................................................................................5-51
5.15.2
Control Valves..............................................................................................5-52
5.15.3
Reheat Stop Valves......................................................................................5-53
5.15.4
Non-Return Valves .......................................................................................5-53
5.16 Casings, Steam Chests, and Nozzle Chests............................................................5-54 5.17 Generator ................................................................................................................5-57 5.17.1
Classifications ..............................................................................................5-57
5.17.2
Generator Stator...........................................................................................5-58
5.17.3
Generator Field ............................................................................................5-62
5.17.4
Generator Electrical Testing .........................................................................5-66
5.17.4.1
Generator Stator Core Electrical Tests ....................................................5-71
5.17.4.2
Generator Stator Winding Electrical Tests ...............................................5-89
5.17.4.3
Generator Rotor Electrical Testing.........................................................5-105
5.18 Excitation System ..................................................................................................5-119 5.19 Using Data on Condition Assessment to Assess Risk of In-Service Failure ..........5-120 6 OIL FLUSHING.................................................................................................................... 6-1 6.1
Preparations and Precautions for Flushing the System ............................................. 6-1
6.2
Resources That Should Be Available While Flushing ................................................ 6-3
6.3
Precautions While Flushing....................................................................................... 6-4
6.4
Oil Cleanliness Criteria.............................................................................................. 6-5
6.5
Heating and Cooling the Oil Without Damaging the Bearing...................................... 6-8
6.6
Minimizing the Use of External Piping While Flushing ............................................... 6-9
6.7
Flushing Without an External Filter...........................................................................6-10
6.8
Techniques to Get Maximum Flow Through Piping ..................................................6-11
7 ROTOR ALIGNMENT AND BALANCING ........................................................................... 7-1 7.1
xvi
Different Tight Wire Techniques ................................................................................ 7-1
7.2
Information Collected from the Unit ........................................................................... 7-4
7.3
Automated and Semi-Automated Alignment Processes ............................................ 7-5
7.4
Slow-Speed Versus High-Speed Balancing............................................................... 7-6
7.4.1
Slow-Speed Balance Requirements/Considerations......................................... 7-8
7.5
When Spin Balancing Is Required............................................................................. 7-9
7.6
On-Line Balancing Devices ......................................................................................7-11
7.7
Potential Consequences of Not Balancing the Rotor ................................................7-12
7.8
Selecting Vibration Limits .........................................................................................7-13
7.9
Balance Limits..........................................................................................................7-14
7.10 Access to Turbine-Generator Rotors ........................................................................7-17 7.11 Turbine-Generator Balance Support.........................................................................7-18 7.12 Turbine-Generator Balance Weights ........................................................................7-20 7.12.1
Split-Weight Design Dovetail Weights ..........................................................7-20
7.12.2
Tungsten-Style Weights ...............................................................................7-22
8 PRE–STARTUP CHECKS................................................................................................... 8-1 8.1
Steps to Minimize Startup Time................................................................................. 8-1
9 POST-OUTAGE ACTIVITIES .............................................................................................. 9-1 9.1
Post-Overhaul Engineering Reports .......................................................................... 9-1
9.2
Documentation for Vendor Signoff ............................................................................ 9-7
9.3
Issues to Review for Future Planning ........................................................................ 9-8
9.4
Recommendations for Planning Future Outages......................................................9-11
9.4.1
Problem Description ........................................................................................9-11
9.4.2
Solution ...........................................................................................................9-12
9.5
Inventory Decision Making .......................................................................................9-12
9.6
Integration with Maintenance Management Systems ...............................................9-13
10 REFERENCES .................................................................................................................10-1 A CONDITION ASSESSMENT DATA SHEETS ..................................................................... A-1 B TURBINE-GENERATOR OUTAGE REPORT.....................................................................B-1 B.1
Outage Report Instructions ....................................................................................... B-1
B.2
Report Table of Contents .......................................................................................... B-3
B.3
Blank Report Format ................................................................................................. B-6
xvii
C DATA SHEETS ...................................................................................................................C-1 D FOREIGN MATERIAL EXCLUSION GUIDANCE ............................................................... D-1 D.1
Introduction and Purpose.......................................................................................... D-1
D.2
Definitions................................................................................................................. D-2
D.2.1
Glossary of Key Terms..................................................................................... D-2
D.2.2
Categorization of FME Areas ........................................................................... D-6
D.2.3
Categorization of FME Events.......................................................................... D-6
D.3
D.3.1
FME Responsibilities for All Personnel............................................................. D-8
D.3.2
Typical Individual FME Responsibilities ........................................................... D-8
D.4
Establishing and Implementing FME Program Requirements ................................. D-11
D.4.1
Introduction .................................................................................................... D-11
D.4.2
General Programmatic Guidance................................................................... D-13
D.4.3
Sources of Foreign Material Contamination ................................................... D-13
D.4.4
Defining the Scope of Equipment Controlled by FME Procedures.................. D-14
D.4.5
Training and Qualification of Individuals......................................................... D-15
D.5
Developing and Implementing FME Control Plans.................................................. D-16
D.5.1
Factors to Consider When Developing an FME Control Plan ......................... D-16
D.5.2
Typical Contents of an FME Control Plan ...................................................... D-17
D.5.3
Establishing the FMEA................................................................................... D-19
D.5.4
Determining Appropriate FME Controls for the Area ...................................... D-20
D.5.5
Establishing an FMEA Boundary.................................................................... D-21
D.5.6
Installing the FMEA Boundary........................................................................ D-21
D.5.7
Conducting Pre-Job Briefings ........................................................................ D-22
D.6
xviii
Plant/Station Responsibilities....................................................................................D-8
Performance of Work Inside the FMEA................................................................... D-22
D.6.1
FMEA Entry Requirements ............................................................................ D-22
D.6.2
Use of Control Logs ....................................................................................... D-23
D.6.3
Monitoring the FMEA ..................................................................................... D-25
D.6.4
Cleanliness and Readiness Inspections......................................................... D-26
D.6.5
Performance of Maintenance Activities Within the FMEA............................... D-26
D.6.6
Examples of Good Work Practices Inside the FMEA...................................... D-28
D.6.7
Implementing Graded FME Controls .............................................................. D-30
D.6.8
Ensuring Cleanliness Inside the FMEA .......................................................... D-31
D.6.9
Use of FME Devices ...................................................................................... D-33
D.7
Recovery of Loss of FMEA Control......................................................................... D-37
D.7.1
Initiation of a Condition Report ....................................................................... D-37
D.7.2
FME Recovery Plan ....................................................................................... D-38
D.7.3
Foreign Material Retrieval .............................................................................. D-38
D.7.4
Recovering Foreign Material After Returning the System to Service .............. D-40
D.8
Close Out of a Foreign Material Exclusion Area...................................................... D-40
D.9
References ............................................................................................................. D-41
xix
LIST OF FIGURES Figure 1-1 Example of the Form Used to Rate the Final Condition of the TurbineGenerator ........................................................................................................................ 1-3 Figure 1-2 Example of NERC-GAD List of Maintenance Outage Events for a Typical Unit.....1-13 Figure 1-3 Typical Orbits Showing Different Problems ...........................................................1-17 Figure 1-4 Location and Function of Basic Turbine Controls ..................................................1-32 Figure 1-5 Location of Typical Turbine Supervisory Instrumentation ......................................1-36 Figure 1-6 Typical Generator Capability Curve.......................................................................1-39 Figure 2-1 Example of an Outage Plan ................................................................................... 2-7 Figure 2-2 Sample of Engineering-Supported Activities .........................................................2-13 Figure 2-3 Dimensional Requirements That May Be Provided Within a Repair Procedure.....2-21 Figure 2-4 Major Repair Times per Inch of Partition Radial Height .........................................2-23 Figure 2-5 Estimate of Total Major Repair Time per Partition .................................................2-24 Figure 2-6 Estimation Tool for Minor Partition Repairs ...........................................................2-25 Figure 2-7 Total Diaphragm Repair Time Divided by the Nonproductive Time .......................2-26 Figure 2-8 Plot to Track Consumable Costs vs. Productive Labor Costs................................2-27 Figure 2-9 Change in HP Section Efficiency After Four Separate Outage Periods .................2-28 Figure 2-10 Examples of Rotor Weight and Coupling Geometry Measurements ....................2-32 Figure 2-11 Examples of Shells Used to Bridge Between the Retaining Ring and Rotor ........2-34 Figure 2-12 Part Location Information ....................................................................................2-39 Figure 2-13 Four-Level Part Location Hierarchy.....................................................................2-39 Figure 2-14 Types of Lifting Cranes .......................................................................................2-54 Figure 2-15 FME Organizational Structure for T-G Contracts.................................................2-67 Figure 3-1 Section Efficiency Change During the Course of Three Outages ........................... 3-2 Figure 3-2 Example of Access Platforms................................................................................3-11 Figure 3-3 Plot of Shutdown Activities....................................................................................3-14 Figure 3-4 Plot of Accelerated vs. Non-Accelerated Cool-Down Rates ..................................3-15 Figure 3-5 Example of a Toggle Blank ...................................................................................3-16 Figure 4-1 Example of a Work Center Lay-Out Plan ............................................................... 4-4 Figure 4-2 Lay-Out Plan with Work Center Layouts to Match the Expected Work Scope ........ 4-6 Figure 4-3 Example of Loading a Reference for a Turbine Deck ............................................. 4-9 Figure 4-4 Example of a Lifting Drawing.................................................................................4-12 Figure 4-5 Example of a Detailed Rigging and Lifting Drawing...............................................4-13
xxi
Figure 4-6 Rigging Fixtures for CRVs.....................................................................................4-14 Figure 4-7 Rigging Fixture for a Control Valve Actuator .........................................................4-15 Figure 4-8 Modification to a Lifting Beam That Allows the Turnbuckle to Remain Attached While Lifting.....................................................................................................4-15 Figure 4-9 Example of a Generator Field Support Modification ..............................................4-16 Figure 4-10 Example of a Diaphragm Transport and Storage Rack .......................................4-19 Figure 4-11 Example of an Oil Deflector Rack........................................................................4-20 Figure 4-12 Example of a Fixture Holding a Control Valve .....................................................4-21 Figure 4-13 Example of a Fabricated Rotor Stand with Rollers ..............................................4-22 Figure 4-14 Example of a Bearing Fitting Mandrel to Check Tilt Pads....................................4-22 Figure 4-15 Example of a Mandrel to Check Cylindrical or Elliptical Bearing Bores................4-23 Figure 4-16 Example of a Rack for Holding an HP Upper Shell..............................................4-24 Figure 4-17 Example of a Fabricated Shell Support ...............................................................4-24 Figure 4-18 Example of a Support for Generator Field Removal Without Cribbing .................4-25 Figure 4-19 Coupling Alignment Nomenclature ......................................................................4-39 Figure 5-1 Access Path for Video Probe Delivery Device .......................................................5-13 Figure 5-2 Examples from Remote Video Probe In Situ Inspection ........................................5-15 Figure 5-3 Elliptical Bearing Construction...............................................................................5-31 Figure 5-4 Diaphragm Construction .......................................................................................5-35 Figure 5-5 Mechanics Describing Rubbing Process ...............................................................5-44 Figure 5-6 Typical Non-Return Valve Construction ................................................................5-54 Figure 5-7 Terminal Stud Hydrogen Seal Construction ..........................................................5-64 Figure 5-8 Location Susceptible to High-Cycle Fatigue and Low-Cycle Fatigue in Certain Main Lead Designs.........................................................................................................5-65 Figure 5-9 Flux Fault Current Path .........................................................................................5-72 Figure 5-10 EL-CID Excitation Setup .....................................................................................5-73 Figure 5-11 EL-CID Analog Equipment ..................................................................................5-74 Figure 5-12 EL-CID Digital Equipment ...................................................................................5-75 Figure 5-13 EL-CID Chattock Theory .....................................................................................5-76 Figure 5-14 EL-CID MMF Theory...........................................................................................5-77 Figure 5-15 EL-CID Signal Interpretation ...............................................................................5-79 Figure 5-16 Toroid Wrap ........................................................................................................5-82 Figure 5-17 Operating Flux Pattern ........................................................................................5-82 Figure 5-18 B-H Curve Example ............................................................................................5-84 Figure 5-19 Flux Test Electrical Setup ...................................................................................5-85 Figure 5-20 Flux Test Mirror Setup.........................................................................................5-85 Figure 5-21 Infrared Hot Spot – Bruce 7.................................................................................5-86 Figure 5-22 Infrared Hot-Spot Flux Test 1 ..............................................................................5-86 Figure 5-23 Infrared Hot-Spot Flux Test 2 ..............................................................................5-87
xxii
Figure 5-24 Flux-Temperature Profiles...................................................................................5-88 Figure 5-25 IR Versus Temperature.......................................................................................5-92 Figure 5-26 Polarization Index Dryness Curve .......................................................................5-93 Figure 5-27 IR Versus Temperature – PI................................................................................5-95 Figure 5-28 DC Ramp ............................................................................................................5-97 Figure 5-29 Stator Hi-Pot Arcing ............................................................................................5-99 Figure 5-30 LKV G5 EE Hoseglow .......................................................................................5-100 Figure 5-31 PD Off-Line Capacitive Coupling.......................................................................5-101 Figure 5-32 Dissipation Factor Tip-Up..................................................................................5-104 Figure 5-33 NO Shorted Turns Traces – Superimposed ......................................................5-109 Figure 5-34 NO Shorted Turns Traces – Separated.............................................................5-109 Figure 5-35 NO Shorted Turns Traces – Summed ...............................................................5-109 Figure 5-36 RSO Single-Shorted Turn – Dual Superimposed Trace ....................................5-110 Figure 5-37 RSO Single-Shorted Turn – Difference Trace ...................................................5-110 Figure 5-38 RSO Dual-Trace – Multi-Shorts.........................................................................5-110 Figure 5-39 RSO Difference Trace – Multi-Shorts ................................................................5-111 Figure 5-40 STD by Open Circuit .........................................................................................5-112 Figure 5-41 STD by Impedance ...........................................................................................5-113 Figure 5-42 C-Core 1 ...........................................................................................................5-114 Figure 5-43 C-Core 2 ...........................................................................................................5-115 Figure 5-44 Rotor Ground – Split Voltage ............................................................................5-117 Figure 5-45 Rotor Ground – Current Through Forging .........................................................5-118 Figure 5-46 Example of a Probability Distribution and Limit State Function..........................5-121 Figure 5-47 Basic Elements of a Probabilistic Analysis ........................................................5-122 Figure 5-48 Example of SPE Inspection Criteria Using Series of Probability of Failure Curves..........................................................................................................................5-123 Figure 5-49 Ratio of Actual Crack Sizes to Measured Crack Sizes ......................................5-125 Figure 6-1 Oil Flushing Piping ................................................................................................6-10 Figure 7-1 Ten-Year Record of Rotor Bowing ......................................................................... 7-4 Figure 7-2 Exaggerated Rotor Motion for the First Three Field Critical Speeds ....................... 7-8 Figure 7-3 Low-Speed Portable Balance Machine .................................................................. 7-9 Figure 7-4 Various Standards for Residual Unbalance...........................................................7-16 Figure 7-5 Offset Modification to a Shell Bore ........................................................................7-17 Figure 7-6 Access to Balance Grooves ..................................................................................7-18 Figure 7-7 Split-Weight Dovetail Weight.................................................................................7-21 Figure D-1 Foreign Material Exclusion Flowchart .................................................................. D-12 Figure D-2 Example of an FME Plan Document.................................................................... D-18 Figure D-3 Example of an FME Boundary Sign..................................................................... D-19 Figure D-4 Example of an FME Boundary Sign..................................................................... D-20
xxiii
Figure D-5 Examples of FME Boundaries and Entry Locations ............................................. D-21 Figure D-6 Example of an Individual Entry Log ..................................................................... D-24 Figure D-7 Example of a Long-Term Placement Log............................................................. D-25 Figure D-8 Example of Lanyard Use ..................................................................................... D-36
xxiv
LIST OF TABLES Table 1-1 Breakdown of Condition Assessment ...................................................................... 1-8 Table 1-2 Critical Components Identified in Assessment Procedure ......................................1-10 Table 1-3 Cause Codes Associated with Critical Components Found in the Assessment Procedure.......................................................................................................................1-12 Table 1-4 Effect of Component Condition Changes on Fossil Cycle Performance Parameters (at Valve Wide Open Operation) .................................................................1-22 Table 1-5 Guidance for Interpreting Turbine Cycle Steam Flow and Unit Load Changes .......1-23 Table 1-6 Effect of Leakage to the Condenser on Heat Rate and Load..................................1-24 Table 1-7 Example of Overall Unit Condition Assessment .....................................................1-50 Table 2-1 Index to Turbine Outage Report: Appendices B and C............................................ 2-2 Table 2-2 Checklist of Pre-Outage Activities ........................................................................... 2-8 Table 2-3 Recommended Process Instruction Sheet and Detailed Work Package Information .....................................................................................................................2-10 Table 2-4 Recommended Information for Parts and Part Use Databases ..............................2-11 Table 2-5 Post-Outage Activities ............................................................................................2-15 Table 2-6 Recommended Parts Purchase Document Information ..........................................2-19 Table 2-7 Recommended Diaphragm Repair Purchase Document Information......................2-20 Table 2-8 Recommended Information for Bucket Replacement or Repair ..............................2-29 Table 2-9 Geometry and Tolerances Required to Support a Repair Procedure......................2-31 Table 2-10 Classification of Generator Components ..............................................................2-33 Table 2-11 Generator Tests and When They May Be Performed...........................................2-34 Table 2-12 Parts and Consumables Used to Support Routine Work ......................................2-38 Table 2-13 Examples of Specialized Sources for Locating Vendors and Supplies .................2-42 Table 2-14 Selected Activities and Estimated Lead Times .....................................................2-43 Table 2-15 Basic Elements of a Scaffolding Plan ...................................................................2-45 Table 2-16 Typical Locations Where Asbestos Is Found Around a Turbine............................2-49 Table 2-17 Items Recommended in an Asbestos Abatement Program ..................................2-50 Table 2-18 Alternative Lifting Devices for Turbine-Generator Components ............................2-55 Table 2-19 List of NDE Equipment Used to Support a Turbine-Generator Outage .................2-63 Table 2-20 Areas of a Turbine-Generator to Be Protected During Disassembly .....................2-71 Table 2-21 Areas of a Turbine-Generator That Should Be Blocked........................................2-72 Table 2-22 Example of FME Measures During Turbine Disassembly and Reassembly..........2-75
xxv
Table 2-23 List of Recommended Training Topics .................................................................2-86 Table 2-24 Typical Consumables Required for an Outage .....................................................2-88 Table 3-1 Steps Typically Involved with the Shutdown of a Turbine-Generator ....................... 3-4 Table 4-1 Checklist for Preparing a Lay-Down Plan ................................................................ 4-2 Table 4-2 Activities Required to Support a Generator Field Rewind ........................................ 4-7 Table 4-3 Tooling and Support for Turbine Deck Lay-Down Plan ...........................................4-10 Table 4-4 Tooling and Support for Turbine Deck Lay-Down Plan ...........................................4-19 Table 4-5 Comparison of Accuracy Between Different Preload Methods [14].........................4-26 Table 4-6 Tooling and Support for Turbine Deck Lay-Down Plan ...........................................4-29 Table 4-7 Outage Fastener Usage Record.............................................................................4-34 Table 4-8 Checks to Determine Compatibility Between Original and Replacement Rotors.....4-41 Table 5-1 NDE Inspection Methods Used on Different Turbine-Generator Elements .............. 5-4 Table 5-2 Visual Inspection Methods Used on Different Generator Elements ......................... 5-8 Table 5-3 Sizing and Mapping Constraints Associated with NDE...........................................5-18 Table 5-4 Cleaning Processes Used for Coated and Non-Coated Components.....................5-19 Table 5-5 Support Information Required to Interpret UT Data ................................................5-23 Table 5-6 Coupling Inspections – Disassembly and Reassembly...........................................5-29 Table 5-7 Recommended Action for Bearing Damage Typically Found at Inspection.............5-33 Table 5-8 Recommended Pre-Outage Preparations for Bearings...........................................5-34 Table 5-9 Separate Areas That Form a Stationary System ....................................................5-35 Table 5-10 Recommended Action for Diaphragm Damage Typically Found at Inspection......5-37 Table 5-11 Blade Damage Typically Found at Inspection.......................................................5-42 Table 5-12 Typical Seal Design Clearances with Field Tolerances ........................................5-49 Table 5-13 Casing Repair Issues ...........................................................................................5-56 Table 5-14 Combined Cooling Designs with Retaining Rings.................................................5-57 Table 5-15 Alternative Processes for Grinding Collector Rings ..............................................5-63 Table 5-16 Generator Electrical Tests....................................................................................5-66 Table 5-17 Capabilities of LAI vs. Conventional Inspections ..................................................5-68 Table 5-18 Summary of Advantages and Disadvantages of LAIs...........................................5-69 Table 6-1 Recommended Cleanliness Criteria ........................................................................ 6-7 Table 7-1 Specifications for an On-Line Active Balancing System..........................................7-12 Table 7-2 Sources for Equipment and Rotor Balancing Standards.........................................7-15 Table 8-1 Recommended Outline for a Startup Document ...................................................... 8-2 Table 9-1 Recommended Information to Be Collected After the Outage Is Complete.............. 9-3 Table 9-2 Examples of Post-Outage Engineering Reporting ................................................... 9-5 Table 9-3 Uses for Engineering Information Obtained in the Outage....................................... 9-8
xxvi
1
TURBINE-GENERATOR CONDITION ASSESSMENT – IN SERVICE
In order to justify the extension or reduction of time between overhauls of turbine-generator components, a systematic procedure is presented in this section to guide in the gathering and evaluation of information that, in turn, may be used to assess the turbine-generator’s condition since the last major overhaul. The assessment procedure in this section deals primarily with the information that can be obtained and evaluated while a unit is operating. The condition assessment presented later in Section 5 of this volume treats the information and processes that are recommended for when the unit is off-line and components or systems are available for detailed inspection and testing.
1.1
Overview
The purpose of this steam turbine-generator condition assessment is to: •
Track and monitor the turbine-generator’s condition since its last overhaul
•
Assemble relevant information from multiple sources (interviews and reviews of actual maintenance or operating information)
•
Offer a basis for assessing the collected results in term of recommendations for future work scope and schedule
This procedure and method of reporting is not to be confused with a system of artificial intelligence or decision-making software. This is a step-by-step process to be undertaken by an assigned team of experts and specialists. It is supported with a series of generic formats (data sheets) to guide in the collection of information for the purpose of making an informed, documented determination of when work on a specific system or component of a unit should be performed. This type of assessment is to be conducted by appropriate technical personnel and specialists acting as a group, with support provided by others within the specific plant or utility as necessary. Personnel likely to be involved in the process include: •
Those who are familiar with plant operations, such as a unit engineer
•
Those who are responsible for unit operations and conducting various tests on the equipment
•
Staff or plant personnel responsible for unit performance monitoring
•
Staff personnel responsible for unit vibration monitoring 1-1
Turbine-Generator Condition Assessment – In Service
•
Plant electrical and mechanical personnel responsible for performing preventive maintenance (PM) or predictive maintenance (PdM) on the various equipment
•
Staff chemistry personnel responsible for maintenance of steam and water purity within specified limits
Input and insight will be required from all the above individuals in order to obtain the best possible assessment of the turbine-generator’s current condition. It is important to note that this procedure is designed so that, when each evaluation process is complete, there is only one of three possible outcomes for the systems, sections or components involved: •
The current condition is rated as excellent, and the inspection interval should be extended.
•
The current condition is rated as acceptable, and the inspection interval can be maintained.
•
The current condition has significantly or drastically changed, and the inspection interval should be reduced on various components or systems.
For each of these possibilities, only three types of recommendations are allowed in terms of when the unit should be taken off-line so that corrective action can be implemented: •
Immediate – This recommends that a weekend shutdown be scheduled to correct a potentially serious problem.
•
Intermediate – This recommends that the maintenance be deferred until the next scheduled outage.
•
Long term – This recommends that major maintenance be performed but not until the next scheduled overhaul, based on the most current evaluated condition.
These recommendations are highlighted in the summary page of the condition assessment (Data Sheet #17) report by the use of color coding where the following definitions are applied: •
Green. There are no perceived problems, and the system or component is expected to perform reliably until the next assessment.
•
Blue. No specific immediate or intermediate action is considered necessary, but the issue is significant enough to be monitored.
•
Yellow. Work is required at the next convenient outage; otherwise, a potential problem will develop that could become serious (threaten a forced shut-down) if not corrected.
•
Red. A specific component or system needs immediate attention. Risk of a component or system failure is considered high, and such a failure would cause loss of the unit for an extended time.
If the condition is rated as good or excellent, the item would be color-coded as green, and the summary report should recommend that the scheduled maintenance for these systems or components be long term, with the options to extend, maintain, or shorten the present intervals based on the further details gathered in the assessment. 1-2
Turbine-Generator Condition Assessment – In Service
In the case of the other two possible outcomes (immediate or intermediate) where some condition degradation is identified, the three possible colors allow the degradation to be further monitored (blue), treated as soon as conveniently possible (yellow), or immediately (red). As a general rule of thumb, unless an unexpected event of drastic proportions has suddenly occurred, the yellow or red rating would be based on more than just an anomaly noted in some indicator or sensor. An issue or problem flagged with a yellow or red code on the summary sheet should have the suspected root cause(s) identified and the appropriate action specified, based on this diagnosis. In other words, every issue associated with degradation of a system or component does not demand a shutdown of the unit. This ultimate action can be deferred until the condition is more precisely defined if the trending of the key criteria or parameters that identified the problem or issue indicates that time permits additional study. This decision of course, must be balanced against the system or component that is involved and the potential risk if a sudden unexpected breakdown of the system should occur. For example, evidence of either a balancing or a misalignment problem might be studied at greater length to identify which of the two is most likely, and the appropriate corrective resources could be planned in the most effective manner. However, the threat raised by a high particle count observed in the bearing lubricating oil system is more immediate and not worth the risk to the bearings and/or the turbine-generator system. A summary form, Data Sheet #17 (see Figure 1-1), is used to document the condition assessment based on information gathered on Data Sheets #1–16. These data sheets are provided in Appendix A.
Figure 1-1 Example of the Form Used to Rate the Final Condition of the Turbine-Generator
The intention for a limited number of outcomes and the use of color codes is to force the assessment to a conclusion and to summarize the results into recommendations that are easy to present to and use by management. Given that two of the three possible recommendations from the assessment could require the unit to come off-line earlier than originally planned, the 1-3
Turbine-Generator Condition Assessment – In Service
financial consequences of such a decision will invariably be a factor. Therefore, for a condition assessment to be effective, it must offer tangible recommendations, not simply provide a chronicle of assembled data. In this regard, it should be recognized that the assessment approach and method offered here is based on the best combined judgment of the personnel who are responsible for the operation and maintenance of the turbine-generator units. These judgments and recommendations are to be supported by the indicators, trends, known problems, and issues that have been collected specific to the unit since the last overhaul or conditions that may have occurred at some time in the past. In certain cases, particularly where the outcome indicates a “red” condition (where immediate attention is recommended), the human judgment of this in-service procedure can then be further supported by an analytical risk assessment in which all of the pertinent facts are considered. This supplemental assessment is most likely to involve issues associated with the rotating components, particularly those whose catastrophic failure would seriously harm or damage the unit, and not just shut it down. Examples where such analytical assessments might be appropriate for high-pressure (HP) and intermediate-pressure (IP) rotors include the potential for creep failures, solid-particle erosion (SPE), high-cycle fatigue (HCF), and erosion, typically for the first HP and first reheat stages. For LP rotors, the probability for stress corrosion cracking (SCC), HCF, low-cycle fatigue (LCF), water droplet erosion, stall/flutter, and corrosion fatigue would most often focus on the last stages of the LP turbine wheel attachments and blades. It is at this point where the responsibility of the condition assessment ends, and the work remaining transitions into a task of life cycle management (LCM). EPRI has produced a series of Life Cycle Management Planning Sourcebooks, each of which contains a compilation of industry experience, information, and data on aging, degradation, and historical performance for specific types of systems and components. These are potentially useful as references to complement and compare with the information assembled during this machine-specific condition assessment. The EPRI report name, number, and date of publication for Volumes 1–10 of the LCM Planning Sourcebooks are as follows Report Name
Report #
Report Date
Volume 1:
Instrument Air Systems
1006609
December 2001
Volume 2:
Buried Large-Diameter Piping
1006616
May 2002
Volume 3:
Main Condenser
1003651
March 2003
Volume 4:
Large Power Transformers
1007422
March 2003
Volume 5:
Main Generator
1007423
July 2003
Volume 6:
Feedwater Heater Controls
1007425
March 2003
Volume 7:
Low Voltage Electrical Distribution Systems 1007426
February 2003
Volume 8:
Main Turbine
1009071
January 2004
Volume 9:
Electrohydraulic Controls
1009072
September 2003
Volume 10:
Feedwater Heaters
1009073
December 2003
1-4
Turbine-Generator Condition Assessment – In Service
1.2
Procedure and Objectives
The principal objectives for applying the approach discussed in this section to perform an inservice condition assessment can be summarized as follows: 1. To assess whether the unit may be able to operate successfully until the next major planned overhaul. This recommendation is qualified based on the limits of the evaluation. The objective is to provide a basis for why an unscheduled outage or an extension of a major overhaul interval is recommended. The major limitation to the quality of this condition assessment is the inability to evaluate the condition of internal steam path stationary and rotating parts by means of direct visual inspection. However, by looking at and trending performance data and startup vibration, by performing visual and limited NDE in the exhaust ends of the turbine, and by reviewing bearing metal temperatures and other turbine supervisory instrumentation (TSI) data, it should be possible to obtain a reasonable picture of the internal health of the turbine-generator. 2. To determine what maintenance work needs to be performed on the unit during a weekend shutdown or upcoming future outages. The objective is to reduce the potential of a forced outage prior to the next major planned unit overhaul. 3. To provide additional input for determining the risk of failure (Pf) associated with extending turbine-generator outage intervals. Analytical calculations can be performed using stress (both dynamic and steady), material property variations, and operating data to analytically predict blade failure probability due to SCC, HCF, LCF, SPE, erosion, and creep. The need for such an analysis is dependent on the uncertainty of assigning failure probabilities to certain turbine components and their value to the utility. 4. To provide a relative risk assessment of the individual turbine sections and systems of a specific plant. When a formal system is in place, it becomes possible to compare the risk assessment to other units within the fleet. Three tasks are to be accomplished in the process of performing this level of condition assessment: •
Completion of evaluation forms (found in Appendix A). These should be kept as records and used as the basis to regularly assess, monitor, and trend unit condition change over time.
•
Development of recommendations for work to be performed on a short-term, intermediateterm, or long-term basis. These are done for each system and component and itemized on the summary sheet.
•
Development of recommendations for extending or shortening the time interval for the next planned major outage with justification as to why this work is required.
The overall unit condition evaluation of the turbine-generator and its systems will require input from technical support personnel involved in turbine-generator and plant maintenance and from operations and performance personnel. This information will be used to complete all attachments and to give an overall risk assessment of the turbine components and systems.
1-5
Turbine-Generator Condition Assessment – In Service
As noted previously, and in addition to the above recommendations, each component should be assessed using a color scheme that makes it easier for management to review and focus on the most critical issues or problems with the machine.
1.3
Planning a Steam Turbine-Generator Condition Assessment
Before proceeding with obtaining the data discussed in this procedure, the following four basic concepts/actions should be clearly communicated to the condition assessment team: •
Responsibility. The organization responsible for planning, scheduling, budgeting, coordinating, and performing the condition assessment of turbine-generators within the system should be clearly identified.
•
Resources. The personnel resources required to coordinate and perform condition assessment of turbine-generators should be identified. Additional resources at a specific station/site for which the assessment is being conducted and the time required of these resources should also be identified. This is necessary to ensure timely completion of data sheets and the turbine-generator unit condition assessment.
•
Budget and Schedules. The group responsible for developing a long-range and/or annual schedule must be identified. This plan and schedule should identify what turbine-generator units will require an assessment with a specified completion date. Included in this should be a budgetary estimate of all costs associated with performing the assessment. Estimated costs of different corrective actions may be found in the previously referenced LCM sourcebook, Main Turbine, 1009071, first published in January 2004.
•
Frequency. The frequency for conducting a condition assessment on a turbine-generator should be identified. In general, the following guidelines are recommended: –
A baseline assessment should be conducted as soon as practical after a major overhaul to provide a benchmark for subsequent comparison and evaluation.
–
A second assessment should be considered at mid-cycle relative to the next scheduled major inspection. For example, if a unit is on a 10-year inspection interval, the next condition assessment should be conducted after five years of operation.
–
The frequency of the inspection interval after this should depend upon the findings of the first condition assessment.
–
If the second condition assessment is satisfactory, the last assessment should be conducted one to two years in advance of the scheduled overhaul.
Contingent upon its findings, the final assessment may show that the planned overhaul can be postponed or must be performed earlier than planned. An interval extension may be supported by calculated failure probabilities based on analytical calculations.
1-6
Turbine-Generator Condition Assessment – In Service
1.4
Documentation of Condition Assessment
The final report of the turbine-generator condition assessment (along with any additional information/documentation) should be filed at the specific plant where the assessment was performed and at the utility general office. A series of recommended forms are provided in Appendix A: Data Sheets #1 through #17. Note that these forms are meant to be both comprehensive and generic in nature. Not all are required to be completed for many units. Only the most relevant information should be incorporated in the final condition assessment report (Data Sheet #17).
1.5
Condition Assessment Procedure
The assessment begins with a systematic collection of information that is typically available while a unit is in service. The data sheets in Appendix A have been organized in a manner and sequence to provide a natural framework for this process and a comprehensive report that can be updated at each successive assessment interval. The 17 separate data sheets are listed in Table 1-1. Generally, each series of data sheets consists of the three same basic parts: •
The first part is identified as an audit. This sheet provides a summary of relevant available data collected from the system as a whole or from individual sections.
•
The second sheet supplements the first by identifying key questions that are to be answered by means of an interview with the specialist, engineer, or operator directly responsible for maintaining and monitoring the system.
•
The third sheet uses a consistent format to summarize the findings drawn from the audit and interview. The perceived risk and considered need for action are both identified.
Four of the series are distinct from the others in that they form the minimum of information that a condition assessment would always include: •
Series #1 provides an overall review and assessment of the maintenance history for the unit, including a record of any upgrades made to replace original systems.
•
Series #15 is a checklist of 69 “out-of-limit” indicators. These highlight the myriad of unexpected problems or issues that are common to large steam turbine-generators.
•
Series #16 is a summary of the current long-range maintenance plans for the individual systems, sections, and components that are contained within the unit.
•
Series #17 is the consequence of the information obtained and reviewed in the preceding series of documents. It provides a one-page synopsis that lists the components/systems and then ranks their present condition. It is the last form completed, but it should be used as the first page of the report, with the subsequent sheets attached.
1-7
Turbine-Generator Condition Assessment – In Service Table 1-1 Breakdown of Condition Assessment Series
System or Component
Data Assembled and Reviewed
1
Maintenance History Summary
Record of modifications and upgrades
2
Turbine-Generator Vibration
Readings at minimum-maximum load, criticals
3
Bearing Metal and Oil Temperatures
Readings at minimum-maximum load
4
Section Performance Parameters
Readings at full load, valves wide open
5
Start-Up Operation
Record of starts, trips, service hours
6
Steam Purity
Frequency of tests, criteria, out-of-spec events
7
Lubricating Oil and EHC Analysis
Particle counts, presence of contaminates
8
Pump Start Tests
Frequency of tests, pressures, out-of-spec events
9
Valve Tightness Tests
Frequency of tests, criteria, sticking events
10
Turbine Trips and Tests
Record of trips, results, consequences
11
Turbine Monitoring Instrumentation
Readings at minimum-maximum load, calibrations
12
Generator-Exciter Condition
Readings, criteria, test results
13
Auxiliary System Operation
Readings, criteria, test results
14
Visual Inspection Results
HP, IP, LP inlets and exhausts
15
Checklist of Out-of-Limit Events
Record of unit upsets, actions, and consequences
16
Current Maintenance Plan
Record of inspections: last, next, frequency
17
Overall Condition Assessment
Summary with recommended actions
As noted, it may not be necessary to complete Data Sheets #2 through #14 for every unit. The extent to which the assessment is performed is at the discretion of the owners and operators. The checklist of indicators provides a snapshot to identify warnings of problems that should also be subsequently covered in the more detailed evaluation of a specific system or component. However, these warnings only highlight a potential problem and do not provide the additional detail that an audit or interview is meant to supply as a basis for estimating the risk of a failure and the type of action that is needed. Data Sheet #17 is essentially a digest of the critical systems and components, itemized for the main steam turbine in Data Sheet #16. It principally focuses on the fundamental issues or problems identified in the assessment, although it should include an assessment of certain systems whose health is critical to the reliable operation of the turbine-generator. In other words, it is just as valid to report that the conditions of these key systems are presently considered good, as it is to highlight the potential or immediate problems. As a general rule, the summary report should not exceed a single page.
1-8
Turbine-Generator Condition Assessment – In Service
The next 14 subsections of this section present a system-by-system breakdown of the checks, inspections, tests, etc., that are recommended on particular systems and components and some guidance in how these would be used to evaluate the operating health of the machine. The subsections are in the sequence of the data sheets previously shown in Table 1-1. Each set (series) of data sheets is referenced to a specific plant and unit number, with the following generic information included throughout the package of information that is finally assembled into a report: •
Turbine, generator, and exciter original equipment manufacturer (OEM)
•
Unit maximum dependable capacity (MDC) and unit design rating
•
Date the unit went into commercial operation and date the unit was last inspected
1.5.1 Turbine-Generator History, Upgrades, and Major Forced Outage Events After the proper planning and organizational measures have been finalized, a condition assessment may be initiated. The process starts by using Data Sheet #1 to prepare a detailed list of past maintenance performed on the specific turbine, generator, or auxiliary system. The purpose for completing this form is to have a concise and up-to-date record on the components and systems that make up the particular turbine-generator under assessment. This step in the assessment is meant to provide a historical perspective, which in turn is used to weigh the importance of issues or anomalies that are identified for certain systems or parts. It identifies what types of problems have been reported in the past, whether a problem appears to be chronic, and if that problem has been eliminated. Part 1(a) of this set of data sheets assembles a baseline of information that reflects the date that critical components of the unit were inspected. Key recommendations made as a consequence of these inspections are also identified, and whether the action taken was considered to be addressing a chronic problem is noted. To assist in defining what is considered to be a “critical component”, in the aforementioned EPRI Life Cycle Management Sourcebooks, the critical components of the main turbine, generator, controls, and support systems were organized by their function. The most prominent components in the assessment procedure are shown in Table 1-2. These components form an integrated system within the overall unit and are subject to a condition assessment as opposed to an assessment of single parts (nuts, bolts, tenons, etc.) that make up a component.
1-9
Turbine-Generator Condition Assessment – In Service Table 1-2 Critical Components Identified in Assessment Procedure Classification
Critical Components (If Present)
1
Pressure Boundaries
HP/LP Outer Casing, HP/IP Inner Casing
2
Piping
Interconnecting, Cross-Over, Cross-Under
3
Nozzles
Impulse and Reaction
4
Turbine Rotor Sections
HP, IP, and LP Rotors; Disks and Blades
5
Packing and Seals
Interstage Packing, Shaft End Seals, Oil Seals
6
Coupling and Bearings
Bolts, Shells, Journals, Pads
7
Front Standard Assemblies
Main Oil Pump, Speed Sensors, Trip Systems, PMG
8
Essential Instrumentation
Vibration, Temperature, Expansion, Speed
9
Generator and Exciter
Stator, Rotor, Windings, Hydrogen Seals, Coolers
10
Controls
Valves, Governors, Trips, Meters, Regulators
11
Lubrication
Bearing Oil System, EHC System
Part 1(b) of this set of data sheets provides a record of any modifications or upgrades that were made to improve the unit’s performance or reliability over its history of operation. In making this evaluation, it should be noted that to improve the reliability of the component does not necessarily require a design modification. For example, a replacement in-kind with a new component would reflect a condition whereby the accumulated damage to the original material is eliminated. A replacement re-sets the aging clock for the time-dependent mechanisms that caused the original material properties of the component or system to deteriorate. Part 1(c) assembles a list of the significant forced outages that have occurred and the components they affected (such as a blade failure, a bearing journal wipe, etc). If the root cause is suspected or known, these should be described. Any specific documentation available should be referenced or attached. When defining what type of information should be included in this section, a significant forced outage event is one in which the unit was automatically removed from service based on a turbine supervisory trip or was removed from service by operators in order to correct a deficiency that could have seriously jeopardized the unit’s operational reliability. Part 1(d) concludes this review with an assessment of the potential risk for failure, based on the past maintenance history. A judgment is required as to whether the forced outage events identified on the prior sheets are likely to be isolated events or symptoms of a longer, chronic problem. The judgment of potential risks is consistent with all the parts forming a condition assessment and is reduced to three possibilities: high, medium, or low. Because this is dealing with the maintenance history, the need to recommend action associated with each itemized problem is not required. Instead, this is determined on a component-by-component basis based on the most recent information made available through audit and interview.
1-10
Turbine-Generator Condition Assessment – In Service
As reflected on the data sheet, the maintenance history does not necessarily have to be limited to any specific item, but it also should not be too detailed. In other words, when subsequent parts of the evaluation identify specific issues, this summary provides a basis for deciding if this is a new problem, a chronic problem, or a routine problem. A new problem may indicate that a system or component is reaching the end of its useful life, or it may reflect the consequence of an action taken during the last overhaul. A repeat or chronic problem suggests that a more permanent solution is warranted, and the condition assessment may need to account for the additional planning and time to find this solution, as well as the potential risk if the problem worsens. A routine problem would be one that can be expected to require action on a periodic basis. Trending the rate at which the system’s or component’s condition appears to deteriorate provides the basis to either shorten or lengthen the interval of inspection and planned replacement. When beginning an assessment, a search of available databases is recommended to assist the condition assessment team in determining an initial list of historical issues that might be considered for further scrutiny in a unit-specific assessment. For example, details such as those compiled by the North American Electricity Reliability Council’s Generator Availability Data System (NERC-GADS) can provide a broad profile of what the plant has officially reported on the unit over time. So as not to be overwhelmed by this raw data, the user should seek information only on a specific unit and focus on entries that are classified as “unplanned events involving required maintenance action.” Events compiled within the NERC-GADS database are organized and reported with “cause codes.” A partial list of cause codes associated with the main turbine is shown in Table 1-3. A search made for a specific unit can produce a simple record like that shown in Figure 1-2. This forms a starting point to build and cross check a historical record of system or component maintenance. It can also be useful in identifying specific issues that would be pursued in the interview part of the assessment.
1-11
Turbine-Generator Condition Assessment – In Service Table 1-3 Cause Codes Associated with Critical Components Found in the Assessment Procedure Critical Components
NERC-Defined Cause Codes
Other Steam 4499: Turbine-Generator Other Problems (All Problems Components) 4000: Pressure Boundaries, (HP/LP HP Outer Casing and Hoods) Casing
4001:
4200:
4201:
HP Inner Casing
LP Outer Casing
LP Inner Casing
Interconnecting and 4270: Crossover Piping Crossover Piping
4279:
Nozzle Boxes
4009:
4010:
4209:
4210:
HP Nozzle Bolting
HP Nozzle Boxes
LP Nozzle Bolting
LP Nozzle Boxes
HP Rotor Sections 4011: 4012: (Disk, Blades, HP HP Buckets Stationaries) Diaphragms Blades
4013:
4014:
4015:
HP Diaphragm Unit
HP Bucket Fouling
HP Wheels HP Rotor Other HP or Spindles Shaft Problems
LP Rotor Sections, 4211: 4212: (Disk, Blades, LP LP Buckets Stationaries) Diaphragms Blades
4213:
4215:
4230:
LP Blade Fouling
LP LP Rotor Wheels - Shaft Spindles
Packing (Interstage 4020: and Shaft End), Oil HP Shaft Seals Seals
4021:
4022:
4220:
HP Dummy Rings
HP Gland Rings
LP Shaft LP Dummy LP Gland Gland Seal Seals Rings Rings System
Bearings and Couplings
4040:
4240:
HP Bearings
LP Bearings
Front Standard 4280: Bearing Pedestal Lube Oil Instrumentation and Pumps Associated TSI Essential Condition 4420: Monitoring (TSI) Turbine Instrumentation Vibration Data Major Turbine 4400 Overhaul >720 hrs Major (All Components) Overhaul
1-12
Miscellaneous Turbine Piping
4300:
4309:
Turbine Supervisory System
Other Turbine I & C Problems
4221:
4030:
4099:
4250: Other LP Problems 4222:
4430:
Turbine-Generator Condition Assessment – In Service
Caution should be exercised in analyzing raw entries in the NERC-GADS data. Scrutiny must be applied to ensure that events are categorized and counted correctly. This principally involves (a) reviewing the cause code component identification and (b) consolidating multiple events. Specific cause codes tend to be assigned subjectively. The same problem can be associated to causes that differ from plant to plant. Some plants use the “Turbine Vibration” and “Other Steam Turbine Problems” categories as a catchall. A review of each entry should be made to ensure that it was assigned to the proper component (when details on the entry make this possible). Multiple events should be consolidated into one when the records indicate that they were actually associated with a single, larger event. The most common example of this circumstance is when several attempts were required to return a unit on-line, with each being logged as a separate entry associated with “turbine vibration.” In such instances, the multiple GADS entries dealing with the same issue (often occurring within a period of hours or minutes) should be treated as a single occurrence for that type of event, and the individual hours for the multiple entries consolidated into a total for the single event. This step prevents overestimation of the frequency of reports made for the specific component.
Figure 1-2 Example of NERC-GAD List of Maintenance Outage Events for a Typical Unit
1-13
Turbine-Generator Condition Assessment – In Service
Beyond the previously listed LCM sourcebooks, additional research published by EPRI that is associated with historical experience on component or system reliability, upgrades, and major forced outage events related to turbine-generators is as follows (listed by year of publication): Survey of Steam Turbine Blade Failures, Project 1856-1, EPRI, Palo Alto, CA: 1985. CS3891. Condition Assessment Guidelines for Fossil Fuel Power Plant Components, EPRI, Palo Alto CA: 1990. GS 6724. Improving Maintenance Effectiveness Guidelines: An Evaluation of Plant Preventative and Prediction Maintenance Activities, EPRI, Palo Alto, CA: 1996. TR-107042. Main Turbine Performance Upgrade Guideline, EPRI, Palo Alto, CA: 1997. TR-106230. Low Pressure Rotor Rim Attachment Cracking Survey of Utility Experience, EPRI, Palo Alto, CA: 1997. TR-107088. Reliability Assessment of North American Steam Turbines, EPRI, Palo Alto CA: 2002. 1006952. Component Failure Database: Version 2.0, EPRI, Palo Alto, CA: 2003. 1004863. Predictive Maintenance Primer: Revision to NP-7205, EPRI, Palo Alto, CA: 2003. 1007350. 1.5.2 Turbine Vibration The signature obtained from the turbine-generator bearing vibration instrumentation may reflect a condition of misalignment or unbalance present within the system. Characteristics (frequency, amplitude, and phase) are typically processed by an expert or specialist in vibration diagnostics. For operators, vibration limits are set to prevent damage caused by exceeding the journal clearances. The condition assessment of turbine vibration is intended to identify the root cause of problems that may be indirectly reflected in the system vibration signature. When trended, the vibration signatures can also provide a sense of whether the problem is stable or deteriorating over time. Data Sheet #2 (a) assembles a complete set of turbine-generator vibration amplitude and phase angle data recorded from the unit at both full load and minimum load. Obtaining data on a unit roll-up and roll-down is also strongly recommended. In addition, a frequency scan should be recorded and attached to the respective data sheet at each bearing location. The specialist providing these data should also be the individual who is responsible for tracking, trending, and evaluating any significant data changes in comparison to readings taken at the last outage or to the first set of benchmark data taken after the unit is returned to service.
1-14
Turbine-Generator Condition Assessment – In Service
On #2 (b), the data recorded in the audit is then analyzed by the vibration specialist through an interview process. A series of questions are identified. These are designed to assist in soliciting details and opinions that go beyond the data collected and summarized on the initial sheet. Ultimately, the results from the audit and interview are summarized in the third sheet #2 (c), where both the risk of failure (low, medium, high) and the need for action (immediate, intermediate, long term) are identified relative to the turbine vibration. To assist in the interpretation of data assembled from both the audit and the interview, typical orbit plots showing symptoms of common problems registered in the vibration signature are provided in Figure 1-4. A more detailed discussion on the symptoms and interpretation of vibration measurements related to common turbine-generator balance and alignment problems can be found in Volume 3 of these Guidelines, specifically the Balance Primer and the Alignment Primer. In terms of possible actions relating to turbine vibration, the most common issues likely to be found in an assessment are: •
Repeated balancing attempts are ineffective. This involves a condition where significant changes in unit vibration have required numerous balance shots, but the running speed vibration amplitude and phase angle at the shaft rotation frequency remain high or are still not improved. For example, the failure to relieve the vibration problem may indicate that the problem is instead with a coupling, or that unbalance in the rotor system is highly static (where phase angles are in-line), indicated by the phase angles recorded at each bearing. When considering recommendations, determine if a coupling has excessive run-out between the coupling halves. If so, this condition will result in high vibration that cannot be improved by balance shots. Coupling disassembly will be required to eliminate excessive run-out. A large static unbalance in the rotor system may require shop balancing of one or more rotors to correct this condition.
•
Significant changes are noted at harmonics of shaft speed. Vibration frequency scans should identify any significant changes associated with one-half X, 2X, 3X, 4X, or 5X harmonics of shaft rotational speed (where X represents rotor speed). For example, a frequency that is less than one-half rotor running speed represents an oil whip instability if seen during unit operation. There would also be a significant difference between the filter-out and filter-in vibration amplitudes. A significant change in the 2X frequency response could be noted if a unit has a large peripheral crack (with orientation orthogonal to the shaft centerline). Such a response would be seen in comparing a 2X scan before and after cooling and re-heating a rotor by changing the main or reheat steam temperature by 50–75ºF (10–24ºC). A large difference between filter-in and filter-out vibration typically signifies significant vibration at other frequencies as noted above. If a vibration problem is identified, it should be considered as serious and treated immediately in order to prevent a potential catastrophic failure of the rotor during operation. Vibration technical experts should be immediately contacted to further assess this problem and for guidance as to possible unit shutdown.
1-15
Turbine-Generator Condition Assessment – In Service
•
A major change is noted when load is added. Major changes in rotor vibration amplitude and phase angle that occur from minimum load to full load can be due to alignment problems on the unit or other reasons (such as partial arc loading issues at minimum load). For example, on some units the static and dynamic vector components may show significant changes in amplitude and phase between minimum load and full load. A subsequent alignment check may reveal significant rim and face misalignment of one rotor to another rotor.
•
A major vibration change is noted when passing through shaft critical speeds. Significant vibration amplitude and phase angle changes at rotor critical speeds could be caused by rotor bowing. Such bowing can be due to operation at high temperature over a long time period or due to rubs in the steam packing or elsewhere within the rotor system. This would typically be seen in the HP or IP turbines.
•
Sudden step increases in vibration are measured at the journals. Large, sudden step increases in journal vibration often indicate a loss of rotating blade material. Such changes generally require immediate inspection to assess damage and determine other corrective actions that may be required.
As a rule, progressive changes in measured vibration over extended periods of time reflect degradation due to wear of the bearings, settling of the foundation, permanent rotor bowing, or the cumulative effect of individual section overhauls that eventually require a major correction. A condition assessment should seek to determine the point in time when a major unit realignment would be worthwhile to restore the unit to its originally aligned condition. An interview of the operators can aid in determining whether unit vibration problems have been chronic, are getting worse, or have only recently started. After multiple rotor overhauls, possible settling of the bearing foundation, and/or thermal distortion of stationary components, the unit will require extensive alignment, This will establish the radial position of the rotor with respect to stationary components and to re-establish the cantenary position of the bearings to the initial cantenary line. If a system becomes badly misaligned, it may become impossible to find a reasonable alignment solution without a complete disassembly of the unit to perform a tops-on and tops-off alignment to correct the problem. Pronounced step changes in vibration typically signify a situation that requires immediate concern and attention. These steps changes reflect a loss of noticeable rotating mass, often caused when portions of a rotating blade such as a tip or cover are lost. This indication is significant in that it may represent an early warning of the progressive deterioration of the structural system as a prelude to a more catastrophic failure. For example, if the blade is designed to have its natural frequencies fall within certain prescribed operating bands; the loss of part of a cover band may shift one of the fundamental frequencies into a condition of resonance. If a large blade fails at the platform or root, the loss of mass can be sufficient to cause an unbalance that will cause extensive damage to the entire machine. Such cases have been recently documented.
1-16
Turbine-Generator Condition Assessment – In Service
Figure A shows the orbit of a shaft with several concurrent whirls at different frequencies taken from unfiltered vibration. Figure B shows the same plot at synchronous speed where nonsynchronous frequencies have been filtered out, showing the unbalance whirl orbit.
Stiffness affects the shape of the orbit as noted by Figures C–E. Identical bearing stiffness gives a circular orbit, dissimilar stiffness gives an elliptical orbit, and cross coupling of stiffness in vertical and horizontal direction gives a rotated elliptical orbit. Misalignment can be indicated as shown in Figures F–H. Note that the orbit is highly elliptical in F, indicating poor alignment. In Figure G the orbit for two bearings on each side of a coupling are banana-shaped, indicating severe misalignment. In Figure H the misalignment is also severe and suggests backward precession. Fluid whirl, a subsynchronous fluid instability, occurs within the range of 30–48% of machine operating speed. Figure I shows precession of vibration in the same direction as shaft rotation, displaying a circular orbit and two key-phasor dots. If the dots slowly rotate against shaft rotation, the subsynchronous frequency is less than 50% of shaft speed; if the two dots remain stationary, the frequency is exactly 50% of shaft speed.
Fluid whip is a subsynchronous excitation at the first critical speed of a rotor that operates far above twice first critical speed and has vibration amplitudes equivalent to bearing clearance. The orbit for this is shown in Figure J. Note the multiple key-phasor dots that this excitation produces.
Rubs are a typical problem seen on steam turbines and are generally at 1X frequency for units that operate at less than twice first critical speed. If a rotor operates well above first critical speed, frequencies generated would be at 1X and 1/2X frequencies. Figure K shows a rub in a unit whose speed is well above first critical speed frequency and has a predominant 1/2X frequency.
Figure 1-3 Typical Orbits Showing Different Problems
1-17
Turbine-Generator Condition Assessment – In Service
In addition to the information referenced in Volume 3 of these guidelines, previous research published by EPRI related to turbine-generator vibration is as follows (listed by year of publication): Periodic Vibration Monitoring: Utility Experience, EPRI, Palo Alto CA: 1987. CS-5517 Symposium Proceedings: Rotating Machinery Dynamics, Bearings and Seals, EPRI, Palo Alto, CA: 1988. CS-5858. Applying Vibration Monitoring, EPRI, Palo Alto CA: 1991. NP-6340. Shaft Alignment Guide, EPRI, Palo Alto CA: 1999. TR-112449. Technology Development for Shaft Crack Detection in Rotating Equipment Using Torsional Vibration, EPRI, Palo Alto CA: 2003. 1009060. 1.5.3 Bearing Metal and Oil Temperatures Pressure-lubricated journal-type bearings support the rotor shaft elements at both ends. A thin oil wedge that is created by rotation within the stationary pads supports the shaft. Maintaining the proper thickness of this oil wedge is critical to preventing bearing vibration due to oil film instabilities. Since steam pressure differential across most turbine stages produces a net thrust along the shaft, the thrust bearing provides a reaction force to this differential and limits the rotor’s axial position to maintain proper axial clearances between the stationary and moving elements. Loss of lubrication, oil temperature excursions, or contamination of the lubricating oil can result in serious damage to the bearings. This step of the condition assessment is meant to ensure that no unusual loading condition is occurring at the rotor journals or bearings. As part of the assessment, bearing metal, oil inlet, and oil outlet temperatures should be recorded at maximum and minimum load and recorded as recommended in Data Sheet #3(a). Recording these data should also be considered throughout the load range. As in the previous Data Sheet, #3(b) identifies a list of questions that should be answered as part of the specialist interview. The series concludes with Data Sheet #3(c) and the assessment of risk and need for action relative to this system or component. The most common issues likely to be found in an assessment are: •
Pronounced changes in metal temperature since the last evaluation which can be an indication of alignment changes or bearing wipes. Both conditions generally represent a high risk to the overall system. As with a step change in vibration, they should be taken seriously and corrected promptly.
•
Sudden spikes in bearing metal temperature may indicate that a bearing may have actually wiped. If a sudden metal temperature change has occurred, this can mean that other bearings are now more heavily loaded. Their journal radial position may have changed, meaning that the rotor may now be running close to a rub condition that could take the unit off-line and potentially damage rotating and stationary components.
1-18
Turbine-Generator Condition Assessment – In Service
It should be noted that bearing vibration amplitude or metal temperature changes could be observed on some units due to valve arc loading. For these units, this may be normal. It should be considered an abnormal condition in the assessment only if this is the first time it has been observed. Actions that may be required to correct a bearing problem can include checking the rotor radial position to either oil or gland bores along with the disassembly of couplings and realigning the turbine depending on the radial position found. Such work can be easily accomplished during a limited outage if planned sufficiently in advance. A partial list of research published by EPRI related to aspects of turbine-generator bearing operation is listed as follows (by year of publication): Guidelines for Maintaining Steam Turbine Lubrication Systems, EPRI, Palo Alto CA: 1986. CS-4555. Manual of Bearing Failures and Repair in Power Plant Rotating Equipment, EPRI, Palo Alto CA: 1991. GS-7352. Bearing Troubleshooting Advisor, Version 2.0, EPRI, Palo Alto CA: 1994. AP-100531-R1. Bearing Technology Topics: Various Technical Papers, Volumes 1 and 2, EPRI, Palo Alto CA: 1999. TR-113059-V1 and V2. 1.5.4 Thermal Performance The state-of-the-art methodology in axial steam turbine thermal design takes into account many factors, including optimal stage-to-stage loading (enthalpy drop). Other design factors include: •
Airfoil shapes used in nozzles, diaphragms, rotating blades/buckets (three-dimensional designs being increasingly used)
•
End-wall shape (that is, the inner and outer boundaries of the flow passage
•
Blade/bucket shroud configuration
•
LP turbine stage wetness removal
•
Blade/bucket tie-wire losses
•
Interstage sealing
The dry internal efficiency of modern steam turbine sections ranges from the high 80s percentage range to the low 90s percentage range. Wetness losses in both the HP (nuclear) and LP (nuclear and fossil) sections of units reduce this efficiency somewhat, depending on the wetness level. Individual stage efficiency, particularly in the superheated early stages of the LP turbine, can exceed 90%. However, the overall multistage expansion is always less than the individual stage efficiencies. 1-19
Turbine-Generator Condition Assessment – In Service
As part of the overall condition assessment, thermal performance degradation is used as an indirect indicator for identifying problems that may be developing within the HP, IP, or LP turbine steam paths that in some manner inhibit or disrupt the flow so that noticeable losses are produced. These losses are reflected as a higher turbine cycle heat rate or loss of power output. To facilitate the interpretation of performance-related test data, Tables 1-4 and 1-5 reflect common changes in the condition of different components of the turbine steam path in terms of their consequence on measurable performance cycle parameters. Table 1-6 provides general guidance in terms of the impact on heat load due to various forms of steam leakage into the condenser. To make this assessment, the most current data should be gathered and recorded using Data Sheet #4. Data assembled on 4 (a) should be trended at the same load point to further assess performance degradation over time. Often, a key in the interpretation of overall performance data is to isolate the cause within the respective section of the overall steam path. Identifying the source of a suspected flow restriction can be approached in a systematic manner: •
Steam extractions, reheat conditions, and moisture separators are logical cycle points that can be used to identify turbine sections or plant systems that may be deteriorating.
•
The first HP (control) stage and last LP stage have a significant influence on the overall unit performance. Although most turbines consist of a large number of individual stages, only the first and last stages tend to be significantly influenced by changes in the flow rate. First stage performance is primarily more sensitive to variations in load. The last stage is more sensitive to variations in both load and backpressure.
•
Throttle flow factor is usually an indication of increased nozzle erosion. Trending this value will give an indication of the rate of nozzle degradation that may be occurring for the specific unit. Evaluation of other performance parameters may show other types of deterioration, such as nozzle area closure or significant deposit formation on stationary and rotating blades.
•
Excessive erosion on the leading edge of a last stage blade may be an indication of high feedwater levels in the neck heaters or of problems in the moisture removal system in the LP section. If significant erosion is seen, such as 1/8" (3.18 mm) or greater on the leading edge of a last stage blade since the last inspection, attention to feedwater heater controls should be considered along with a detailed inspection of the LP section moisture removal system at the next scheduled overhaul. If numerous tube leaks have occurred in these heaters, immediate actions should be taken to eddy current inspect and plug suspect tubes. A long-term fix may be to re-tube these feedwater heaters.
•
Operating at low loads and high backpressures can result in excessive moisture droplet erosion in the last stage blades or the development of fatigue cracks due to a vibratory condition referred to as flutter. This can occur as a result of increased cycling duty. In the current utility financial operating environment, many units designed in the 1960s and 1970s for base load operation are now being run to match peak consumption swings, sometimes on a daily basis. When performing a condition assessment, changes to the condenser system or operating pattern of the unit should be noted, and the condition of the LSB monitored for these signs of distress during limited outages where the condition can be visually inspected.
1-20
Turbine-Generator Condition Assessment – In Service
Performance-related problems such as blade erosion or nozzle deterioration are generally not catastrophic, unless a severe condition is allowed to persist. The incentive to address the thermal performance issues is therefore typically based on the loss of efficiency produced by the deterioration of the steam path. However, the location and extent of work that may be required to restore a component should be seriously evaluated so that damage that would require only small repairs is not allowed to escalate into a major scope of activities that compound the risk of an extended outage due to unforeseen circumstances. For example, erosion in the HP nozzles can normally be tracked by periodic visual inspection with corrective actions planned for the next overhaul. Monitoring performance parameters such as those identified in Tables 1-4 and 1-5 can assist in identifying consequences of such erosion. However, if erosion cuts or erosion are allowed to become deep, the time required for repair will be extended based the amount of weld repair required to correct the problem. The location of the repair may also increase the level of risk assigned to a reliability issue. For example, extensive erosion to the HP nozzle plates can then result in significant erosion to the HP first stage buckets. Unintended removal of material from the buckets can make them more susceptible to resonant vibration, particularly at certain valve points. It is this type of secondary consequence that should always be considered when assessing the potential need and urgency to address an original problem. A partial list of research published by EPRI related to turbine-generator performance is as follows (organized by year of publication): Assessment of Supercritical Power Plant Performance, EPRI, Palo Alto CA: 1986. CS-4969. Heat-Rate Improvement Guidelines for Existing Fossil Plants, EPRI, Palo Alto CA: 1986. CS-4554 Fossil Unit Performance: 1965-1984, EPRI, Palo Alto CA: 1987. CS-5627. Solid Particle Erosion Technology Assessment, EPRI, Palo Alto CA: 1994. TR-103552. Thermal Performance Engineering Handbook, Volume 2: Advanced Concepts in Thermal Performance, EPRI, Palo Alto CA: 1998. TR-107422-V2. Turbine Steam Path Damage: Volumes 1 and 2, EPRI, Palo Alto CA: 1998. TR-108943.
1-21
Turbine-Generator Condition Assessment – In Service Table 1-4 Effect of Component Condition Changes on Fossil Cycle Performance Parameters (at Valve Wide Open Operation) Condition
Throttle Low #/hr
PT
P1st
PHRH
PLP
H. P. Efficiency
I. P. Efficiency
Increase TT
↓
N. C.
-
↓
↓
↓
-
Increase THRH
-
N. C.
-
↑
↑
-
-
Increase A1st (SPE) in HP
↑
N. C.
↑
↑
↑
↓
-
Increase AHRH (SPE) in IP
-
N. C.
-
↓
-
-
↓
Decrease A1st (Deposits and Peening) in HP
↓
N. C.
↓
↓
↓
↓
-
Decrease AHRH (Deposits and Peening) in IP
-
N. C.
-
↑
-
-
↓
Decrease A2nd (Deposits) in HP
↓
N. C.
↑
↓
↓
↓
-
Increase A2nd (Rubs) in HP
↑
N. C.
↓
↑
↑
↓
-
Decrease ALP (Deposits and Damage) in LP
-
N. C.
-
-
↑
-
-
Note that every change in turbine condition results in a different three-key pressure pattern.
1-22
Turbine-Generator Condition Assessment – In Service Table 1-5 Guidance for Interpreting Turbine Cycle Steam Flow and Unit Load Changes Type Problem
Timing
Throttle Flow
Section Efficiency
Electrical Load
SPE
Gradual
Increase
Decrease (-HPη decrease greatest at light load)
Increase or essentially constant
Deposits
Gradual
Decrease (may increase after shutdown)
Decrease (may increase after shutdown)
Decrease (may increase after shutdown)
Foreign Object (Wrench, Bolts, etc.)
Abrupt following outages
Decrease
Decrease
Decrease
Peening (weld bead)
Abrupt following boiler repairs
Decrease
Decrease
Decrease
Mechanical Failure
Abrupt anytime, usually during operation
Usually decrease
Decrease
Decrease
Water Induction
Abrupt anytime during operation
Slight increase
Decrease
Decrease
Vibration
Abrupt, usually most severe at first startup
Slight increase
Decrease
Decrease
Steam Whirl
Abrupt at first startup
Slight increase
Decrease
Decrease
Internal Leakage (Balance Hole Plug)
Abrupt following overhaul
Increase
HP turbine – decrease
Decrease
Internal Leakage (Inner Shell)
Gradual
Slight increase
Decrease
Slight increase
Broken Valve Stem
Abrupt
Decrease
Decrease
Decrease
1-23
Turbine-Generator Condition Assessment – In Service Table 1-6 Effect of Leakage to the Condenser on Heat Rate and Load Effect of 1% Leakage to the Condenser on Fossil Reheat Origin of 1% Leakage Flow
On Heat Rate
On Load
Throttle
0.83%
0.94%
HP Turbine Exhaust
0.53%
0.69%
Ahead of Intercept Valves
0.69%
0.56%
Crossover
0.44%
0.44%
Rules of Thumb 1% ηHP
=
0.16% heat rate or 0.3% kilowatt
1% ηIP
=
0.12% heat rate or 0.12% kilowatt
1% ηLP
=
0.5% heat rate or 0.5% kilowatt
1% flow increase =
0.94% increase in kilowatts
1°F (0.56°C) temperature increase = 0.08% decrease in kWs and 0.024% better heat rate (VWO) 1 Btu TEL (total exhaust loss) = 0.1% poorer heat rate 10°F (5.6°C) decrease in TT increases ηHP 0.11%. A 5% increase in stage pressure flow relationship is cause for alarm. 1% ∆P increase in steam path = 0.1% poorer heat rate A 1% change in P1st due to a change downstream indicates a 1.5% change in flow for a 1.25 pressure ratio control stage (Curtis stage, not single Rateau stages). nd
A 1% change in P1st due to a change upstream of the 2 stage indicates a 1% change in flow . Bench mark η HP 2% is better than heat balance. A 10% nozzle area increase due to SPE results in a 6½% loss in stage efficiency for the control stage and 3–4% for the other stages. A 10% decrease in control stage nozzle area decreases the flow passing capacity by 3%. % ∆P SV and CV 4% (VWO) % ∆P IV 2% % ∆P crossover 3% A 1% increase in HP and IP turbine stage pressures due to a restriction downstream of the stage results in a 0.6% increase in pressure upstream for an impulse type stage and 0.7% for a 50% reaction stage.
1-24
Turbine-Generator Condition Assessment – In Service
1.5.5 Unit Start and Load Data Unit start and load data are primarily used to assess (a) rotor life, (b) issues associated with creep damage in the first HP or reheat stages, and (c) low-cycle fatigue in the last two or three rows of the LP turbine, depending on the size of the blades and the loads they exert on the blade root attachments. The condition assessment should evaluate the record of accumulated start-stop events to determine the potential for these damage mechanisms to affect the future reliability of the rotor components. To perform the assessment, the number of hot, cold, and warm starts are gathered since the last overhaul along with the unit service hours and recorded on Data Sheet #5. These totals (since the unit went into commercial operation) should also be determined to assess creep (total hours) or low-cycle fatigue damage (total start-stop cycles). Unit start/load information for the time period from the last inspection can be entered into Equation 1-1 to assess the potential need for a major inspection. This equation equates unit start types and trips to equivalent operating hours (EOH) for a unit. EOH = 20 x CS + 10 x HS + 5 x WS + SH+ 40 x FLT + 10 x LFLT
Eq. 1-1
EOH = Equivalent operating hours CS
= Number of cold starts
HS
= Number of hot starts
WS
= Number of warm starts
SH
= Synchronized hours since last overhaul
FLT
= Number of unit trips above 75% of full load
LFLT = Number of trips on unit at less than 75% full load (includes annual overspeed trip). Using this formula, the maximum number of equivalent operating hours between outages should not exceed 80,000 equivalent operating hours (EOH) for base load and load cycling units. This assumes that unit condition is acceptable and that normal predictive/preventive maintenance and system tests are being performed on the unit as specified by the OEM. It should be noted that higher EOH might be allowed for newer designed machines. Experience may show that 80,000 EOH is not advisable in some situations, especially in supercritical HP and IP sections. Judgment, experience, and a continuing condition assessment program for a unit should be the basis for selecting EOH limits. If detailed component life consumption models exist for the critical elements, such as blades and rotors, this information can be used to determine the probability of problems due to creep, highand low-cycle fatigue, stress corrosion cracking, or other blade issues without having to use Equation 1-1 above. The additional accuracy provided by such models offers a basis for changing the turbine overhaul interval. 1-25
Turbine-Generator Condition Assessment – In Service
A partial list of research published by EPRI on turbine-generator unit start-up and loading is as follows (organized by year of publication): Variable Pressure Operation: An Assessment, EPRI, Palo Alto CA: 1990. GS-6772. Steam Turbine Start-Up and Loading, EPRI, Palo Alto CA: 1998. CD-110966. 1.5.6 Unit System Steam/Water Purity Inadequate control of steam or water purity can introduce contaminants into the steam path, particularly at the LP section Wilson Line (“wet zone”), causing corrosion damage and/or plugging to the steam path. As such, a change or deterioration in steam or water purity can affect both reliability and output and therefore is included as a separate part of the condition assessment. Data Sheet #6 outlines the purity tests that are typically conducted and at what locations. In general, purity is defined by measurements of: •
Specific conductivity
•
Cation conductivity
•
Presence of sodium
•
Presence of silica
•
The pH level
Typical limits are identified on Data Sheet #6(a). Purity is monitored at the polisher or economizer outlet, the main steam or reheat steam inlet, and the condensate pump discharge. Any out-of-limit chemistry excursions that have occurred since the last assessment should also be noted. If specific problems have been experienced, determine by means of the interview when it occurred, how the problem was resolved, and then assess the impact on future turbine reliability. In terms of interpreting the potential impact of water impurities as a consequence to unit condition, the following guidance is offered: •
If a steam or water chemistry upset has occurred, there may be an increased risk of caustics, sodium, chlorides, copper, or sulfates being deposited on steam path components.
•
Copper typically deposits in HP turbine sections with chlorides, sulfates, and caustics in the LP section near the phase transition zone. These deposits can lead to pitting and eventually stress corrosion cracking (SCC) or corrosion fatigue (CF) of turbine blades, wheel dovetails, or other highly stressed rotating components. The leading indicators of this problem can be determined by visual observation of pitting on LP wheels and blades and the coloration of the deposit (white or light grey). Such deposits, if seen, should be analyzed to determine their composition in that specific deposit at the stage in which they were located.
1-26
Turbine-Generator Condition Assessment – In Service
•
Copper deposition in the HP section can be determined by noting a gradual loss of steam flow passing capability and turbine performance degradation. There have been cases where copper deposits were removed by flooding the specific turbine section with a mixture of ammonia and water. Several utilities have reported the success of such an activity.
A partial list of additional research published by EPRI on turbine-generator steam and water purity is as follows (organized by year of publication): Guide for the Use of Corrosion Resistant Coatings on Steam Turbine Blades, EPRI, Palo Alto, CA: 1987. CS-5481. Cycling, Startup, Shutdown and Layup Fossil Plant Cycle Chemistry Guidelines for Operators and Chemists, EPRI, Palo Alto CA: 1997. TR-107754. Steam, Chemistry and Corrosion in the Phase Transition Zone of Steam Turbines: Volumes 1 and 2, EPRI, Palo Alto CA: 1997. AP-108184. Corrosion of Low Pressure Steam Turbine Components, EPRI, Palo Alto, CA: 2000. 1000557. Turbine Steam Chemistry and Corrosion: Electrochemistry in LP Turbines, EPRI, Palo Alto, CA: 2001. 1006283. 1.5.7 Lubricating Oil and EHC Fluid Testing Contaminated lubricating oil can damage bearings, thrust runners, and rotor journals. It can also cause problems on units that use lubrication oil in control systems, such as mechanical hydraulic controls. This increases the risk of overspeed events and result in problems in other turbine systems. Any sudden increase in particle count size, water, color, or neutrality number may require plant actions to correct such problems before the turbine-generator experiences problems. Data Sheet #7 is provided to guide the assessment of lubricating oil and EHC fluid condition. As with the preceding data sheets, it begins with an audit #7(a), moves to the interview with the system specialist in #7(b), and concludes with an assessment of the risks and recommendation of action in #7(c). Typical problems that can be experienced due to contaminated lubricating oil or EHC fluid are discussed below. Lubricating oil: •
A copy of the lubricating oil particle counts, water neutralization number, and color should be obtained, and this data should be recorded on Data Sheet #7a. Oil samples are normally taken near the booster oil pump or eductor, on the inlet side of the oil tank return screens and from the discharge side of the oil purification system.
1-27
Turbine-Generator Condition Assessment – In Service
•
These values should be trended and compared against the OEM limits. High particle counts can result in scored journals and increase the risk of bearing wipes (thrust and heavily loaded journal bearings) along with having a detrimental affect on the hydraulic control system. The control system components can seize due to dirt accumulation and, in some severe cases, fail to properly function during transient events on the machine.
•
Water in the lubricating oil can result in reduced oil film thickness at operating speed and loss of insulation on generator and exciter bearings and the insulated hydrogen seal casing. Electrical discharge from the journals to the bearing babbitt material can result in the frosting of bearings and journals and increase the risk of a wiped journal or thrust bearing during unit operation.
•
Contaminated lubricating oil can be corrected by installing a vacuum dehydration system at the oil tank to remove both water and particles from the oil. There are recirculation pump and filter units that can be installed at the oil tank to clean particles from the oil. These actions will clean the lubricating oil, but they do not clean the piping, which may also be contaminated.
EHC Fluid: •
A copy of the EHC fluid sample test results should be obtained and recorded on the same set of data sheets. It is important that fluid quality be maintained within the limits specified in Table 11-1 of the EPRI report EHC Fluid Maintenance Guide (1004554) to prevent hydrolysis on valve spools. Hydrolysis can cause sticking or binding of EHC components or cause stress corrosion cracking in stainless steel parts of the EHC system. A copy of this table is shown on the next page for reference.
1-28
Turbine-Generator Condition Assessment – In Service
EHC Test Results 1
Parameter
EPRI Limit of 4.0 max or monthly increase of 0.5
Color
Frequency
Reference
Achievable2
Comments
Monthly
4.4.3.1
1.5 (Light Tan)
ASTM D-1500 Color Criteria; OEM’s limits are included in Appendix H. Fluid Color Scale Comparison Chart is included in Appendix G5. System contamination measurement most important after breach of system
Viscosity
+/- 10% initial value
Monthly
4.4.3.2
N/A
Acidity (mg KOH/g)
< 0.1
Monthly
4.4.3.3
< 0.05
Chlorines (ppm)
< 50
Quarterly
4.4.3.4
< 10
System contamination measurement most important after breach of system
Water (%, ppm)
< 0.1, 1000ppm
Monthly
4.4.3.5
< 0.05, 500ppm
2 main methods for improving this are dry air purge and vacuum dehydration
Mineral Oil 3 (%)
< 0.5
Quarterly
4.4.3.6
5 - 10
Monthly
4.4.3.7
> 20
Can cause erosion problems on stagnant areas of system, fluid types have different values
Particulate4
482°C]). Creep damage (material plastic flow) is a function of the material, operating temperature, time, and stress.
Operating data may also be required for the rotor analysis. Operating data are used in the damage accumulation model of the analysis and the remaining life prediction. A history of steam temperature, pressure, rotor starts, ramp rates, operating projections, and other data provide the details for a remaining life assessment. Future or planned component utilization is also important 5-23
Turbine-Generator Condition Assessment
as an input for life prediction until the next planned inspection interval, or the data may be used to plan the next inspection period. 5.8.3 EPRI-Supported Rotor Boresonic Inspection Over the years, EPRI has undertaken many programs to evaluate the inspection technologies available to the industry. Probably the most recognized is the program to evaluate techniques for performing ultrasonic inspection from the bore surface (boresonic) of steam turbine and generator rotors. During the period immediately after the TVA Gallatin rotor failure in 1974—a failure that was ultimately determined as caused by a crack on the bore surface—organizations that had experience in rotor inspection began to develop and implement techniques for inspecting rotor forgings from the central bore hole. Prior to that time, only the OEM had been able to perform the inspection. During this period, the number of vendors performing the inspection increased, and although the inspections they were performing were similar in nature, variations did occur. Utilities that earlier had only their OEM to rely on to perform the inspection now had several vendors to choose from, and each claimed to be better than the last. In the early 1980s, EPRI developed a program to evaluate the inspection systems that were rapidly being deployed to meet the needs of the utility industry. A series of test blocks were designed with flaws fabricated in them to resemble the types of flaws thought to exist in rotor forging. To evaluate the test blocks, utilities would urge their vendors to participate in the program by performing a blind test of the test blocks and have their results evaluated and reported by EPRI. That program is still in effect today, and by the mid1990s, most of the domestic companies known to provide boresonic inspection services had participated. The following reports and system evaluations provide a good foundation for understanding boresonic capability as it is applied today. The EPRI report Rotor Boresonic Inspection Guidelines [24] is recommended as a detailed reference. It provides guidance for planning and implementing a viable rotor bore inspection program for steam turbine and generators. It also presents information useful to making a runretire decision on rotor use. Sections in the report cover: •
History, including the Gallatin failure
•
Rotor material
•
Boresonic inspection principles
•
Discussion of various inspection systems (from a survey)
•
Description of the boresonic performance demonstration plan
•
Conclusions and recommendations for performing an inspection
Also developed by EPRI in the 1980s was a boresonic inspection data evaluation DOS-based computer program, SAFER (Stress and Fracture Evaluation of Rotors), used to analyze the data obtained from turbine-generator boresonic inspections and evaluate the remaining life of these rotors. In 2004 EPRI released SAFER-PC (Stress and Fracture Evaluation of Rotors- Personal Computer), product number 1010003, with numerous upgrades. 5-24
Turbine-Generator Condition Assessment
SAFER-PC combines transient thermal-elastic finite element stress analysis, fracture mechanics, material property data, and the clustering and linking of surface defects identified from nondestructive examination (NDE) data to assess the remaining useful life (RUL) of steam turbine or generator rotors. SAFER-PC can also perform probabilistic analysis of remaining life and material creep. It includes modules that allow imported boresonic NDE data in ASCII format to generate a flaw file, and a flexible approach to curve-fitting fracture toughness data is also available. SAFER-PC’s powerful user interface enables the program to be run on a number of current operating systems, with a variety of options for displaying and archiving analysis results. SAFER-PC can be used to assess the remaining life of critical rotating equipment in lifeextension studies, potentially savings millions of dollars in replacement rotor costs. The program can be used to reduce the uncertainty in risk analyses of older turbine-generator rotors, reducing the possibility of rotor burst that is an issue involving both a safety and consequential cost. Many plants are being cycled more frequently with larger daily variations in steam inlet temperatures. SAFER-PC can evaluate the increased rate of damage associated with this mode of operation, enabling a more accurate assessment of the costs and benefits of flexible plant operation. The new version of SAFER, SAFER-PC Release 2.2, 1013044, [54] includes improvements to the user interface developed as a result of the May 2004 user training held in Charlotte, North Carolina, were formally incorporated in the official software release. 5.8.4 Boresonic System Evaluation Procedures A boresonic system evaluation begins with the member utility making a request to sponsor a particular inspection vendor. The evaluation blocks are then shipped to the participant where a series of 15–25 scans of the block is performed. At the end of the inspection phase of the program, the participant is provided with a map of the flaw locations. Included are several locations where either geometry reflections exist or there are no flaws at all. The boresonic measurements taken by the vendor that describe the flaws are sent to the EPRI NDE Center for statistical evaluation. The mean value and standard deviation for all scans are computed for each dimension included in the study for the 70-odd flaws in the blocks. A linear least squares fit of all the data, for each measured dimension, and for each defect type is then used to determine the best fit of the measured values versus the true dimension. The linear best fit is represented by a slope, intercept, correlation coefficient, and the root-mean-square (rms) error. The slope and the intercept are indicative of the systematic error. The correlation coefficient is a measure of the strength of the linear relationship between the true and indicated value. Standard deviation and rms error are measures of the spread and accuracy (relative to the best fit line), respectively, in the data.
5-25
Turbine-Generator Condition Assessment
Further details associated with specific inspection vendors can be found in the following EPRI reports: •
NEI Parsons Ltd. Boresonic Inspection System Evaluation, TR-102126, [25].
•
Northeast Inspection Services, Inc. Boresonic Inspection System Evaluation, TR-102256, [26].
•
WesDyne International UDRPS Boresonic Inspection System Evaluation, TR-106234, [27].
•
General Electric Company Boresonic Inspection System Evaluation, TR-107174, [28].
•
Reinhart & Associates, Inc. Boresonic Inspection System Evaluation, TR-108423, [29].
•
Boresonic System Performance Guide, TR-104355, [30] provides a comparison of all system evaluations prior to 1994.
5.8.5 Inspection of Boreless Rotors Nondestructive examination is generally less essential for solid (boreless) turbine rotors than for bored rotors because stresses are lower without a central bore hole. Occasionally, situations arise in which examination of a boreless turbine rotor is not only advisable but essential to maintain confidence in the rotor’s capacity for continued safe operation. Ultrasonic examination of rotors from their central bore holes has become an accepted in-service inspection method throughout the utility industry. Without a bore, ultrasonic examination of the rotor must be conducted from the outer periphery, a task that is made difficult by the periphery geometry and lack of a continuous, uniform surface from which to conduct the inspection. Boreless turbine rotors can be inspected reliably if the utility is willing to invest the time and effort needed to conduct multiple inspections from the limited surfaces that are available for transducer placement. Multiple direction angulation techniques must be used to inspect regions that are otherwise inaccessible. To maintain sensitivity, calibration-correction factors must be applied to account for the fact that the beam cannot be introduced orthogonally to the major dimension of the flaw, which, on the most conservative assumption, is assumed to lie in a radialaxial plane. Correction factors must also account for the possibility that the flaw does not lie in the center of the beam where sensitivity is optimized. Each test must be carefully designed to cover a specific region where normal inspection procedures cannot reach, and sensitivity corrections must be used to analyze the data and estimate reflector sizes properly. The EPRI report Guide for In-Service Ultrasonic Inspection of Boreless Turbine Rotors and Other Solid Shafts, TR-101836, [31] is recommended as a reference document. It presents details on methods of inspecting boreless rotors as well as determining appropriate inspection angles and corresponding sensitivity correction factors. 5.8.6 Inspection of Steam Turbine Disk Blade Attachments Although many innovative design changes have been developed to address stress corrosion cracking (SCC) of turbine disk keyways and bores, the blade attachment regions (disk rim) remain susceptible to SCC. Blade attachment designs are varied, and their geometrical shapes complicate in-service inspection. Accurate knowledge of the attachment geometry is needed to develop and apply reliable NDE, but this information is not always available. 5-26
Turbine-Generator Condition Assessment
The reliability of turbine disk blade attachments and the capability of in-service inspection methods to detect and accurately characterize cracking have received considerable attention from utilities. Although reliable NDE can be performed after blade removal, this approach is costly. It has created a strong incentive to develop reliable methods that do not require blade removal. The EPRI report Inspection of Turbine Disk Blade Attachment Guide, TR-104026-V1, [32] is recommended as a reference document. It introduces the topic of attachment NDE and describes conventional and advanced NDE methods and principles. It reviews the types of cracking and the potential crack locations in both axial and circumferential entry blade attachment designs. Current conventional NDE techniques and approaches are described including visual, liquid penetrant, magnetic particle, eddy current, and ultrasonic inspection. Application principles and inspection advantages and limitations are also presented. The guide features ultrasonic inspection because, at present, it is the only method available that is capable of examining the interior regions of the blade attachment. Complex design geometry introduces many opportunities for false calls during ultrasonic inspection. Reflections from the geometrical features of the various attachments can interfere with proper interpretation of inspection results. The report discusses these effects and the difficult problem of determining the geometry of blade attachments that is necessary for design and application of proper inspection procedures. An ultrasonic technique for measuring attachment geometry is also presented. EPRI report Steam Turbine Disk Blade Attachment Inspection Using Linear Phased Array Ultrasonic Technology, 1000122, [33] is also recommended as a reference. It provides a brief history of disk blade attachment inspection and the fundamental principles of ultrasonic linear phased array technology. In 1997, EPRI developed a technique to inspect steam turbine disk blade attachments using linear phased array technology. In 1998, a company to commercialize the technology was sought to make the technique available to utilities with turbines that incorporated the General Electric straddle-mount dovetail design. By the end of 1999, General Electric had commercialized the technology, had refined the technique, and was offering it to their utility customers. The cited report describes the development of the project as well as its status up to November 1, 1999. EPRI has developed a technique to inspect steam turbine disk blade attachments utilizing linear phased array technology. The EPRI guide Field Application for Ultrasonic Linear Phased Array Inspection of Straddle-Mount and Axial-Entry Disk Blade Attachments, 1000663, [34] provides a brief history of disk blade attachment inspection and the fundamental principles of ultrasonic linear phased array technology. It also describes the development of techniques for the inspection of straddle-mount and axial-entry disk blade attachments using this technology. The EPRI guide Ultrasonic Inspection of Steam Turbine Blade Roots, 1011680, [55] details the research findings, transducer designs, application methodology, and observed detection and sizing performance for the inspection of fir-tree root designs of large curved axial-entry configurations. Axial-Entry Blade Attachment NDE Performance Demonstration, 1011677, [56] consists of flaw detection and sizing performance of commercial inspection providers, using a blind test approach similar to that being undertaken currently for straddle-mount blade attachment designs. The final technical report provides test methodology and inspection performance results and comparisons, including guidance for specifying inspection services. 5-27
Turbine-Generator Condition Assessment
5.8.7 Inspection of Nonmagnetic Generator Retaining Rings Nonmagnetic retaining rings were introduced as an alternative to magnetic rings to minimize heat losses and improve generator efficiency. The predominant alloy, 18% manganese and 5% chromium (18-5), as well as other nonmagnetic alloys, were all found to be susceptible to SCC in the presence of moisture. Although relatively few catastrophic failures of nonmagnetic retaining rings have been reported, the economic consequences of a ring failure are too severe to ignore. Early detection of SCC in nonmagnetic retaining rings by NDE can mitigate such a catastrophe. NDE, however, has proven difficult because of the geometry of the ring and the method of assembling the ring on the rotor. The EPRI report Evaluation of Nonmagnetic Generator Retaining Rings, TR-104209, [35] is recommended as a reference document. It provides a road map to help inspectors evaluate approaches for the various inspection scenarios and detailed technical information for use in developing an effective retaining ring inspection program. Investigation of NDE techniques for retaining ring inspection revealed that no single inspection method or technique alone provides the high reliability required for retaining ring inspection. High reliability can be achieved only by implementing complementary methods or techniques to address the special considerations of a retaining ring inspection. In addition to its assessment of state-of-the-art NDE techniques, this investigation included an assessment of advanced techniques, resulting in recommendation of several for inclusion in retaining ring inspection programs. Retaining ring inspection requires considerable expertise to fully comprehend the capabilities and limitations of the various techniques. Inspections must be tailored to address various inspection situations including in-frame inspections, out-of-frame inspections with rings installed, and out-of-frame inspections with rings removed. Section 8.12 in Volume 2 of these guidelines contains a detailed procedure for removing, inspecting, and reinstalling generator retaining rings. EPRI report TR-102949 gives detailed methods to be used during operations, stand-by, and maintenance to keep generator retaining rings free from exposure to moisture.
5.9
Inspection of Shrunk-On Components
When the turbine experiences any vibration or balance problem, the cause can sometimes be related to the couplings. For this reason, a complete coupling inspection should be performed when balance problems exist. Shrunk-on couplings are used on turbine and generator shafts when it is not possible to use a solid coupling. The checks used for solid couplings should be performed when inspecting shrunk-on couplings. In addition to the solid coupling checks, there are special checks that should be performed when inspecting a shrunk-on coupling. Normal coupling checks are listed in Table 5-6.
5-28
Turbine-Generator Condition Assessment Table 5-6 Coupling Inspections – Disassembly and Reassembly After Disassembly Normal Inspections
Special Inspections
Coupling rim and face run-out
Locking key tightness
Face flatness
Bolt tightness on retainer plates
Visual inspection
Secure staking of retainer bolts
Magnetic particle inspection
Loose fits
Bolt elongation Bolt and hardware damage
Upon Reassembly
Locking plate damage
Face flatness
Windage cover damage
Rim and face run-out
Coupling spacer tightness
Loose fits
Coupling spacer gear tightness
Locking key tightness
Gear and spacer shifting
Retainer plate bolt tightness
Bolt hole scoring or galling
Secure staking of retainer plate bolts
Shrunk-on couplings should not be removed during normal maintenance. If the shrunk-on coupling needs to be removed, the manufacturer should be contacted, and a representative should be on-site to supervise the process. When a shrunk-on coupling is removed, the coupling should be inspected for fretting, dimensional checks, key tightness, keyway damage, and galling or scoring of any components. All retainer plate bolting should be inspected for damage. All bolting and shrunk-on rings should be magnetic particle tested.
5.10 Bearings – Journal and Thrust Types Understanding bearing construction, operation, damage mechanisms, and operational indicators will help the turbine engineer prepare and plan for outage contingencies. Because bearing repairs are typically performed off-site, inspections to document the condition of the bearing and journal should occur early in the outage cycle. Four critical steps are involved in the bearing repair process: 1. Incoming inspection and removal of the old babbitt 2. Liner preparation to receive the new babbitt 3. Casting the babbitt 4. Restoring dimensional integrity and final inspection 5-29
Turbine-Generator Condition Assessment
Each phase of the repair process can affect the quality and serviceability of the repaired bearing. Incomplete incoming inspections can cause delays in the final machining process. Improper preparation of the liner and improper tinning can result in disbonding of the babbitt. Improperly cast and cooled bearings can cause porosity, segregation, and disbonding. Improper preparation and lack of fusion can cause separation and dislodging during a minor repair. Improper machining can cause complete spin casting of the bearing again. Journal bearings and thrust bearings are the two types of bearings generally found in turbinegenerators. Journal bearings confine the rotor in the radial direction, support the rotor weight, and are usually located at the ends of each rotor. Some designs share a common bearing between rotors. The term journal is associated with the area of the rotor that is encircled by the bearing. While the journal bearing confines the rotor in the radial direction, it also allows the rotor to spin safely at high speeds on a wedge of oil. The bearing is lined with a soft material, usually tin-based babbitt. The babbitt can act as a sacrificial material if the harder rotor comes into contact with the bearing surface, and it also can absorb foreign particles too large to pass between the rotor journal and bearing babbitt without causing significant damage to the journal surface. Turbine-generator journal bearings are usually constructed in two halves that are split at the horizontal centerline. The halves are bolted together and use dowel pins in the horizontal joint to ensure alignment of the bearing halves. Journal bearing construction designs are usually tilting pad or elliptical. Tilting pad designs use either single or double tilting pads; the heaviest loaded pads are located in the bottom half of the bearing. The first step in forming the ellipse in an elliptical journal bearing is machining a cylindrical bore through the bearing with shims in the horizontal split line. The elliptical form is obtained by removing the shims when the bearing is assembled. The elliptical bearing has a primary loading zone in the bottom half of the bearing and a secondary zone in the upper half. The elliptical design with overshot grooves in the upper half passes more oil than a plain cylindrical bearing of comparable size; therefore, it should have fewer heating problems than a plain cylindrical bearing. Some elliptical designs are reduced in width and are designated as shortened elliptical bearings. Shown in Figure 5-3, elliptical and shortened elliptical bearings are a two-part construction. The inner bearing component, also known as the liner, contains the babbitt on the inside diameter and the outside is machined to a convex sphere also known as the ball. The outer component is the ring, and its inside diameter is machined to a concave ball seat.
5-30
Turbine-Generator Condition Assessment
Figure 5-3 Elliptical Bearing Construction
Turbine and generator rotors “sag” from the rotor weight distributed between the bearings. The ball seat allows alignment of the bearing assembly in the turbine pedestal to the sag geometry of the rotor journal. In an unbolted state, the upper ring has “ears” with a clearance between the ears and the horizontal split centerline of the bearing. After the bearing is aligned, the upper ring hold-down bolts in the ears are tightened. The tightening of the upper ring “pinches” the liner so that the bearing is locked in place. The amount of pinch is determined by the deflection of the “ears” as the ring hold down bolts are tightened. A thrust bearing absorbs and limits the axial movement or operating thrust of the rotor. The rotor contains integral collars that restrain the rotor to the limits imposed on it by the thrust bearing and the thrust bearing housing. Tilting pad and taper land thrust bearings are used in turbine / generators. Tilting pads come in a number of designs, but usually rely on six pads, three within each half of the bearing. The pads fit under lips machined in the bearing casing to prevent the pads from moving in a radial direction. Locating pins through the bearing housing into the pads prevent circumferential movement. To prevent seizing and to allow free tilting of the pads, the protruding pins have a smaller diameter than the holes in the pads. The tilting of each pad takes place between the back radius of the pads and the inner bore diameter of the casing in the region of the pins on a line contact. By making the back radius of the pads smaller than that of the housing produces the line contact. Tilting pads are self-aligning and therefore do not require bearing casing adjustment features like a ball seat of an elliptical bearing. Lubricant is provided to the bearing by flooding the bearing casing. Seal and drain orifices control lubricant flow out of the casing. Higher bearing temperatures are found on tilting pad thrust bearings because: •
Rotating journal forces oil away from the entrance spaces of the bearing
•
Constant churning of the oil in the flooded casing
•
Increased turbulence 5-31
Turbine-Generator Condition Assessment
Therefore, tilting pad thrust bearings have larger oil requirements and higher power losses. On the positive side, tilting pad bearings are an effective corrective measure for shaft instabilities. The increased bearing loads provide enhanced stability at light thrust bearing loads. The tapered land thrust bearing is comprised of two stationary plates that react with two collars on the rotor. The thrust plates are contained in a casing and are aligned to the rotating thrust collars. The surface of the thrust plates is babbitted and contoured to assist in the formation of the oil wedge between the rotating and stationary components. Oil is forced into the bearing and radial feed grooves provided between each of the contoured surfaces allow oil to be fed both at the base of the thrust plate and outwardly. The outer edge of the groove is damned to maintain a positive pressure within the bearing. Orifices in the oil feed pipes control the amount of oil entering each end of the bearing. The outer diameter of the assembly is spherically machined like an elliptical bearing to allow the thrust bearing assembly to be aligned to the rotor thrust collars. The thrust plates are pinned to the bearing assembly preventing rotation. Thrust transfer and axial movement is restricted by a tongue and groove machined between the bearing assembly and the standard. Table 5-7 lists a number of bearing conditions and damage mechanisms that may be found during an outage inspection.
5-32
Turbine-Generator Condition Assessment Table 5-7 Recommended Action for Bearing Damage Typically Found at Inspection Bearing Condition
Outage Activity
Remedial Action
Possible Consequence
Abrasion
Depending on severity:
Investigate lube oil and remove the contaminant and its source.
Scored rotor journal
Hand scrape Re-babbitt Corrosion
Re-babbitt
Investigate lube oil and remove the contaminant and its source.
Corroded rotor journal
Disbonding
UT inspection
Investigate the previous repair process.
Babbitt separation from liner and bearing wipe
Re-babbitt if unbonded area exceeds specifications Disbonding edge
PT inspection
Electrolysis
Depending on severity:
Re-babbitt if unbonded area exceeds specifications Hand scrape
Hot bearing Investigate the previous repair process.
Investigate the current source and inspect the shaft grounding system.
Minor repair
Hot bearing Scored rotor journal
Re-babbitt Excessive clearance
Babbitt separation from liner and bearing wipe
Bearing stability
Re-babbitt for elliptical Measure and re-shim for tilt pad Re-babbitt if required
Fatigue
Depending on severity:
Babbitt separation from liner and bearing wipe
Minor repair
Scored rotor journal
Re-babbitt Lead contamination
Re-babbitt
Investigate the previous repair process/vendor.
Babbitt separation from liner and bearing wipe
Wiping
Depending on severity:
Any one or more combinations of the above conditions
Hot bearing
Hand scrape
Scored rotor journal
Minor repair Re-babbitt Note: Any breakdown of the adhesion bond between the babbitt and the backing in the lower bearing half can begin to impact heat transfer and subsequent cooling through the bearing.
5-33
Turbine-Generator Condition Assessment
Pre-outage preparations that can be completed in anticipation of outage repairs are listed in Table 5-8. Table 5-8 Recommended Pre-Outage Preparations for Bearings System
Action
Journal bearing
Repair procedure prepared, minor repair - local tungsten inert gas (TIG) repair, major repair - re-babbitt, ball seat, qualified vendors selected, purchasing documents and process ready
Tilt pad bearing
Repair procedure prepared, minor repair - local TIG repair, major repair - rebabbitt all pads, qualified vendors selected, replacement pads, replacement hardware, purchasing documents and process ready
Journal repair
Repair procedure prepared, qualified vendors selected, purchasing documents and process ready
Thrust bearing
Repair procedure prepared (if parts unavailable), major repair - re-babbitt, qualified vendors selected, replacement parts, shims, thermocouples, thrust plates, purchasing documents and process ready
Proximity probes and data collection support software are excellent diagnostic and preparation tools. Unusual bearing conditions and circumstances can be “red flagged” before an outage, giving an early warning for inspection and preparation. Existing bearing thermocouples can provide limited information on a bearing condition. A bearing running hotter than normal may be the result of an operational change or the result of in-service wiping of the bearing. Babbitt can dislodge and limit journal movement within the bearing, causing increased loading and temperature. Unexpected reduction in the temperature of a hot bearing may be the result of a wipe or restriction clearing and the journal returning to normal position. A bearing improperly installed during an outage may provide operational information before the telltale wear pattern in the bearing is observed. Consequently, it is important for the turbine engineer to be aware of the bearing operating conditions as a forecasting tool for outage planning and preparation.
5.11 Stationary Components Figure 5-4 shows a typical diaphragm. The three major areas of a stationary system are listed in Table 5-9.
5-34
Turbine-Generator Condition Assessment
Figure 5-4 Diaphragm Construction
Table 5-9 Separate Areas That Form a Stationary System Steam Path
Structure or Body
Seals
Partitions
Ring and its components
Spill strips (removable, integral, insert)
Inner sidewall
Web and its components
Packing (addressed further in Section 5.13)
Outer sidewall
Each area is inspected during the outage to determine the exact repair scopes, but pre-outage planning can be accomplished by understanding the repair histories for each stage, wear, and damage mechanisms. Before the outage, the following should take place: •
The repair procedures are prepared.
•
Qualified vendors are selected.
•
Purchasing documents and process are ready. 5-35
Turbine-Generator Condition Assessment
Previous outage reports and repair recommendations should be the first in line of information used to create the work scope for diaphragm and nozzle repairs. That information can be confirmed or augmented by a review of each stage and the possible damage mechanisms associated with the stage. The following information is a review of damage mechanisms typically associated with diaphragms and nozzles: •
Solid particle erosion (SPE) is the result of very small particles having a very large economic effect on repairs and loss of turbine efficiency. The source of solid particle erosion is exfoliation of oxides from the inner surfaces of boiler tubes and steam piping. The oxide grows as these systems operate at elevated temperatures and the exfoliation occurs during lower loads. For a comprehensive study, see Reducing Solid Particle Erosion Damage in Large Steam Turbines [36]. The particles are carried to the turbine and do damage when the right flow and particle size distribution exists. The action is comparable to grit blasting the turbine components. The damage is most severe in supercritical units. Both diaphragm pressure side and suction side SPE can occur within the turbine. The primary effort to reduce the effects of SPE has been to combat the effects of SPE by coatings, increased set back, and partition re-designs. Damage from SPE tends to be limited to the HP section and first stages of the IP sections.
•
Water erosion caused by water droplets in the saturated steam is found in the wetter sections of the LP turbines. Component joints that have leakage paths can be heavily damaged from water erosion. Bucket tips, leading edges, and root trailing edges of the last stage buckets are also areas where water erosion is prevalent.
•
Foreign object damage (FOD) results when foreign objects either left within the steam path or introduced in the steam path come into contact with rotating and stationary components. FOD can result from weld slag or a tool left within the turbine; even turbine components themselves that come loose in service can cause dings, dents, and major damage. FOD can be found in any stage.
•
Deposits in the steam path are caused from boiler carry over, water treatment chemicals, and metals that can plate out on turbine buckets and diaphragm partitions. These deposits can block discharge areas causing a reduction of stage efficiency or creating initiation sites for corrosion damage. Stage conditions need to match the right steam conditions for the deposition of material to occur. The main corrosion mechanisms occurring in a turbine are corrosion fatigue (CF), stress corrosion cracking (SCC), and erosion-corrosion (EC). Pitting (P) and CF of blades and SCC of discs tend to dominate as the costliest problems. Corrosion in a turbine is complex and related to the environment including steam purity, moisture evaporation, crevice concentrations, oxides, pH, velocity, turbulence, “lay-up” duration and conditions, stress, stress concentrations and material properties. Metallic deposits can accumulate in the HP section, but corrosion deposits and P, CF, and SCC tend to occur in the latter stages of the LP section.
5-36
Turbine-Generator Condition Assessment
•
Cracking in the steam path is the start of breakage and the start of a component failure. Location, service environment, service requirements, etc. are evaluated when crack indications are found in steam path components. Run-repair-replace decisions for crack indications can occur anywhere in the steam path. Each turbine section is significant because of the potential problems that may result.
•
Component distortion is a common condition in the hotter turbine sections. Time, temperature, and stress on the steam path components influence the amount of distortion.
Table 5-10 presents a matrix that provides a review of damage mechanisms and corresponding planning options for each diaphragm area. Table 5-10 Recommended Action for Diaphragm Damage Typically Found at Inspection Location
Item
Discussion
Planning
Rings or webs
Centering pin fits
Some damage should be expected.
Include as a repair in the diaphragm repair procedure.
Rings or webs
Corrosion
Usually occurs in the LP section.
Rings or webs
Crush pins
Some repair/fitting will be required during outage, as a result of expected wear or physical damage.
Include as a repair in the diaphragm repair procedure for weld buildup
Fitting requires measuring the corresponding location in the shell and machining it to fit.
-or-
Repair includes welding and refitting for the clearance to the centering pin.
Repair information (material identification and work package) will be available for on-site utility repair
Rings or webs
Distortion dishing
HP and IP sections. This develops over time and should be anticipated.
Include as a repair option in the diaphragm repair procedure. It will require repair and re-machining.
Rings or webs
Distortion out of round
HP and IP sections. This develops over time and should be anticipated.
Include as a repair option in the diaphragm repair procedure. It will require repair and re-machining and may require replacement packing.
Rings or webs
Dowel pins
There is a low probability of in-service or disassembly damage.
Repair information (material identification and work package) are available for on-site utility repair.
The dowel pin requires removal if the horizontal joint is repaired. Rings or webs
Hook fit packing
There is a low probability that a repair will be required unless associated with dishing or out of round condition.
Include as a repair option in the diaphragm repair procedure. Machine support is required after repair buildup. Include in the pre-outage, development work package for spill strip hook fit configurations.
5-37
Turbine-Generator Condition Assessment Table 5-10 (cont.) Recommended Action for Diaphragm Damage Typically Found at Inspection Location
Item
Discussion
Planning
Structure
Hook fit Spill strip
SPE damage usually occurs in HP and IP sections.
Include as a repair option in the diaphragm repair procedure. Machine support is required after repair buildup. Pre-outage, development work package for spill strip hook fit configurations.
Rings or webs
Horizontal joint
Damage such as seal weld cracks, erosion, damaged keyways, and excessive opening can be anticipated.
Include as a repair in the diaphragm repair procedure.
Rings or webs
Horizontal joint bolts
Damage could possibly occur during disassembly.
Mixed sizes are available as spares.
Rings or webs
Horizontal joint keys
There is a high probability that higher temperature/pressure stages will require replacement.
Repair information (material identification and work package) is available for on-site utility repair.
This will require removal if the horizontal joint is repaired. Rings or webs
Rings or webs
Horizontal joint miscellaneou s hardware
There is a high probability that higher temperature/pressure stages will require replacement.
Horizontal joint threaded holes
Damage could possibly occur during disassembly.
Repair information (material identification and work package) is available for on-site utility repair.
This will require removal if the horizontal joint is repaired. Include as a repair option in the diaphragm repair procedure. -orRepair information (material identification and work package) is available for on-site utility repair.
Rings or webs
Rings or webs
5-38
Steam seal face and inserts
Support bars
Typical damage occurs from excessive movement and SPE. Significant damage may occur in the HP section from fretting, galling, and SPE.
Steam seal faces can be touched up and hand dressed if the damage is limited. Repair information (material identification and work package) is available for on-site utility repair.
There is a low probability of in-service mechanical damage, erosion, or corrosion.
Drilling and re-tapping if any work is done to high temperature diaphragms.
This may require disassembly for diaphragm alignment.
Replacement bars/shims should be stocked or material and machining process should be available during an outage.
Turbine-Generator Condition Assessment Table 5-10 (cont.) Recommended Action for Diaphragm Damage Typically Found at Inspection Location
Item
Discussion
Planning
Rings or webs
Welds
Although a low probability, some of the original factory welds may become distressed after time and require repair.
Pre-outage activity may be limited to understanding diaphragm construction and an awareness of repair options.
Repair may require mechanical securing and then welding. The amount of weld deposited may require stress relief. Seals
Spill strips removable
There is a high probability that replacements will be required for HP and IP section diaphragms.
A replacement can usually be ordered and received during an outage.
Damage includes erosion, corrosion (LP section), cracks, and FOD.
This requires early identification and correct part identification.
Approximately 50% of the heat rate loss associated with diaphragms is spill strip leakage losses. Seals
Spill strips integral
Located in the HP and IP sections. Damage of these strips includes erosion, cracks, and FOD.
Since this repair requires weld buildup and machining, early outage identification is helpful. Pre-outage planning includes work package development.
Seals
Spill strips insert
Located in the HP and IP sections. Damage of these strips includes erosion, cracks, and FOD.
Since this repair requires weld buildup and machining, early outage identification is helpful. Pre-outage planning includes work package development and part resource identification.
Steam path
Partition & sidewall damage
Surface damage includes: FOD, DPE, thinning (erosion), corrosion, and general in-service partition damage Generally, SPE occurs in HP and IP sections, and water erosion occurs in the LP section. Approximately one-third of heat rate loss associated with diaphragms is due to the surface condition of the partition. Trailing edge cracks are caused by differential cooling between heavier and thinner sections.
Include in the repair procedure. The general planning work scope should come from previous outage reports and recommendations. Pre-outage preparation should include compilation of diaphragm partition and machining dimensions. Plans to obtain data during the outage should be considered if the information is not available.
Fillet welds are structural in nature and weld cracks may result from differential thermal stress.
5-39
Turbine-Generator Condition Assessment Table 5-10 (cont.) Recommended Action for Diaphragm Damage Typically Found at Inspection Location
Item
Discussion
Planning
Steam path
Partial partitions at joints
The pins supporting these partial partitions may be damaged with the same mechanisms as full partitions.
Include in the repair procedure.
Steam path
Bridges
Bridges are structural in nature.
Include in the repair procedure.
Cracks and physical damage are typical discrepancies found.
5.12 Buckets/Blades Turbine buckets/blades are comprised of three basic sections: attachment, vane, and tip area (to include the bucket/blade area). The attachment of the buckets/blades to the rotor may be the same design and size for any number of stages on a rotor. The method in which the final bucket/blade is attached to close or complete a row may vary. The vane section and the tip configuration may be the distinguishing difference between rows of buckets/blades on a rotor. Vane sections are a combination of impulse and reaction designs. Most vane designs will differ from stage to stage restricting the interchangeability of stage from one turbine to another. An exception may be the L-0 and L-1 rows. Turbine stages are normally numbered with the first stage starting as the steam enters the turbine and continues in sequence until the last stage and steam exiting the turbine. An alternate method of identification usually used in the LP section of the turbine is to begin with the last row of buckets/blades and identify it as either the “L” stage or “L-0” stage, and then identify each preceding row in reference to the last stage. The next-to-last stage becomes “L-1,” then “L-2,” and so on as the reference moves against the steam flow. This is a useful convention because the L-0 and L-1 groupings share the same respective designs, operating conditions, and problems. Although turbines vary in number of stages, they can readily be identified and grouped according to last stage designation. Previous outage reports and repair recommendations are the first place to look for an outage action plan for bucket/blade replacements. External inspection access locations may also provide condition information before major outages. Understanding steam path damage mechanisms and having replacement component design information available prior to the outage should provide an edge against outage delays if the unexpected inspection finding occurs. Manufacturing of the buckets/blades would still be required if an outage inspection finding dictated replacements. It may not be practical to have design information created for all stages, but instead, a combination of complete design for priority stages, attachment details for shared attachments, and bucket cover and tip information will all help reduce the time required during an outage to make decisions and obtain replacements. Understanding bucket damage mechanisms and 5-40
Turbine-Generator Condition Assessment
sensitive stages will help in assessing which stages should have a design pool of information available. The following is a brief review of damage mechanisms typically associated with buckets/blades. Fatigue is widely classified as either high cycle or low cycle. High-cycle fatigue is associated with a high local stress in a moderate dynamic stress. Initiation may take a long time, but time to failure may be short after a crack begins. Growth may stop and restart as the dynamic stimulus is removed. Such intermittence may be the result of changes in operating conditions such as excessive condenser back pressure or partial arc operation. Low-cycle fatigue is associated with fewer cycles to failure and higher alternating strain ranges. Fewer cycles are required to initiate a crack and fewer cycles to propagate a crack than in highcycle fatigue. Often, fatigue is associated with some other initiating mechanism to create the stress riser. Stress concentrations may be formed as a result of: • • • • •
A geometrical change in the original design Created during the manufacturing process A maintenance process A heating or cooling process either during operation (for example, a rub) or repair A chemical attack such as pitting
Corrosion-assisted bucket/blade failures typically occur in the attachment area, but they are not limited to the attachment. Observing deposit buildup patterns before a rotor is grit blasted helps to identify possible locations of corrosion-assisted problems. Corrosion-assisted failures can also occur in cover/shroud areas, tie wire, or dampening attachment areas; anywhere that deposit buildup can occur. The presence of a corrosive environment can affect the material’s fatigue strength and, therefore, degrade the ability of the component to withstand the steady and dynamic stresses imposed on it. The bucket/blade and rotor material endurance limit can be significantly affected when a rotor is out of the turbine during an outage. Rotors left unprotected in moist outdoor environments can have local area corrosion accelerated. Physical damage to the exposed area of the bucket assembly occurs from three primary mechanisms: •
Erosion. Erosion within the steam path typically will take two forms: solid particle erosion (SPE) and water erosion. SPE damage is associated with the HP and IP section, and water erosion takes place on the outer-half leading edges of later LP stages. Covers and tenons are susceptible to both types of erosion. A general thinning of the vane section can also occur. SPE and water erosion tend to remove gross quantities of material and have very noticeable results; thinning is a gradual loss of material and is noticeable at the trailing edges.
•
Foreign object damage. Foreign object damage (FOD) is akin to erosion damage in that the rotating element is striking objects. The difference is the size of the objects. FOD damage can occur quickly and severely.
•
Rubs. Rubs occur in a radial or axial direction. Radial rubs can smear tenon and cover material or groove the root area of the bucket from the root spill strip. Axial rubs can also damage covers and root areas of the buckets. 5-41
Turbine-Generator Condition Assessment
Table 5-11 associates typical damage mechanisms to turbine sections. Table 5-11 Blade Damage Typically Found at Inspection Type
Attachment
Vane
Tip Area
Fatigue cracks
Notch keys Dovetail pins
Trailing edges Tie wire holes
Cover lifting Covers – predominately axial toward tenon
LCF
Notch area lifting
HCF
Notch area lifting Root radii
Creep
High temp. stages
Tie wire holes
Corrosion Pitting
Especially L-2 through L-0
Especially L-2 through L-0
SCC
Especially L-1
Especially L-1
Physical SPE
HP & IP sections
Tenons HP & IP Sections
Water erosion
L-1 through L-0 Leading edge Outer radius
L-3 through L-0 Covers Tenons
L-0 Trailing edge Inner radius Thinning FOD
Anywhere along vane section
Cover & tenon
Rub
Predominately radial. Concern if severe on: First HP & IP stages Last LP stages Evaluate axial
Predominately radial. Concern if severe on: First HP & IP stages Last LP stages Evaluate axial
Other
Fretting
EPRI report State-of-the-Art Weld Repair Technology for Rotating Components: Volume 2: Repair of Steam Turbine Blading, TR-107021-V2, provides weld repair details for turbine blading airfoils, erosion shields, tenons, and cover bands. Volume 2 of these guidelines, in Sections 5.2 and 5.3, also provides detailed procedures for blading tenon, tie-wire hole, and erosion shield repairs. 5-42
Turbine-Generator Condition Assessment
5.13 Rotors 5.13.1 Causes of Rotor Bowing First on the rotor problem list is rotor bowing. The following are causes for rotor bowing and are listed from the most common to the least common. 1. Severe rubbing 2. Water induction 3. Metallurgical •
Non-uniform material properties
•
Non-uniform residual stresses
5.13.1.1
Severe Rubbing
Rotor bowing caused by rubbing is a result of non-uniform local yielding and residual stresses. Rubbing occurs when the rotor body comes in contact with a stationary component such as oil deflectors, interstage packing, or end packing. The rub is usually a secondary effect of some other machine condition such as overly tight clearances or excessive vibration of the system. Rubs are usually transient in nature. For example, a rub on the interstage packing may eventually remove enough material to become a lighter rub and eventually eliminate itself. Rubs can be in either the radial or axial direction. A full annular type of rub would occur when the rotor is in continual contact with a stationary component. A partial contact rub is the more common of the two and can be either a single-point rub with the stationary component or a multi-point rub where contact is at multiple locations When the rub is a combination of impact and rubbing friction on the stationary component, the impact usually creates a secondary effect in the form of a rebounding motion. This type of rub will cause circumferential temperature gradients from the friction that is created and is usually more severe on one side of the rotor. If the rotor bows, it will do so gradually toward the rub or “high spot.” This is as a result of the fact that the hot area expands and yields in compression, which causes the rotor to bow toward the rub. A simplified sketch of the orbital motion of a rotor is shown in Figure 5-5 [37]. Note that the high spot always faces out. The high spot comes in contact with the stationary component, causing the rub.
5-43
Turbine-Generator Condition Assessment
Figure 5-5 Mechanics Describing Rubbing Process
The following is the sequence of a rub at constant rotational speed below the rotor first resonance speed: 1. Due to the rub, the shaft bows in the direction of the high spot and a new unbalance force occurs. 2. The original and rub-related unbalances add together, producing an “effective” unbalance force. 3. At a constant rotational speed, the phase lag between the effective force and response (high spot) is constant. This means the shaft has to rotate, yielding a new high spot. The magnitude of the rotor orbit also increases as the center of mass moves away from the center of rotation. This causes greater unbalance, causing the rotor to rub harder. 5-44
Turbine-Generator Condition Assessment
Severe vibration effects may occur during constant speed while the rotor is operated close to or below its first critical speed. Some turbines use vibration detection equipment where the response of the pickup falls off rapidly below 800 rpm and is virtually ineffective below 500 rpm. Therefore, it is difficult to detect potentially destructive levels of self-accelerating vibration from rubbing below 800 rpm when the vibration equipment detection capability is limited. Operation at a constant speed below the vibration equipment detection level should be avoided. Vibration levels for slow-speed balanced modern HP and IP rotors at 1000 rpm should be nearly zero. A 5 mil (0.127 mm) vibration should be unusually high for those rotors at 1000 rpm and may be indicative of a rub. Two conclusions result from the above: •
Operating turbine rotors at a constant speed below vibration equipment detection levels should be avoided.
•
Unusual vibration at slow speeds may indicate packing rubs, and the turbine should be returned to turning gear to “straighten out” the rotor. The rotor should not be operated at low speed to clear the rub.
Rubs occurring above the first critical speed will cause the center of mass to move toward the center of rotation and decrease the magnitude of unbalance. Very hard rubs above the first critical speed are required to excite the rotor. The effects of in-service rubs are reduced if the unit is operating at high load and high steam flow; the steam flow tends to cool the hot spot, reducing its effect. Having “pushed” through a rub to above the first critical speed is not insurance that the rub has cleared; it will probably be experienced again on a future coast-down. In most cases, the machining action of a rub will eventually provide the increased clearance for smooth operation. The exception to increasing clearances by rubbing is associated with non-metallic materials such as textolite and some packing designs that use a slant-tooth design. Slant-tooth design uses a different material, tends to flex out of the way instead of rub away, and increases its crosssection as it is rubbed. The consequences of increasing clearances by rubbing are: •
Reducing sealing efficiency at the rub location
•
Damage to the stationary component
•
Damage to the rotating component
It is important to minimize rotor deflections during the rubbing process to minimize the clearance increase caused by rubbing. The efficiency of the seal will be reduced because of the damage to the sealing profile, but reducing the amount of clearance opening can minimize the consequences. Several seal designs are currently available to withstand the effects of rotor deflections during operation. The rotor maintains a permanent bow if the overheating caused by the rub is severe enough that sufficient area is heated that has expanded and yields in compression. This helps move the rotor toward the rub. Then, as the rotor normalizes in temperature, the material in tension tends to bow the rotor in the opposite direction. 5-45
Turbine-Generator Condition Assessment
5.13.1.2
Bows Caused by Water Induction
Rotor bows caused by water induction result from a hot rotor being quenched so that it yields in tension, thereby causing the rotor to bow away from the cooled spot. As the rotor normalizes in temperature, the yielded material tends to bow the rotor in the opposite direction. Water induction into a turbine section can also cause distortion of shells, which “hump” into the rotor causing severe rubs. 5.13.1.3
Bows Caused by Metallurgical Problems
The least common cause of bowing is from metallurgical problems in the rotors. The cases are few but documented. Metallurgical bows are not well understood, but they are usually associated with high-temperature rotors and may be the result of non-uniform yield strength and material creep properties. Turbine manufacturers changed the heat treatment requirements for turbine rotor forgings in the early 1960s to control impurities and provide more uniform material properties. The progression of the bow is slow and, therefore, corrective action can easily be planned. 5.13.1.4
Corrective Actions
The following actions are available for bowed rotors: •
Balance to offset unbalance
•
Straighten mechanically or thermally
•
Heat lathe to straighten
•
Re-machine new compromised center of mass
Rotor run outs should be taken at each outage inspection on high-temperature rotors. Historical readings will help to plot any rotor bow and predict a growth rate. Run-outs should be taken early in the outage to provide a warning for any unexpected condition. In-service balancing information will also provide a warning of any changing condition. Requirements for in-service mid-span or static balance shots may be indicative of an increasing rotor bow. Rotors are manufactured to a run-out tolerance of one mil (0.0254 mm). Anything in the field that is 3 mils (0.0762 mm) TIR or less is considered acceptable and is only slightly bowed. Rotors with 6–7 mils (0.152–0.177 mm) TIR of bowing (between bearings) are considered to have a minor bow that should be correctable with slow-speed balancing. Rotors with 10 mil (0.254 mm) TIR or greater are considered severely bowed, and corrective action should be applied. The corrective action for a severely bowed rotor may be a combination of heat lathe straightening and machining. The heat lathe may remove a portion of the run-out ranging from 25% to 33%. The heat lathe also reduces residual stresses created during the bowing process.
5-46
Turbine-Generator Condition Assessment
A machining plan is developed after taking a comprehensive set of run-out readings. A new center is determined for the rotor. New diaphragm packing, end packing, coupling bolting, and resized journal bearings are the typical components required to install a re-machined rotor. 5.13.2 Other External Rotor Problems Other types of external rotor problems include: •
Seal area damage associated with rotor rubbing
•
Corrosion - seal areas/rotor body, especially the LP/bucket/blade attachments
•
Attachment creep in high-temperature rotors
•
Journal scoring
•
Oil deflector area scoring
•
Coupling bolt hole damage
Except for the possibly of journal damage, the external rotor damage would be determined during an outage. Damage and its extent are determined by visual inspection, NDE replication, MT, and UT during the outage. UT can be used to inspect rotor areas that are not available to direct sight inspection, such as the rotor dovetail with the bucket/blade installed. The type of repair needed would normally be machining of the damaged area and redesigning the corresponding component. An extreme repair would include a weld repair for the damaged area. Possible rotor body problems include cracks. Rotor cracks have been detected by using a vibration detection system that uses proximity probes. The information provided by the proximity probes, displayed in cascade, polar, 2X amplitude phase time (APHT) plots, and 2X orbits associated with APHT plots, are used to evaluate the rotor condition. The frequency of inservice rotor cracking is increasing because rotors are running longer between inspections and more units are being called upon for peaking duty. A transverse crack occurs most frequently with a slight variation of a cup shape occurring under a shrink fit and has the appearance of a tension fracture in ductile material. Torsional cracks appear less frequently than transverse cracks and are readily identifiable. Longitudinal and transverse symmetrical cracks are rare. The following are a few things to remember about cracked shaft detection: •
A cracked shaft is a bowed shaft.
•
A rotor system with an asymmetric shaft and a radial side-load force rotating at a speed near half of any resonant frequency may experience high 2X vibration amplitude and 2X phase shift.
•
In roughly 75% of the cases, the 2X component does not occur at operating speed.
Taking torsional coupled with lateral vibration measurements will assist in vibration analysis for rotor crack detection. 5-47
Turbine-Generator Condition Assessment
As a result of two generator shaft cracking events experienced at a large nuclear generating station in 2004, EPRI produced a technical report containing a tutorial on the subject. Generator Rotor Shaft Cracking Management Guide, 1011679, [61] includes a discussion of turbinegenerator shaft torsional dynamic behavior, sources of excitation from the generator or grid, locations of damage and mechanism involved, viability of ultrasonic testing for shafts with shrunk-on couplings, effects of unit uprates, and monitoring schemes.
5.14 Shaft Seals Typical damage that occurs to steam deflectors, interstage (diaphragm) packing, and shaft packing during operation is the following: •
Erosion – solid particle and water
•
Corrosion – chemical attack
•
Cracks – brittle
•
FOD
•
Electrolysis
•
Abrasion
A rub is the other damage that occurs to the rotor body shaft and diaphragm packing in addition to the above list. Since spill strip clearances are greater than rotor body sealing clearances, they do not normally rub. The predominant wear for spill strips is erosion. Each packing strip and each spill strip have a spring pushing on them to keep them in their smallest diameter position. These springs are inspected for freedom of movement and cracking when the sealing component is inspected. Unlike turbine seals, modern generator hydrogen seals made with steel backing and babbitted seal surfaces are designed to float. The seal moves “freely” with the rotating shaft radially but is restrained from rotating with the shaft. The hydrogen seal is made of two sealing rings. Oil is forced between the outer ring (air side) and the inner ring (gas side) at a pressure (typically 4.5 psi [31 kPa]) greater than the generator hydrogen pressure. Oil flows between the constricted space between the seal and the rotor, preventing hydrogen gas from leaking along the rotor. Smaller clearances are found in the hydrogen seals than in any other rotating sealing areas within the turbine-generator. If, for some reason, the rings are prevented from floating, damage to the hydrogen seal can occur by: •
Contact between the shaft and the seal
•
FOD or particles entrained in the oil trying to pass between the tighter clearances
•
Electrolysis
An offset seal will cause “burning” or “varnishing” from heat buildup and insufficient oil flow to cool the seal. Damage to the hydrogen seal will reduce sealing capability and cause hydrogen “consumption” to increase.
5-48
Turbine-Generator Condition Assessment
Some seals or deflectors are assembled to compensate for rotor sag. The seals or deflectors are assembled with equal clearance on each side, but biased to the vertical clearances. The top usually has two-thirds of the clearance, and the bottom portion has one-third of the clearance. For example, if the total seal clearance is 30 mils (0.76 mm): •
Each side is set at 15 mils (0.381 mm) clearance.
•
The top clearance is set at 20 mils (0.51 mm).
•
The bottom clearance set at 10 mils (0.25 mm) clearance.
Contact and damage to these seals or deflectors can occur if they are not assembled properly. Table 5-12 lists typical design clearances for sealing areas. The listed operating clearance is based on experience and falls on the outer tolerance edge of the design clearance. Maintaining design clearances requires that a machine be aligned within tolerances, but this does not account for the operating variables that can increase the sealing clearances. Table 5-12 Typical Seal Design Clearances with Field Tolerances Seal
Design Clearance
Field Tolerance
Operating Clearance
Spill strip
See the clearance diagram provided by the OEM.
± 15 mils (± 0.381 mm)
Root radial
See the clearance diagram provided by the OEM.
± 10 mils (0.254 mm)
Root axial
See the clearance diagram provided by the OEM.
± 10 mils (0.254 mm)
Diaphragm packing
15 mils (0.381 mm)
+10 to -5 mils (+0.254-0.127 mm)
25 mils (0.635 mm)
Steam packing
15 mils (0.381 mm)
+10 to -5 mils (+0.254-0.127 mm)
25 mils (0.635 mm)
Hydrogen seals 3600 rpm, babbitted steel
10 mils [< 20” Ø] (0.25mm)
± 1 mil (± 0.025 mm)
10–15 mils (0.25-0.38 mm)
Hydrogen seals 3600 rpm, babbitted steel
12 mils [20” Ø] (0.30mm)
± 1 mil (± 0.025 mm)
15–20 mils (0.38-0.51 mm)
Hydrogen seals 1800 rpm, babbitted steel
10 mils (0.25 mm)
± 1 mil (± 0.025 mm)
20 mils (max) (0.51 mm max)
Hydrogen seals 3600 rpm, bronze
8 mils (0.20 mm)
+1 –0 mils (+0.025 –0 mm)
Hydrogen seals 1800 rpm, bronze
3.5 mils (0.09 mm)
5-49
Turbine-Generator Condition Assessment Table 5-12 (cont.) Typical Seal Design Clearances with Field Tolerances Seal
Design Clearance
Field Tolerance
Operating Clearance
Air deflectors aluminum Alterrex
50 mils (1.27 mm)
+25 –0 mils (+0.64 – 0 mm)
Oil deflectors 1800/3600 rpm
20 mils [8-15” Ø] (0.51 mm)
+5 –0 mils (+0.13 –0 mm)
2 mils/inch Ø (0.02 mm/cm)
Oil deflectors 1800/3600 rpm
30 mils [15-21” Ø] (0.76 mm)
+5 –0 mils (+0.13 –0 mm)
2 mils/inch Ø (0.02 mm/cm)
Oil deflectors 1800/3600 rpm
40 mils [21-31” Ø] (1.02 mm)
+5 –0 mils (+0.13 –0 mm)
2 mils/inch Ø (0.02 mm/cm)
Oil deflectors 1800/3600 rpm
50 mils [31-40” Ø] (1.27 mm)
+5 –0 mils (+0.13 –0mm)
2 mils/inch Ø (0.02 mm/cm)
Oil deflectors Textolite exciter
25 mils (0.64 mm)
+5 –0 mils (+0.13 –0mm)
Oil deflectors aluminum exciter
15 mils (.38mm)
+5 –0 mils (+0.13 –0mm)
From experience, nominal clearance may be as high as 2 mils per foot (0.0508 mm per 30.48 cm) of rotor span between bearings. This value has been observed from many inspections and is an average row clearance after normal operation. A clearance of 34 mils (0.8636 mm) can be expected for a rotor with 17' (5.18 m) between bearing spans. Packing and spill strips are inspected for radial rubs. The corresponding rotor, bucket, and cover areas should also be inspected for severity of rub. Sharp edges and deep grooves should be addressed and blended. Some spill strip rubs are the result of a corner lifting on a bucket cover. The following is a list of conditions that might be found regarding rubs: •
A consistent rub around the circumference of the rotor may indicate tight clearances or that the rotor was bowed during operation.
•
The diaphragm may be out of round if sealing teeth are rubbed on the top and bottom.
•
Alignment may be the issue if the teeth are rubbed only on the top or bottom. It may be general alignment; it may be a single diaphragm out of alignment. It may be necessary to perform a tops-on/tops-off alignment procedure.
•
Rubbing only on the bottom of the rotor may indicate shell humping from water induction.
5-50
Turbine-Generator Condition Assessment
Axial ledges on wheels and bucket cover edges should be carefully inspected for axial rubs. The corresponding areas of the diaphragms should be inspected. The damaged area may require either hand blending or re-machining, depending on the severity of the rub. The high-low labyrinth teeth of packing and the corresponding rotor location should be inspected for axial rubs. Incorrect packing may be the cause of a single location rub or differential thermal growth. The repair may require re-machining a new labyrinth sealing area. New designs are available to increase both the sealing efficiency and the longevity of the seals. Packing is designed and in service that at no and low loads retracts from the sealing position and provides an addition clearance of approximately 60 mils (1.52 mm). The additional clearance is a benefit during startup when rotor vibrations may be excessive, which may be the case during a startup after an outage. The first roll or two may have higher than desirable vibration levels until the rotor is balanced for operation. EPRI report Evaluation of Replacement Interstage Seals for Turbine Upgrades, 1010214, [57] evaluates the typical economic benefits of improved interstage packing, identifies the various seal and packing design alternatives available, describes their functionality, and describes any O&M issues that have been observed in their use. Evaluation Tool for Cost Effective Steampath Upgrades, 1004565, [58] provides engineers with a basic understanding of the underlying principles of the new advanced designs of replacement steam path components and analysis tools for large turbines. The report will also include guidance on the relative value of various new design features on the overall improvement. The unique design issues associated with fossil and nuclear turbine performance will also be covered, as well as the overall plant cycle issues that are involved in upgrade or uprate decisions.
5.15 Valves 5.15.1 Stop Valves The stop valve primary function is to provide a second line of defense against energy from the boiler failure in the event of a control valve failure. The main stop valve also closes as a routine activity when the turbine is tripped. Some stop valve designs also incorporate a bypass valve that is used as low load control valve. The components that make up typical stop valves are: •
Lid
•
Disk assembly – cap/bypass valve/disk
•
Seat
•
Pressure seal head assembly – pressure seal head/bushings/valve stem
•
Steam strainer
•
Valve actuator – hydraulic components
5-51
Turbine-Generator Condition Assessment
Typical maintenance for stop valves addresses the wear areas of the valve. The primary wear areas of the stop valve assembly for high-pressure turbines are the bypass valve and the disk assembly. The primary wear mechanism is SPE, but water during startup can also do significant damage. Water can accumulate and erode the valve components if the drains that are incorporated in the valves are not functioning before and during startup. The bypass valves are used for full-arc admission during turbine starting. Bypass valves come in a variety of designs and configurations; they can be coated or uncoated; skirted or unskirted; solid stellite, overlay stellite, or no stellite. Upgrades for bypass valves and stems are provided to combat the effects of SPE. The upgrades include redesign of components, diffusion coatings, overlay coatings, material changes, etc. Other potential valve internal wear areas include stem bushings, disks, and seats. Intervals, methods, and details for the valve body (casing) internal inspections are typically recommended by the original equipment manufacturer. The typical internal areas to be inspected after grit blasting are: •
Intersections – valve inlet/outlet/equalizer
•
Connections – valve body/stop valve seat ledge region
•
Stop valve dam-to-casing welds
Internal inspections look for the effects of age and operation, typically displayed as cracks. Some components may need to be modified as operational changes take place (base load to cycling). Some of the internal areas that may require modification are sharp radius interfaces such as the steps in diameters of the lid. External connections of main stem piping should also be included in a comprehensive inspection program. The effects of age, operating temperature, local stresses and operating practices in the right combinations will be reflected as some level of creep damage. It is important to pay appropriate attention to valve casing inspections especially as units extend the intervals between major outage periods. Pre-outage planning should incorporate some level of preparation for crack indications from the outage inspection. The planning could include gathering and collecting the necessary information if a weld repair would be required for a main steam attachment weld. Condition Assessment Technology for Steam Valves, 1010211, [59] is a generic guide that describes methods and procedures for valve disassembly, assessment of condition and wear, and proper reassembly for long-term operation. The guide contains specific procedures for 12 commonly used steam valves in full-speed and half-speed turbines. 5.15.2 Control Valves Control valves provide the first line of defense against turbine overspeed during emergency conditions. The control valves control turbine speed and load by increasing or decreasing steam flow into the HP turbine. A typical control valve consists of: •
Stand assembly – stand/bushings/valve stem/disk
5-52
Turbine-Generator Condition Assessment
•
Seat – seat stellite/seat pins
•
Valve actuator – linkage/hydraulic components
Control valves are exercised continuously during operation; therefore, the greatest wear area is bushings. Bushings support and restrain the valve stem and are also found in each interface of the operating linkage. Control valves require the same internal and external inspection as stop valves, but in addition, the control valve seat stellite inlay and seat pin welds need to be inspected. The critical external weld inspection area is the weld to the main steam leads if the valves are separately mounted. Control valve condition assessment is covered in EPRI report 1010211 mentioned in the previous section. 5.15.3 Reheat Stop Valves The reheat stop valve (RHSV) is similar in maintenance activities to stop valves and serves similar functions. The RHSV provides emergency protection as the second line of defense against turbine overspeed from stored energy in the reheater. The intercept valve (IV) is the first defense. Control circuitry provides trip anticipation and valve actuation for the IV. Normal operation of both valves is wide open. Reheat valves can be either separate or combined into one valve called a combined reheat valve (CRV). Reheat stop valve condition assessment is covered in EPRI report 1010211 mentioned in the previous section. 5.15.4 Non-Return Valves The non-return valves (NRV) are located in the turbine steam extraction system. The turbine extractions provide steam to the unit condensate and feedwater heater systems. The available energy stored in the extraction piping and heaters is often sufficient to significantly contribute to turbine rotating speed during a turbine trip. The NRVs are comparable to a free swing or a power-assisted check valve. In acting as a free swing check valve, the flapper or disk allows flow to enter and freely move through the valve. The disk pivots downward as flow diminishes. The disk will seat if flow is stopped or if a flow reversal occurs. The power assistance is used for valve closure, and it causes the valve to close before flow reversal occurs. An NRV may be equipped with a single-acting spring to close the cylinder connected to the swing arm. Upon loss of signal or a turbine trip, air is removed from the cylinder, allowing the spring to close the valve at low flow and before reversal of flow. The NRV prevents a flow of steam to the turbine that would either cause or contribute to a turbine’s overspeed. The typical NRV shown in Figure 5-6 is not designed to handle reverse water flow, but it may be called upon to act as a line of defense against water induction, even though valve design does not guarantee tight closure. 5-53
Turbine-Generator Condition Assessment
Figure 5-6 Typical Non-Return Valve Construction
Typical maintenance requirements for this type of NRV address the following wear areas: •
Bushings for the swing arm shaft
•
Swing arm shaft
•
Disk and post assembly
•
Seat
•
Seals-double bearing covers-not external closure assistance/soft packing/mechanical seal
•
Operating cylinder
Seat and disk closure condition is important to the operation of the valve. Outage inspections typically reveal some amount of seat damage from disk closure impact or FOD. Often, the damage can be lapped out easily. Other times, either repair to the inlay or replacement of the valve is required. Seat inlays are usually made from stainless steel or a hard-face material like stellite. Sealing condition checks can be performed during non-major outage periods. By removing the lid during non-major outage periods, a visual internal inspection and a tissue paper test between the disk and seat can be performed, providing valve condition and sealing information.
5.16 Casings, Steam Chests, and Nozzle Chests The pressure-containing components of turbine sections are comprised of: •
Outer shells
•
Inner shells
•
Casings
•
Exhaust hoods
•
Nozzle boxes
5-54
Turbine-Generator Condition Assessment
The high-pressure section is comprised of thick-walled, high-quality alloy castings. The intermediate- and low-pressure sections may be made from castings of thinner section thickness, fabrications of welded construction, or combinations of both. The higher temperature materials will generally be a variation of 1.5% chromium, 1% molybdenum, and similar to ASTM standard materials for the same temperature and pressure service. Lower temperature materials may be cast steel or cast iron. Temperature, pressure, and service are considerations for the material selection. Repair welding can generally be done on steel and alloy castings. The appropriate weld procedure, consideration of component geometry, and thermal expansion are essential for serviceable repairs. Multiple shell construction is used as pressure boundaries, diaphragm carriers, and extraction locations between turbine stages. Unit construction and operation induces stress in the components as they are heated and cooled to meet operating conditions. The internal areas respond quicker than the external areas. The turbine rotor responds thermally much quicker than shells and if not warmed or cooled uniformly with the shells may cause axial rubs. The lower temperature section is not as susceptible to temperature stresses as the higher temperature section. Lower temperature sections are relatively lighter in construction and more flexible. However, thermal cycling of a unit has an impact on all the sections. The typical maintenance concerns of turbine casings from operation are: •
Distortion
•
Cracking
•
Erosion
•
Galling
Casing distortion is a common problem associated with service at elevated temperatures. The most common form of distortion is “clam shelling” or “smiling.” The horizontal joint is open toward the center of the section and closed at the outer edges when viewed along the axis of the machine. Another form of distortion is “humping,” where the horizontal joint is open on the ends, closed in the center at the outer flange, and open in the center of the inner flange. Shell distortion is often linked to water induction incidents. The effects of distortion range from a nuisance to disassembly/reassembly problems. Efforts may be made for thermally relieving some of the stresses or re-machining the shell. Planning preparation including resource identification and repair options can be evaluated as preparation for an outage if a section has experienced a water induction incident. Most cracking in high-temperature shells is a result of thermal stresses where stress concentrations, section changes, etc., respond to differential heating and cooling, causing high stresses. A typical repair for cracks is to grind out and blend them. Severe cracks require stitching, weld repair, or other action to restore the integrity of the shell. Shell geometry should be reviewed to reduce stress concentrations as preparation for increased cycling duty. Lowpressure and temperature sections also experience increased cracking from operational changes. Fabricated sections should also be reviewed for stress risers and pre-outage planning accomplished to address repairs and preventive measures. 5-55
Turbine-Generator Condition Assessment
Solid particle erosion occurs in higher pressure and temperature sections while water or steam erosion occurs in lower pressure and temperature sections. Welding and machining in higher temperature sections typically repair the erosion damage. Ductile filler materials are typically chosen for repair. Low-pressure and low-temperature sections can be repaired with material replacement products that can be cast. Pre-outage planning would include identifying resources and materials for repairs. Component assembly face galling may be the result of distortion, oxide buildup, tight clearances, or interference fits. Repair is typically cleaning and removing the torn material but weld repair and re-machining may be appropriate. Other problems that may be encountered while doing shell inspections are: •
Casting defects
•
Welding defects
•
Fabrication defects
Each defect should be evaluated for its impact on the serviceability of the component. Not all defects are harmless and not all are serious. An attempted repair may increase the scope of work without increasing the serviceability of the component. Therefore, considerations should be given to: •
The cause of the indication
•
The stresses in the area
•
Probability of propagation
•
History of the section and indication
•
Possible operation changes that might impact the indication
A brief list of things to consider when reviewing the need for a casing repair is provided in Table 5-13. Table 5-13 Casing Repair Issues
5-56
Item
Relevant Factors
Base Material
Composition, weld ability, wall thickness, minimum wall to retain pressure, allowable stress at temperature, operating stresses
Indication: General
Location (surface, subsurface, proximity changes), geometric, thermal, pressure
Indication: Weld
Geometry, depth, extent, orientation, type
Crack
Pinhole, casting flaw
History
NDE (methods, results), repairs, operating conditions, operation pattern
Repair Options
Mechanical, welding
Turbine-Generator Condition Assessment
5.17 Generator 5.17.1 Classifications Generators are classified not only by their capability but also by their speed (number of poles), type of cooling, number of coolers, type of coolant, coolant flow path, retaining ring attachment, and type of frame construction. Table 5-14 presents examples of combinations of cooling design and retaining rings: Table 5-14 Combined Cooling Designs with Retaining Rings Cooling Design
Retaining Ring Design
Diagonal flow
Body mounted
Radial flow
Body mounted
Conventional cooled
Spindle mounted
Conventional cooled
Body mounted
Direct cooled
Spindle mounted
Direct cooled
Body mounted
Nested within the generator identification are the types of auxiliary systems that support the generator. This identification system identifies additional equipment that may require pre-outage maintenance planning, outage inspection, repairs, etc. The coding may also help identify a “family” of generators that are experiencing very specific maintenance problems and require special planning before and attention during the outage. Generators can be broken down into two major areas: stator and field. The purpose of the generator inspection is to determine the condition of these two areas and evaluate them for continued reliable service. This may mean repairing conditions found during the outage or postponing action until a future outage. To accomplish this, each area will receive both mechanical inspection and electrical testing during the outage. The EPRI report Guide for On-Line Testing and Monitoring of Turbine Generators, 1006861, [38] describes a variety of generator on-line monitors and generator detectors. The guide is in the form of a spreadsheet where a user can specify the generator they are inspecting and then get extensive technical help in the form of a detailed failure mechanism and monitor description. These results can assist with defining the most appropriate on-line detection system. When considering the replacement of the generator rotor or stator, Sections 8 and 9 and Appendices E and F in Volume 4 of these Guidelines will guide the utility engineering and purchasing organizations through the process of procuring a generator stator rewind or an entirely new generator stator.
5-57
Turbine-Generator Condition Assessment
While preparing these procurement specifications, many engineers may not be entirely familiar with the terms and technical implications of these specifications. The following is a list of references that will further the technical understanding of these terms: •
G. Klempner and I. Kerszenbaum. Operation and Maintenance of Large Turbo-Generators. IEEE InterScience Series, Wiley Press, 2005.
•
M. G. Say. Alternating Current Machines. Pitman Publishing, 1978.
•
E. Fitzgerald and C. Kingsley. Electric Machinery. McGraw-Hill, 1971.
•
IEEE Guide for Operation and Maintenance of Turbine Generators. IEEE Std. 67-2005.
•
General Electric Company. Generators for Utility and Industrial Applications. GE Industrial Power Systems, GET-8022, October 1992.
5.17.2 Generator Stator The stator core serves a mechanical and an electrical purpose. The mechanical purpose is to support the stator winding. The electrical purpose is to provide a return path for the lines of magnetic flux induced by the field. A visual inspection of the stator core starts with carefully inspecting for signs of localized heating or damage of the core inner surface. The stator core is comprised of enameled insulated “punchings” that are assembled to a key bar. Overheating of the punchings can occur at the inner diameter (air gap) if enough punchings are damaged and are shorting together. In some cases, as few as two laminations shorted together would mean a repair and restacking of the core would be necessary. Hydrogen-cooled generators require paying particular attention to this when inspecting. An oxide builds up on the overheated area and limits conduction, eventually limiting the overheating in air-cooled generators, but in hydrogen-cooled generators, there is no oxygen to limit the process. Removing the damage with a de-burring tool and re-insulating the punchings can restore these areas. During their inspection, the windings can be tested to determine if they have been affected by thermal aging. Electrical insulation used in stator and rotor windings has a major impact on the reliability of large motors and generators. Insulation failure, whether direct or indirect, will result in machine failure, which leads to forced outages, reduced reliability, increased maintenance, and repair costs. EPRI report Testing of Stator Windings for Thermal Aging: Interim Results, 1004557, [39] deals with this issue by correlating thermal aging of stator coil/bar insulation to dielectric changes measured at frequencies other than 60 Hz. This project has not been completed, but the results so far have shown that measuring the dielectric changes in the test insulation with non-60 Hz techniques can identify aging. This project should be complete in 2003 and can help improve maintenance of stator windings. The punchings are also inspected for movement. Normally, clamping force from the stator clamping flange holds the punchings tightly together. The telltale sign is dusting or “greasing,” which is a dark, spotty accumulation if a punching is moving. Generator construction normally
5-58
Turbine-Generator Condition Assessment
limits access to the backside of the punchings and the attachment to the key bar. This area is inspected from the ends and ventilation areas to look for movement between the punchings and the key bars. Baffles can be either installed at the ends of the core or mounted to the inner end shield. The baffles should be inspected for cracking in the mounting hardware or the supports, vibration, or fretting, especially at any possible locations of contact with the windings. The attachment hardware should all be nonmagnetic. Stator windings pass through slots in the punchings and wedges retain the windings. The wedges are inspected for looseness or movement by using a small ball peen hammer and tapping the wedge, feeling and listening for changes in sound. Some wedge movement and fretting may be the result of slight movement of the core during normal operation. The corrective action will depend on the type of generator insulation and wedges. A numbering convention should be used to identify the wedges. Generally, a numbering convention begins at one end of the generator and numbers in ascending sequence along the length of the slot to the other end. The usual convention is to number from the turbine end to the collector end of the generator. The numbering convention will also include a radial orientation for the slot location. Typically, the radial location begins with a top bar referenced to one of the terminal connections and increases with reference to the rotation of the field. The stator windings have three areas of inspection: the slot portion through the core; the end winding area past the core; and the end windings support system. Borescopes are used to increase the range of the inspection when examining the stator windings. Excessive winding movement and vibration may be detectable by borescope examination through the ventilation ducts. Dusting on the side of the bar may be evidence of undesirable bar movement. Bar action may also cause chafing on the slot armor cutting through the armor. High-current generators are more susceptible to bar movement and vibration because the forces exerted on a bar during operation varies approximately as the square of the current. The type of insulation used in the construction of the bars is also a factor in how much movement and vibration may exist during operation. The end windings should be inspected for signs of relative movement that may be expressed as cracking in the taping. The copper bars expand at a different rate than the iron core; therefore, the copper bars may have a tendency to expand faster and more than the core as generator loading is changed. Insulation changes that better contain the windings have helped to reduce this condition. Inspection and repair requirements may extend beyond the end windings into the ends of the core if cracking conditions are found. Inspection of the end windings also includes looking for any looseness or relative movement between components. The condition of the stator winding insulation system for a turbine-driven generator can be checked by partial discharge (PD) and electromagnetic interference (EMI) on-line testing. These tests can offer advantages in avoiding prolonged generator shutdown for off-line tests and inspections. PD is a time domain measurement, and EMI measures activity with a frequency scan. Both tests evaluate high-frequency currents that flow as a result of electrical discharges 5-59
Turbine-Generator Condition Assessment
occurring within the structure. The EPRI report Assessment of Partial Discharge and Electromagnetic Interference On-Line Testing of Turbine Driven Generator Stator Winding Insulation Systems, 1007742, [40] evaluates the tests and gives an appraisal of their effectiveness. The results show that systems being able to monitor activity over time can add another dimension to the diagnostic process. Broken ties, loose or missing blocks, distortion, or cracking may all be evidence of high-level generator response to faults or system disturbances. These are repaired by securing, reinforcing, and re-taping. The end windings may also display signs of corona. Typically, these would be seen as whitish, brown, or yellowish accumulation of discolorations. Minor corona deposits are typically just cleaned off, but more intense activity may require providing “bridges” to dissipate the surface charge. Many later design generators must deal with stator bar liquid connection problems. Generally, the hose connections are inspected for movement, abrasion, fretting, cracks, or surface contamination. It is important to keep the hose surfaces clean because dirt or other substances on these surfaces can form a creepage path to ground. The corrosion of the brazed bar end connections is an additional problem you must be aware of. This area is inspected using a hydraulic integrity test (HIT). Leakage in this area may result in wet bars, ends, etc. A variety of repair methods are available today. Some repairs include epoxy injections that repair and seal the area from addition corrosion. It has been found that water-cooled generators have clip-to-strand leakage due to localized crevice corrosion as referred to above. EPRI has investigated this problem, and report Conversion to Deaerated Stator Cooling Water in Generators Previously Cooled w/ Aerated Water: Interim Guidelines, 1000069, [41] outlines how generator stator coolant water can be safely and economically converted from an oxygen-rich (aerated) to an oxygen-rich (deaerated) condition. With generator cooling systems, there are problems with recirculation and deposition of solid matter in the cooling water, which can lead to flow restrictions and inadequate cooling of electrical components. Depending on the magnitude of the flow restriction, various malfunctions may occur. One example of these is local overheating of a stator bar above the long-term temperature limit of electrical insulation. Another problem could be the loss of cooling of a stator bar that leads to overheating or melting of the bar. Due to unexpected problems, it is useful to have a process that monitors the system for flow restrictions and to have options for removing them. The EPRI report Guidelines for Detecting and Removing Flow Restrictions for Water-Cooled Stator Windings, 1004704, [42] gives guidance in this area. The guide reviews various approaches for detecting and removing flow restriction in hollow stator bar strands. It provides an appropriate procedure to conduct a visual inspection that is supported by mechanical means in order to provide the most reliable information. The guidelines will provide information relating to maintaining the system during outages and describe improved early monitoring and detection procedures.
5-60
Turbine-Generator Condition Assessment
The bushing box (attached to the lower portion of the generator) should be opened and the highvoltage bushing and stand-off insulators inspected. Both porcelain insulators should be inspected for damage, cracks, movement, and greasing. All hardware should be inspected for tightness and appropriate sealing. Leaking of asphalt material on the airside of the bushing box is an indication of high-voltage bushing overheating. Abrasion or FOD is the typical damage mechanism to the ID of babbitted steel hydrogen seals. Rings severely scored (> 0.005" [0.127 mm] deep) over 75% of the surface area should be replaced. Rings should also be replaced where the design clearance is worn beyond 1.5 times for 3600-rpm machines or two times for 1800-rpm machines. Air tightness test results or hydrogen consumption beyond design or expected values may also indicate excessive clearance. Electrolysis damage may also occur to hydrogen rings. The hydrogen seal casing is fastened to the outer end shield, and typically, the collector end hydrogen seal assembly is insulated from the outer end shield. The insulated hydrogen seal casing is assembled with insulating sleeves around the bolts and insulating washers under the bolt heads; no conduction paths should exist though the assembly. Insulation resistance should be checked with a megger and should be greater than 100,000 ohms. If megger readings are low, the assembly should be disassembled one bolt at a time to determine the possible shorted location. Oil deflectors should be checked for clearance to the shaft. The inner oil deflector should especially be checked if there is oil leakage into the generator. The inner oil deflector may also leak oil into the generator from a previous poor assembly or lack of seal to the bolting face or at the horizontal joint. Generator bearings may experience the same problems and would require the same repair techniques as turbine bearings; both issues are addressed and outlined in Section 5.10. An insulation requirement may be the one difference from the turbine bearings. One bearing, usually the collector end bearing, will be insulated. This bearing should be checked with a 500 volt megger, and a minimum of 100,000 ohms resistance is required Units with a hydrogen cooler should be inspected for signs of vibration damage or breakage to the structural supports strengthening the cooler. Repairs are made by: •
Protecting the tube bundles and fins
•
Taking precautions to reinforce the cooler structure by attaching the cooler to wide flange structural beams or “strong backs” to prevent distortion, sagging, etc. during the repairs
•
Weld repairing or strengthening and then weld repairing the damaged area
•
Recoating the structure
An inspection of the cooler support structure within the generator should be done if repeated cooler damage is observed. The cooler is designed to be equally supported along its entire length. If there are large changes in elevation between the support rails and the cooler openings or if the support structure is not level, the cooler may significantly deflect or vibrate in service causing cracking in the cooler bundle structure. 5-61
Turbine-Generator Condition Assessment
The EPRI report Generator Cooling System Operating Guidelines: Cooling System Maintenance and Performance Guidelines During Start-Up, Operation, and Shutdown, 1004004, [43] provides operating guidelines that apply to systems with either high or low dissolved oxygen (DO) concentrations. This report provides a cooling system maintenance and performance guide for application during startup, operation, and shutdown. An easy way to assess this condition is by stretching a wire between the cooler openings and evaluating the position of the support rails in relation to the cooler openings. The cooler support structure should have less than 1/8" (3.2 mm) deflection between the structure and the wire at any point. Modify the rails’ position or support structure as necessary. Another indication of cooler support structure issues may be reflected in the difficulty of cooler removal or insertion. The cooler may “hang-up” during these activities. The tube and fins of the cooler should be inspected for any signs of leaks. The cooler heads are inspected for corrosion. The cooler heads may require sandblast cleaning and then adding a corrosion-protection coating. The generator produces an electrical potential induced into the generator field during normal operation. Shaft grounding brushes are usually installed near the coupling between the turbine and generator to provide a ground path for the potential difference. Recent preliminary studies have shown that it is not only the generator contributing to the potential difference of the HP turbine to the generator rotor train, but the LP rotors also build up a potential as water droplets are stripped from the last stage buckets. The LP rotors may also retain a magnetic field as they age, induced by any repair welding on the rotor, induction heating, or by the MT inspection performed on the rotor. Residual magnetic fields should be inspected and, if significant, degaussed as an outage activity. Further studies of the turbine are being performed and should shed some light on the total impact and significance of LP rotor contribution to the total rotor train potential. If the potential difference were not grounded, it would seek a ground path though either the oil film of a bearing (or bearings) or the hydrogen seal. Electrolysis of either location would result. It is good maintenance practice to monitor the grounding brushes during normal operation to ensure that they are clean and in contact with the rotor. During an outage, they should be inspected, reassembled, and tested as soon as the unit is returned to service. 5.17.3 Generator Field The field collector rings should be checked for vibration prior to the outage to provide a preview of the collector ring condition. A dowel rod is used between the transducer and the brushes to obtain the collector ring vibration. Choose a brush location in approximately the same angular position as the other vibration pickups on the turbine-generator (yields approximately the same reference location and orientation for vibration equipment). A good operating collector ring and brushes will have a vibration level in the range of 2–3 mils (0.05–0.76 mm); up to 6 mils (0.15 mm) is satisfactory. Readings of 10–20 mils (0.25–0.50 mm) may be experienced when there are problems. Vibrations normally increase as the collector ring wears. The wear is usually seen as peaks and valleys caused by the different brushes. 5-62
Turbine-Generator Condition Assessment
Vibration may be caused by a condition known as “photographing” where the brush bounces and loses contact with the collector ring, resulting in an “image” of the brush on the surface of the collector ring. High vibrations may mean mechanical problems with the rings such as potentially loose parts or misalignments, but often this vibration is caused by the condition of the rings. It may be possible to balance the collector rings, but because the cause of the vibration is the condition of the rings, it is necessary to correct this vibration by machining or grinding the collector rings. Rings can be machined as long as there is sufficient spiral groove depth remaining (plan on replacing rings when this groove in a minimum 1/16” [1.6 mm] depth) after machining the rings to an 8 microinch (0.203 micrometer) finish. Collector rings may either be ground “in the machine” while on turning gear, in a rotor turning device during an outage, or “in the machine” at speed. Each process has its advantages and disadvantages; some of those are listed in Table 5-15. Table 5-15 Alternative Processes for Grinding Collector Rings Operation
Advantage
Disadvantage
Turning gear
Usually non-critical path. Done between getting on turning gear and unit startup at the end of the outage.
May be centrifugal effects on rings.
Rotor turning device
Non-critical path.
May be centrifugal effects on rings. Equipment availability.
At speed
Remove centrifugal effects on rings.
Critical path activity.
A review of brush wear rate may also provide information regarding the condition of the collector rings. Typical brush life is 3–6 months, and the brushes should wear at a rate of approximately 350 mils (8.9 mm) per 1,000 hours of operation. The corresponding wear rates of the collector rings should be 1 mil (0.3 mm) per 1,000 hours of operation. Excessive wear rates indicate poor performance of the brush-ring assembly. A collector ring wear rate of 5 mils (0.13 mm) per 1,000 hours of operation would indicate poor performance. Wear comes from both mechanical abrasion and electrical arcing. Minute electrical arcing begins to occur after the ring is worn a few mils and the brush contact begins to change. The brushes may tend to chatter and chip as brush temperature rises from the loss of contact if the condition worsens. The rings can be visually inspected after the field has been removed. The rings should be inspected for grooving, cutting, uneven wear, and surface coloration. Only minor surface damage from handling may be dressed and blended with a stone. A darkish-brown coloration on the collector rings is an indication of good brush-ring contact and operation. The coloration is from a film of hydroscopic brush material residue that has collected and is adhering to the collector ring surface. The composite film then acts as a lubricant between the brush and the ring, reducing wear. The collector rings sit outboard of the collector end outer end shield, and the field windings are inboard of the collector end inner end shield. There is a connection between the collector rings 5-63
Turbine-Generator Condition Assessment
and the field windings that is made through a conductor that passes through the field bore and then emerges between the fan ring and the field windings under the retaining ring. The bore conductor is connected to the main lead through a field terminal stud shown in Figure 5-7. There is a seal around the terminal stud on hydrogen-cooled units to prevent the hydrogen gas from leaking into the bore and exiting around the collector rings.
Figure 5-7 Terminal Stud Hydrogen Seal Construction
The seal is inspected for tightness as part of the outage inspection. If possible, check the hydrogen seal while the unit is still on turning gear and pressurized with hydrogen using a hydrogen “sniffer.” Certain main lead designs (solid lead, hollow and ventilated) experience a combination of highcycle and low-cycle fatigue in the 90-degree bend where the lead turns away from the field body and radially toward the windings. See Figure 5-8. The problem is a result of centrifugal loading caused by the main lead; the loading is transferred to the lead wedge and the #1 coil. Therefore, damage may also be seen in the lead wedge and the #1 coil end strap.
5-64
Turbine-Generator Condition Assessment
Figure 5-8 Location Susceptible to High-Cycle Fatigue and Low-Cycle Fatigue in Certain Main Lead Designs
These components are visually inspected with a borescope and red dye PT. The end turns can also be inspected with the retaining rings in place for distortion, loose or moved blocking, discoloration, or debris using articulating mirrors and borescopes. The retaining rings must be removed if damage is found. The retaining rings have been identified as the highest stressed component of either the turbine or the generator. Therefore, special care must be taken to prevent adding any stress risers from nicks or dings to the retaining rings. The outside diameter of the retaining ring may be PT or MT inspected with the retaining ring in place. But the ring must be removed to inspect the internal diameter with either PT or MT. UT techniques are being developed to inspect the condition of the inside diameter with the retaining ring in place. Current retaining ring removal practice uses high-frequency induction heating, but older methods include both flame heating and resistance blankets. Flame heating required additional protection to prevent flame impingement on the nonmagnetic retaining ring surface. The field slot wedges can be visually inspected with the retaining rings in place. The wedges are inspected for mechanical damage of nicks, gouges, etc., and electrical damage of overheating. Excessive circulating surface currents cause the overheating. The wedges may require removal if the overheating damage is severe. Good contact between the wedges and body is required to prevent overheating. Arcing and overheating reduce the conduction between the wedge and the body. The retaining rings must then be removed to remove the field slot wedges. The wedges are then removed, glass-bead cleaned, inspected, and reinstalled. Hydrogen-cooled generators use body-mounted fans on each end of the field to circulate hydrogen through the generator. The fans may either be single stage or multiple stages. Typically, single-stage fans are used with liquid-cooled stators, and multiple-stage fans are required for gas-cooled stators. The field-mounted fan segments should be inspected for tilting, gaps between the blade hubs and blades for single-stage fans, proper fan ring clearances, rubbing, and other problems while being 5-65
Turbine-Generator Condition Assessment
removed from the rotor. The segments should be numbered in the order of their removal and referenced to a specific rotor location. The removed fan segments should be inspected using an appropriate NDE method. Older units may have steel segments, but newer units are made from aluminum. The anticipated indications would be cracking or other mechanical damage. The rotor surface at the hydrogen seal areas, oil deflectors, and rotor journals should be inspected for damage, including grooving, scoring, and damage from previous assembly or disassembly. Minor damage may be “strap lapped,” but severe damage will require machining of the surface. Strap lapping cleans and polishes the rotor surfaces. The grit of a coarse grit emery paper is “broken down” first before using on the rotor. Straps 3/8–1/2" (9.5–12.7 mm) wide are wrapped one-and-one-half turns around the shaft and pulled back and forth. Strap lapping also enhances the oil-carrying capability of the rotor surface. 5.17.4 Generator Electrical Testing Electrical testing is performed to both generator components as a normal part of generator maintenance. Electrical testing can provide damage expectations and condition assessment when it is used as a diagnostic tool. During reassembly, electrical testing provides assurance that the assembled components will function as intended after the outage. Table 5-16 summarizes some of the tests that can be performed. Sections 5.17.4.1, 5.17.4.2, and 5.17.4.3 provide additional details on most of these electrical tests.
Assembly
Test
Diagnostic
Table 5-16 Generator Electrical Tests
Area
Condition Tested For
ac impedance
X
Field turn insulation
Turn shorts
Capacitance mapping
X
Stator water-cooled windings
Wet ground wall bar insulation
Copper resistance
X
Field and stator
Poor connections and opens
dc leakage current
X
Stator winding
Deterioration or contamination
Doble
X
Stator winding
Insulation integrity
EL CID
X
Stator core insulation
Weak or damaged core
Helium gas
X
X
Stator water-cooled windings
Leak detection
Hi POT
X
X
Stator winding
Insulation integrity
HIT
X
Stator water-cooled windings
Water leaks
Megger
X
Stator, field, bearing, H2 seal
Insulation: condition, assembly
Partial discharge
X
Stator windings
Localized deterioration
Stator assembly
Leak detection
Stator wedges
Loose wedges, losing tightness
Sniffer Wedge tightness
5-66
X
X X
X
Turbine-Generator Condition Assessment
An alternative to taking the generator field rotor from the stator to perform inspections is to use a limited access inspection (LAI). Limited access inspection (LAI) provides generator component assessments using robotic technology without removing the rotor. This type of testing is comparable and very often superior to conventional inspections with the rotor removed. The EPRI report Experience with Limited Access Generator Inspections: A Study of Inspections Done with Robotic Equipment and their Effectiveness as Compared with Conventional Inspections Where the Generator Rotor Is Removed, 1000100, [44] is a compilation of 68 LAIs from 1995– 1999 of two major OEMs (Siemens and General Electric) and original video tape of the inspections. The report lists the applications and capabilities of LAIs, demonstrates how LAIs compare to conventional assessments, and shows a cost comparison of LAIs versus conventional “rotor out” inspections. LAI has been accepted by the Nuclear Electric Insurance Limited (NEIL) as an equivalent rotor-out generator inspection. Applications and Capabilities of LAIs
The LAI is used to replace the routine and periodic rotor-out inspections where the rotor is physically removed from the stator by major disassembly of the end shields, bearings, oil deflectors, baffles, seals, etc. Conventional methods can possibly cause major damage to a rotor, and the inspection is time consuming and very costly. Although some disassembly is still required by LAIs, it is quite minimal compared to conventional methods and will vary depending on the LAI vendor, the size of the unit, the type of unit (nuclear or fossil), the manufacturer of the unit, and so on. LAIs have many of the same capabilities as conventional inspections. LAIs can still visually inspect a generator for “tell-tales” and “defects,” perform a stator slot and bar assessment, complete a stator end-winding inspection, and execute a rotor field visual inspection. The video records can now be recorded and archived for future reference. Very often, many specialists like to use a “touch and feel” method with some areas of the generator to better inspect components. Although LAIs do offer a similar capability, some say that it is not exactly comparable. Table 517 shows a comparison of LAI capabilities compared to conventional inspections. Note that the LAI has all of the same capabilities as the conventional rotor-out method.
5-67
Turbine-Generator Condition Assessment Table 5-17 Capabilities of LAI vs. Conventional Inspections Generator Component
5-68
LAI
Conventional Inspection
Wedges/Slots
-
-
Dusting/greasing
X
X
Tightness
X
X
Bar movement
X
X
Discharge (corona. . .)
X
X
Contamination
X
X
Wear/damage
X
X
Core
-
-
Dusting/greasing
X
X
Overheating/shorts
X
X
Wear/damage
X
X
Blocked Vent Ducts
X
X
Field
-
-
Surface Heating
X
X
Wedge movement
X
X
Vent holes
X
X
Contamination
X
X
Hardware
X
X
Retaining rings
X
X
Blocking/filler movement
X
X
End turn problems
X
X
Turbine-Generator Condition Assessment
Advantages and Disadvantages of LAIs
There is no doubt that LAIs will save maintenance costs when compared to rotor-out inspections. The amount of savings depends upon the individual plant’s situation and economics such as: •
Amount of disassembly required
•
Confined space restrictions
•
Amount of station support required
•
Length of inspection
•
Reassembly material required
•
Crane requirements
•
Rotor storage/protection
•
Rotor alignment
•
Risk of damage to rotor, stator, and other components
•
Risk of oil leak introduction
•
Foreign object material damage potential avoidance
LAIs have some small risks associated with it; however, they are smaller than the potential risk associated with pulling the rotor. (Table 5-18 presents a list of advantages and disadvantages of LAIs.) Issues such as a stuck robot inside the unit have not really presented themselves. Loose hardware falling off the robot and getting lost in the stator has been controlled by the use of standard industry locking devices (that is, safety wire, thread locking compound, plastic hardware, self-locking nuts, etc.). There have been some cases where the robot was unable to access the generator due to the baffling or other unusual dimensional access issues; however, these are usually the exception.
5-69
Turbine-Generator Condition Assessment Table 5-18 Summary of Advantages and Disadvantages of LAIs Advantages
Disadvantages
Better technology – cameras, wedge tappers, El Cid
Limited touch and feel capability in slots and on rotor body
Convenience
Exaggerated condition possible based on poor evaluation
Reduced cost
Evaluation more dependent on evaluator’s experience
Reduced outage duration
Possibility of compromised and damaged shaft insulation
Deferred rotor removal
Other inspections possibly overlooked (bearings and seals)
Potential damage from rotor removal eliminated
Interpretation and evaluation – too often simplistic – not enough relevant comments on condition
Permanent video record Frequently higher quality reports Outage interval extension facilitated Excellent application for mis-operation internal inspection More accurate tell-tale positioning data Better trending capability/records Decreased potential for foreign object damage Consistent with turbine LAI strategies Over 10 years of experience with LAI
LAIs should be strongly considered as a universal substitute for rotor-out inspections. LAIs, in addition to maintenance and operations history and standard tests, provide some of the essential ingredients for an overall comprehensive maintenance program. There may be extenuating cases where further investigation may require pulling the rotor after an LAI is performed based on the need for further inspection and testing of unique situations; however, this should be the exception.
5-70
Turbine-Generator Condition Assessment
5.17.4.1
Generator Stator Core Electrical Tests
5.17.4.1.1
EL-CID Testing
The EL-CID technique was originally devised as portable test equipment for inspection and repair of rotating electric machine stator cores. It was devised as a low excitation power alternative to the high power level stator core flux test, for looking for stator core inter-laminar insulation problems. Its application has been shown to be applicable to turbine generators, hydraulic generators, small generators and large motors. The subject of this book however, is confined to the class of large 2 and 4 pole, round-rotor machines, commonly referred to as Turbine-Driven Generators. The information contained in this section is a brief discussion of the EL-CID test technique and basic interpretation of results. Traditionally, stator core inter-laminar insulation testing has been done using the “Ring” or “Loop” flux test method, in which rated or near-rated flux is induced in the stator core yoke. This in turn induces circulating currents from the faulted area usually to the back of the core, at the core-to-keybar interface (see Figure 5-9). These circulating currents cause excessive heating in areas where the stator iron is damaged. The heat produced is generally detected and quantified using established infrared techniques. This method has been proven to be successful over the years, but requires a large power source and considerable time, manpower and resources to complete. Starting at the early 1980s, the EL-CID test has been developed as an alternative to the ring flux test. The technique is based on the detection of core faults by measuring the magnetic flux resulting from the current flowing in the fault area, at only three to four percent of rated flux in the core. Furthermore, the test usually requires only two or even one man to complete (using the latest version) in less than one eight-hour shift.
5-71
Turbine-Generator Condition Assessment
Figure 5-9 Flux Fault Current Path (Courtesy of IEEE © 2004)
5.17.4.1.1.1
EL-CID Test Procedure
The level of excitation to produce the desired flux in the stator core-back area is generally determined by a combination of the stator design parameters, and the power supply available to achieve the required flux level. For most generators, the standard 120 V AC (North America, etc.) or the 230 V AC (Europe, etc.) outlet, with a current capacity of 15 to 20 amperes is usually adequate. The characteristics of most stators are such that 4 to 7 turns of a #10 AWG insulated wire (2.5 mm2) can be used to carry the excitation current for the test. The winding is then energized to the required volts per turns, to produce approximately 3 to 4 percent of rated flux, usually corresponding to around 5 volts per meter across the stator iron. A Powerstat or Variac is best used for voltage and supply current control. The signal-processing unit of the EL-CID test equipment measures detected fault current (in QUAD mode) in mA. By theory and experimentation, a measurement of at least +/- 100 mA is required at 4% excitation of the core before it is considered that the core has significant damage affecting the inter-laminar insulation.
5-72
Turbine-Generator Condition Assessment
The excitation winding and power supply are set-up during the test as shown in Figure 5-10.
Figure 5-10 EL-CID Excitation Setup (Courtesy of IEEE © 2004)
5-73
Turbine-Generator Condition Assessment
The EL-CID equipment is set up as shown in Figure 5-11 (original analogue set), and Figure 5-12 (newer digital set).
Figure 5-11 EL-CID Analog Equipment (Courtesy of IEEE © 2004)
5-74
Turbine-Generator Condition Assessment
Figure 5-12 EL-CID Digital Equipment (Courtesy of IEEE © 2004)
In the older analogue sets, a separate coil is placed in the bore over undamaged iron as shown in Figure 5-11, to supply the reference signal. In the newer digital version, a CT (shown in Figure 5-12) is placed around the excitation winding to reference the supply signal. The CT was also an option on later analogue sets. The digital equipment uses a laptop computer to store the axial traces, whereas a plotter was used in the original version. The sensor head (chattock potentiometer) is pulled axially along the core at a speed slower than one meter every twenty seconds and always bridging two stator teeth as shown in Figure 5-13. (The slower speed is important, as the standard chattock coil has a magnetic sense area of only 4mm diameter, and both the Digital and Analogue systems have a definite time needed to record the Phase/Quad signals to sufficient resolution. The Digital set records the Phase and Quad values every 2mm. Any faster testing results in some missed test points in the Digital system or potential inaccuracy due to settling time on the Analogue system).
5-75
Turbine-Generator Condition Assessment
Figure 5-13 EL-CID Chattock Theory (Courtesy of IEEE © 2004)
The fault current signal is read directly off the signal processor meter, and input to a computer or chart recorder to trace out the readings as a function of the axial position along the stator core. When the sensor head is over undamaged iron, the meter should read zero if it is calibrated previous to the test for a condition where no fault current is circulating. In actual practice, no insulation system is perfect and some background signal is usually detected. In addition, the contact resistance of the core to keybar interface is not zero and can be found to vary between near 0 to 2 ohms. This also affects the EL-CID signal that is measured. Usually anywhere from a 0 to +/- 20 mA EL-CID signal (in Quad mode) is found to be normal when good core is measured. The above is somewhat similar to the rated flux test where the undamaged iron slowly heats up producing a background level due to eddy current losses in each lamination. During flux testing, this is recorded as the ambient core temperature rise. Where there is damaged or deteriorated core insulation, the core overheats and is detected as a hot spot above core ambient due to high fault currents circulating locally. In the EL-CID test, when the sensor head is placed over damaged core areas, the primary indication of a fault is obtained by detecting the flux produced by a current flowing in phase quadrature with respect to the excitation magnetizing current (the PHASE current). This flux is then converted back to an indicated current (the QUAD current), assumed to be flowing in the fault (see Figure 5-14). For this reason the QUAD current detected by the EL-CID processor is frequently referred to as the fault current (although for large faults the PHASE current may be affected as well, especially where the fault current path is highly inductive). The QUAD current is indicated on the signal processor meter and the traces recorded on the plotter (original analogue EL-CID equipment) or computer (newer digital EL-CID equipment). 5-76
Turbine-Generator Condition Assessment
Figure 5-14 EL-CID MMF Theory (Courtesy of IEEE © 2004)
5.17.4.1.1.2
EL-CID Experience
EL-CID has proven to be extremely reliable in detecting and locating core problems. It can cut the time and manpower requirements for core testing to within one eight-hour shift, where a flux test may have taken a few days to set up, and then a day to test, and another day to dismantle the test equipment. In the large majority of the EL-CID tests on turbo-generator machines, the experience has been that EL-CID is very reliable in determining that actual core faults or inter-laminar insulation deterioration exists. In other words, if a core defect exists, then EL-CID is likely to find it. And if the core is indicated to be defect-free by an EL-CID test, there is a very high probability that it actually is free of defects. Large signals may be found at tooth-top locations on the core, and only indicate a significant surface fault. Local surface faults are generally indicated by faults that show very localized signals, either high or low in magnitude and positive in polarity, if within the test coil span (assuming the standard EL-CID test set-up). Deeper faults can generally be seen over a larger scanning area, and also often become opposite in polarity as the sensor head gets away from the fault area. This is because the fault is outside the flux path of the chattock coil sensing the fault current, and the magnetic potential difference is reversed. Figure 5-15 shows a general basic interpretation of the EL-CID signals that can be expected to be seen based on fault location. The magnitudes in Figure 5-15 are only relative to one another, to give an idea of what might be expected for faults of roughly the same severity, at different locations. The peaks, and widths of the peaks, will vary from fault to fault as their size varies, and as they are more or less severe.
5-77
Turbine-Generator Condition Assessment
Attempts have been made to correlate EL-CID signal readings to temperatures that would be created in the defect area, during a flux test. The basic premise of the EL-CID test significance level of +/-100 mA is, that this level represents a 5 to 10ºC temperature rise that would be seen on flux tests, and therefore, just at the level of temperature rise where most OEMs and experienced stator core experts would carry out repairs to the core iron. There are a number of issues that makes questionable the assumption that correlates EL-CID’s signal to temperature. Firstly, many core testers carry out flux tests at widely differing flux levels. Some prefer to test at 100% of rated flux level while others test at about the 80% level. Different operators also apply the flux test over widely varying time periods. Yet, by all indications, all seem to work on the same temperature rise criteria. Obviously, a 10ºC rise at 100% of rated flux is much less significant than at 80% of rated flux. This is so because the 80% level is generally at the knee of the B-H curve and the curve is exponential. Increasing to 100% when the temperature rise is already 10ºC will increase the temperature in the fault. There is also the influence and undetermined significance of core-to-keybar contact resistance at the back of the core. The concern here is that the resistance can be generally measured to vary from near 0 to 2 ohms and may affect the EL-CID signal as well as the temperature measured during a flux test. For the low flux levels of the EL-CID test, it may be quite significant. This is one of the unknowns for which there is little data to support this statement, one way or another. But it should be noted that on ungrounded cores (i.e. cores with insulated keybars and infinite core to keybar contact resistance), the EL-CID and the flux test are both ineffective unless there are two faults in proximity, to allow circulating current to flow and hence be detected by either test. This has been well proven by actual tests done on such cores, where single faults do not show up until the core is artificially grounded at the back near the keybar that is behind the location of the fault.
5-78
Turbine-Generator Condition Assessment
Figure 5-15 EL-CID Signal Interpretation (Courtesy of IEEE © 2004)
Generally, there is only one keybar that is grounded on such core arrangements (for purposes of the stator ground fault relay), and the testing is effective in that lamination region only, for single faults. To test such a core arrangement by either EL-CID or flux test, the core must be purposely grounded at the back, between every keybar, circumferentially, and on every lamination axially. One can easily see the difficulty in this, as it is not even possible on most stator designs, because of the frame construction and arrangement. One has to consider this when purchasing a new stator. If a single core fault occurs, then the insulated core may allow operation where the grounded core would fail based on the progression of the fault. However, if two faults occur in the same proximity, then a much more severe fault may occur and go undetected until the effect of the failure causes additional collateral damage, which takes the machine out of service. This is generally a matter of user preference and both philosophies are sound, each in their own way. In addition, there is also the problem that core faults can manifest themselves in many forms and levels of severity. It is not uncommon for a surface iron smudge to show a very high EL-CID signal and yet not produce much heat when looked at under infrared in a flux test. And the opposite is also true. It is not uncommon for an EL-CID signal that is not much higher than the 5-79
Turbine-Generator Condition Assessment
manufacturer’s recommended significance level, to produce significant heat. In particular, with very small faults on the surface where EL-CID does not produce a significant signal, there are sometimes high spot temperatures detected by infrared, but there is insufficient power to cause damage. One advantage is that EL-CID can detect deep-seated faults, which may often not show as a particularly large temp rise on the surface, but can be quite damaging to the body of the core or adjacent conductor bar insulation. This is due to the fact that the attenuation of EL-CID signals is generally less than the attenuation of temperature rises with depth of fault. The difficulty in co-relating EL-CID and Flux test temperatures therefore comes from many issues as stated above and a number of other possible influences as listed below: •
Core-plate grade (i.e. grain oriented vs. non-oriented steels)
•
Lamination insulation grades
•
Axial length of the fault
•
Total size of the fault
•
Electrical resistance of the inter-laminar fault (i.e. deteriorated or fretted insulation type damage as opposed to hard contact, low resistance type faults)
•
Geometry differences in core structure from one machine to another
•
Limitations of the earlier EL-CID test equipment, in relation to the size of the chattock coil itself and the relative size of any fault being measured. (Current standard Chattock coils have only a 4mm diameter magnetic sense area, thus are able to detect very small faults, particularly if the suspected fault area is investigated/scanned slowly enough).
All these issues can have a significant effect on both the EL-CID signal seen, and the temperature produced during a flux test. Some are better known and quantified than others. Trying to correlate temperatures to EL-CID signals under so many variables is difficult, unless all of these parameters can be taken into account. In other words, the core under test must be well known to be able to make such a correlation. There is one other factor regarding EL-CID signal interpretation, and it has to do with readings taken in the Phase mode, as opposed to the normal Quad mode reading that Figure 11-8 is based on. Basically when a stator has (for example) 4 turns of an excitation winding and is carrying 12 amps, then it has an excitation level of 48 Ampere-Turns. When the EL-CID signal processor is set to Phase mode and a reading is taken from tooth centre to tooth centre across one slot, a signal of (48 A-T divided by 48 slots) 1 amp should be read. Generally, for most fault areas this is the reading that will be seen. However, in some cases, much higher current is read in the Phase mode than the simple magnetic potential based on excitation and slot geometry. One of the things that has been seen when this type of situation occurs, is that very high Quad readings are generally also present and the fault is usually at the bottom of the slot, or in the core yoke area. Correspondingly, there is not always much heat given off during flux testing, and the two tests do not always correlate when this occurs. There is very little experimental data on this point, and again it shows that some uncertainties remain in interpretation of EL-CID test results. It is 5-80
Turbine-Generator Condition Assessment
believed that Phase readings are also significant and should be factored into the test interpretation. Just what that interpretation should be is unclear to date, due to the difference in faults from case to case. Probably the main difficulty for the test interpreter is, when nothing is known about the core under test nor the type of fault found. Most often, the core defect is not visible, and what the tester is trying to determine is how deep it is and how severe it is. The general consensus of the people surveyed on this issue, is that more often than not, they cannot tell how severe a detected fault is, and thus require a flux test with infrared scan to help in that determination. The Ring Flux test remains the best test to determine the actual temperature rise of any fault, and if repairs are required. If the suspected fault is believed to be deep-seated from the EL-CID test result, the Ring Flux thresholds should be appropriately adjusted. Once the core is repaired, an EL-CID test can usually show that the repair is successful by the absence of a defect signal. This is perhaps the best value in EL-CID testing. There seems to be general consensus that if an EL-CID test is performed and no damage is found, then the core is defect free. EL-CID has gained good credibility in its ability to determine and locate the presence of faults, and to verify repairs when faults are found. The general consensus also appears to be that more work is required on EL-CID signal co-relation with temperature rise in fault locations. The general feeling to date is that both the EL-CID and Flux Testing together are still required to give the best information on any core defect found. 5.17.4.1.2
Rated Flux Test with Infra-Red Scan
The Rated Flux test is a high-energy test, used to check the integrity of the insulation between the laminations in the stator core. It is also commonly referred to as the “Ring Flux” test, in which near-rated flux (normally about 80%) is induced in the stator core yoke. This in turn induces circulating currents and excessive heating in areas where the stator iron is damaged (see Figure 5-9). The heat produced is detected and quantified using established infrared techniques. Flux is produced in the iron by looping a cable around the core in toroidal fashion (see Figure 516), and circulating a current at operating frequency. The flux required for the flux test is half the normal operating flux due to the difference in the way the flux is induced in operation (see Figure 5-17) from that of the flux test. The power supply for the cable is usually taken from two phases of one of the high voltage breakers (i.e. 4 kV) in the plant, or a portable motor generator set. The correct number of turns are looped around the core to produce the required level of flux. IEEE Std. 432 [11] provides the following expression to find the rated volts per turn required on the stator core: Voltage per turn of test coil = (1.05 * VLL)/(2 √3 x Kw * N)
Where: VLL Kw N
= = =
line-line voltage winding factor Number of turns/phase in series in the stator winding 5-81
Turbine-Generator Condition Assessment
Figure 5-16 Toroid Wrap (Courtesy of IEEE © 2004)
Figure 5-17 Operating Flux Pattern (Courtesy of IEEE © 2004)
5-82
Turbine-Generator Condition Assessment
From the above, the available power supply voltage is divided by the calculated volts per turn above, to give the correct number of turns to loop through the core. If the number of turns includes a fraction, then the next highest number is used, to reduce the flux level to below 100% or below rated. Using too high of a flux level can create core damage since there is no cooling on the core during the test. Once the number of turns is known, the current capacity is required to size the cable and ensure the power source can handle the current that will be drawn. Knowledge of the specific B-H characteristic of the subject core being tested is required for this. In cannot be stressed too highly that exact B-H characteristic of the stator core should be known in relation to the flux volts per turn, and the current that will be required from the power source (see Figure 5-18). In many cases it is unknown, and therefore the number of ampere-turns required must be estimated based on industry curves for the most likely grade of core-plate that would be used in the machine under test. A higher end and lower end core-plate grade are usually selected to provide a range of possible operating characteristics for the subject core. These are selected to provide a range of possible excitation requirements, based on B-H curves taken from small and large turbogenerator applications. From the winding configuration for the subject generator, the power supply available, and the BH curves, an estimate can be made for the number of turns required to achieve the required level of flux for the test. This is generally in the 70 to 90 percent range of rated flux. The current that would be flowing in the flux cable will depend on the actual B-H characteristic of the stator iron and therefore, this must be carefully estimated for safety of both personnel and the equipment. When the B-H curve is in doubt, adding a higher number turns will reduce the level of flux. Then it follows to successively remove turns and keep recording the current attained as the flux volts per turn increases. Successive voltage application in this manner can be made until a B-H curve is created and the proper number of turns found (see Figure 5-18).
5-83
Turbine-Generator Condition Assessment
Figure 5-18 B-H Curve Example (Courtesy of IEEE © 2004)
The flux test is set up as basically shown in Figure 5-19. The power supply is selected and connected as shown. The cable is wound through the stator bore the correct number of times, and connected back to the power supply. Protection for the test cable is set up to provide “ground fault” and “over-current”. The stator core, frame and the windings were all grounded for their protection and that of the test personnel. The CTs should also shorted at the terminals and grounded. Metering is set up to provide measurements of supply voltage and current. A single loop of cable is installed additionally, to measure the actual flux volts on the stator core during the test. This is done to provide an accurate measurement of the induce voltage across the core and the level of flux as well. In some cases, an infrared, non-reflecting mirror is used to monitor the temperature of the stator core when angled viewing from outside the stator bore is difficult (see Figure 5-20). The mirror provides a known surface to accurately measure the temperatures, so that the absolute and relative rise of temperatures in the core defect areas can be recorded.
5-84
Turbine-Generator Condition Assessment
Figure 5-19 Flux Test Electrical Setup (Courtesy of IEEE © 2004)
Figure 5-20 Flux Test Mirror Setup (Courtesy of IEEE © 2004)
5-85
Turbine-Generator Condition Assessment
Once the flux is established in the core, it is kept for at least thirty minutes to one hour. The temperature of the core should be maintained within values not significantly higher than those encountered during operation. Under these conditions, the temperature rises in the core are monitored and recorded while the existence of hot spots is investigated with infrared monitoring equipment (and possibly a non-reflecting mirror) (Figures 5-21 to 5-23).
Figure 5-21 Infrared Hot Spot – Bruce 7 (Courtesy of IEEE © 2004)
Figure 5-22 Infrared Hot-Spot Flux Test 1 (Courtesy of IEEE © 2004)
5-86
Turbine-Generator Condition Assessment
Figure 5-23 Infrared Hot-Spot Flux Test 2 (Courtesy of IEEE © 2004)
The temperature rises of the “good” core areas (ambient core temperature rise) are then compared to the temperature rise profile of any defective locations found. Once the defects are located and characterized, repair solutions can then be addressed. Figure 5-24 shows the relative experience in the industry with the typical types of core faults encountered, and how they appear during flux testing. It should be noted that these are general examples and may not be the case for a particular core tested. However, they show the general trend that the large majority of faults seem to follow.
5-87
Turbine-Generator Condition Assessment
Figure 5-24 Flux-Temperature Profiles (Courtesy of IEEE © 2004)
5.17.4.1.3
Core Loss Test
Stator core loss is a function of terminal voltage. The Open Circuit Saturation Curve for the core iron determines it. The core loss for any particular generator is always determined at the factory by the manufacturer, and is not a test that is generally done at site. The core loss is determined by the generator being coupled to, and driven by a calibrated motor. The friction and windage (mechanical) losses are calculated and separated out from the electrical losses, to provide a value of core loss for the stator [7]. If there is suspected wear of the inter-laminar insulation in the core, on a large scale, it may be possible that a core loss test could be done to compare the present value to the ‘as new’ value, to determine the extent of deterioration occurring. However, the serious challenge of driving the generator at site with a calibrated motor, for all practical purposes limits this test to the OEM’s factory. 5.17.4.1.4
Through-Bolt Insulation Resistance
There are a few manufacturers that provide through-bolts in their stators to pull the cores tight. These through-bolts are full-length bolts, inserted axially through the core, through holes in the core iron. There are many of them located symmetrically around the circumference of the core, a few inches below the stator winding slots. The ends are threaded and terminated at each end 5-88
Turbine-Generator Condition Assessment
through a pressure plate, where a nut is installed to maintain compression, after the core is pressed to a few hundred psi. The entire through-bolt assembly is insulated, generally by cured epoxy/glass tape wrap or a phenolic tube through the core, and an arrangement of insulators at the pressure plates and nuts. This is done to ensure that the through bolts do not create any short circuits across the stator core laminations and cause a core failure by circulating currents. To ensure that the insulation is in good condition, the insulation resistance of the through-bolts is checked by meggering at 500 V DC. A good reading should be in the hundreds of MΩ range. 5.17.4.1.5
Insulation Resistance of Flux Screens
Most large generators are provided with some form of flux screening for the stator core-end. This is to prevent overheating in the core-ends due to stray flux from the stator endwinding. When flux screens are used, they are insulated from the core end, to ensure that no additional circulating currents flow between the core and the flux screens, which would create additional un-wanted heating in the core-end. To ensure that the insulation is in good condition, the insulation resistance of the flux screens is checked by meggering at 500 V DC. A good reading should be in the hundreds of Mega-ohms range. 5.17.4.2
Generator Stator Winding Electrical Tests
Stator windings are comprised of materials with specific resistive and dielectric qualities. The materials used comprise: mica, Dacron tapes, glass tapes, asphalt binders, polyester resins and epoxy resin binders, and so on. There are also insulating, resistive and stress grading paints applied to various portions of the winding to ensure controlled distribution of the voltage on the individual stator conductor bars. All of the materials used, and their application, are done in such a manner as to ensure proper functioning and a reasonable degree of long term reliability of the winding. The stator winding insulation system is complex and requires a variety of tests to establish its present condition and expected long term reliability. Therefore, to fully test the stator winding, so that the best possible determination of the winding condition can be made, it is desirable to perform both AC and DC tests. DC tests are generally sensitive to the presence of cracks, moisture, particle contamination or a general degradation of the electrical creepage path. During DC application, the voltage is divided according to the DC leakage resistance. Basically, DC is used to test the conductivity of the insulation system. DC testing has the advantage that it less damaging to the insulation due to the absence of corona and partial discharges associated with AC. AC testing on the other hand applies the more realistic electrical stress to the winding, since it operates AC when in service. When the AC test voltage is applied, it is actually applied across several dielectric components of the winding insulation, which are effectively in series. 5-89
Turbine-Generator Condition Assessment
Therefore, the leakage current must go through each of the dielectrics until it reaches ground potential. Under AC, the voltage is therefore divided according to the relative permittivity of each of the dielectric materials. AC testing is in fact far more searching than DC. In addition to the conducting properties of the insulation, AC testing is also capable of determining the loss or power factor characteristics and the dielectric properties. In addition, the mechanical integrity of the insulation can be also be alluded to by the capacitance characteristics of the winding in terms of insulation de-lamination. Like with many other issues about testing insulation of large electric machines, experts and operators have different opinions about which one of the tow tests (DC and/or AC) is more convenient. Some only prefer DC tests, while other prefer AC testing. Still, others prefer using both. 5.17.4.2.1
Pre-Testing Requirements
If the stator winding is water cooled, it must be completely dried prior to all testing to obtain meaningful results. If there is stator cooling water left in the winding it will alter the test results and give a distorted picture of the insulation condition. All three phases must be isolated to ensure all testing is carried out on the stator winding only. This means that each phase should be completely separated at the neutral point and floated from ground. The line ends of the stator winding should be separated from the Isolated Phase Bus (or cables, in smaller units) just outside the generator, at the stator terminals. The generator current transformer windings should be shorted and grounded to avoid induced high voltage and possible discharge failure of the insulation. All instrumentation leads should be grounded to also avoid induced high voltage and possible discharge failure of the insulation. Before conducting any high-voltage testing of the unit, consult vendor and/or pertinent standards. 5.17.4.2.2
Series Winding Resistance
This test is used to measure the ohms resistance of the copper in each phase of the stator winding. Given the relatively low DC series resistance of windings of large machines, the measurement accuracy requires significance to a minimum of 4 decimal places. The purpose of the test is to detect shorted turns, bad connections, wrong connections and open circuits. Acceptable test results consist of the three resistance values (one per phase) to be balanced within a 0.5% error from the average. The test is very sensitive to differentials of temperature between sections of the winding. The machine should be at room temperature when the test is performed. 5-90
Turbine-Generator Condition Assessment
As with any other electrical test, the results should be compared with original factory data, if available. This test can be performed on stator and rotor windings. 5.17.4.2.3
Insulation Resistance (IR)
The purpose of this test is to measure the ohmic resistance between the conductors in each of the 3-phases and ground (i.e. the stator core). This test is generally regarded as an initial test to look for gross problems with the insulation system, and to ensure further high voltage electrical testing may “relatively” safely continue, in terms of danger of failing the insulation. Normally, the measurements of IR will be in the mega-ohm range for good insulation, after the winding is subjected to a DC test voltage usually done anywhere from 500 to 5000 V, for one minute. The minimum acceptable reading by IEEE Standard 43 [4] is (VLL in kV + 1) MΏ. The test is carried out with a “Megger” device. However, resistance bridges may also be employed. The DC test voltage level is usually specified based on: the operating voltage range of the machine, the particular component of the generator being tested, operator’s policy and previous experience, and knowledge of the present condition of the insulation in the machine. Although the readings obtained will be somewhat voltage-dependent, this dependency becomes insignificant for machines in which the insulation is dry and in good condition. This is why it is essential that the stator winding is completely dried before any testing, so that any poor readings will be due to a “real” problem, and not residual moisture from the stator cooling water. The readings are also sensitive to factors like humidity, surface contamination of the coils, and temperature. Readings should be corrected to a base temperature of 40oC by the following: R (40oC) = K * Rmeasured (oC) Where K is a temperature-dependent coefficient, which can be obtained from IEEE Std 43 (see Figure 5-25). The following equation can be used to obtain K to some degree of accuracy in lieu of the standard: K = 0.0635 * exp(0.06895 * Tmeasurement in oC) Insulation Resistance tests are performed on both stator and rotor windings, core-end flux screens, and core-compression or through-bolts as mentioned previously.
5-91
Turbine-Generator Condition Assessment
Figure 5-25 IR Versus Temperature (Courtesy of IEEE © 2004)
5-92
Turbine-Generator Condition Assessment
The insulation between the core-end flux screen and the stator core-end iron ensures that the flux screen maintains its capability to shield the core-end from axial flux, and keep the resulting circulating currents within the flux screen, without providing a current path to the core. The insulation between core-compression bolts and the iron keeps the through-bolts from shortcirculating the insulation between the core laminations. Otherwise, large eddy currents generated within the core would produce heat and temperatures, which could further damage the interlaminar insulation, as well as the insulation of the windings. Figure 5-26 shows typical IR behavior as a function of time.
Figure 5-26 Polarization Index Dryness Curve (Courtesy of IEEE © 2004)
5-93
Turbine-Generator Condition Assessment
5.17.4.2.4
Polarization Index (PI)
Insulation resistance is time-dependent as well as being a function of dryness. The amount of change in the IR measured during the first few minutes depends on the insulation condition, and the amount of contamination and moisture present. Therefore, when the insulation system is clean and dry, the IR value tends to increase as the charge is absorbed by the dielectric material in the insulation. When the insulation is dirty, wet or a gross insulation problem is present, the charge does not hold and the IR value will not increase, due to constant leakage current at the problem area. Therefore, the ratio between the resistance reading at 10 minutes and the reading at 1 minute produces a number or “Polarization Index” which is essentially used to determine how clean and dry the winding is (see Figure 5-27). Class B and F windings tend to show higher PI values than windings made of Class A insulation. It is also dependent on the existence of a semi-conducting layer. The recommended minimum PI values are as follows: Class A insulation: Class B insulation: Class F insulation:
1.5 2.0 2.0
The same Megger used for the IR readings should be used to determine the PI. The PI readings should be done on a per phase basis at the same voltage as the IR test, and can be used as a go/no-go test before subjecting the machine to subsequent high voltage tests, either AC or DC. The IR readings for the PI test should also be corrected to 40oC as in the IR test. Performing the high voltage tests on wet insulation may result in unnecessary failure of the insulation.
5-94
Turbine-Generator Condition Assessment
Figure 5-27 IR Versus Temperature – PI (Courtesy of IEEE © 2004)
5.17.4.2.5
Dielectric Absorption During DC Voltage Application
Dielectric absorption current characteristics can be used to measure the aging of the resin binder in the groundwall insulation. When applying DC voltage to insulation material, a time-dependent flow of current is established. This current has a constant component, called the conduction current or leakage current, and a transient component, called the absorption current.
5-95
Turbine-Generator Condition Assessment
Absorption current is a function of the polarization of the molecules in the binding material. The older the binding material, the more polarized it becomes, and the more absorption current flows. Therefore, this test is best used as a comparison test between the winding condition at different times, and between similar windings. Absorption current is also temperature dependent. This fact should be taken into consideration when performing the test and interpreting the results. Absorption current is also dependent on the amount of voids in the insulation. The dependence is inverse, i.e. an increase in the number of voids in the insulation, will tend to reduce the magnitude of absorption current. The contradictory effects regarding voids density and aging of the binding material, renders this test difficult to interpretation. It is best when used in conjunction with other dielectric tests, such as Partial Discharge and Dissipation Factor Tip-Up tests. 5.17.4.2.6
DC Leakage or Ramped Voltage
The DC leakage or ramped voltage test, is a controlled DC voltage application designed to test the winding in such a manner as to monitor the DC leakage current, at the same time the DC voltage is increased. The leakage current is plotted against the DC voltage applied to give early warning of any impending insulation breakdown. This helps in limiting damage by shutting down the test prior to a full breakdown occurring (see Figure 5-28). When applying DC voltage to the winding, a time-dependent flow of current is established. This current has a constant component, called the conduction or leakage current, and an initial component, called the charging or absorption current. Therefore, it is advisable to raise the voltage to the first level of the kV/min. rate, and hold for 10 minutes, to get beyond the charging phase of the voltage application, and test while dealing primarily with the leakage current. In this way, charging current influence on the leakage current rate of rise will be minimized. The final DC test voltage level is generally in the range of 125% to 150% of (VLL x 1.7) kV DC (ANSI Standard C50.10). The value actually chosen between 125% up to 150% of the test voltage is dependent on the age of the machine insulation, and knowledge of its general condition. The ramp rate is selected at 3% of the final test voltage level in kV DC/minute (IEEE Standard 95). The ramp rate usually is in the range of 1.5 to 2 KV per minute. Generally the ramping portion of the test is automated to allow a steady increase in the voltage. A DC Hi Pot set with the capability of a timed and steady voltage increase is required. If this is not possible with the equipment available, then a basic DC Hi Pot set can be used to raise the voltage in the pre-determined 3% voltage steps, holding each step for 1 minute.
5-96
Turbine-Generator Condition Assessment
Figure 5-28 DC Ramp (Courtesy of IEEE © 2004)
5.17.4.2.7
DC Hi-Pot
The DC Hi-Pot test is used to ascertain if the winding is capable of sustaining the required rated voltage levels (without a breakdown of the insulation), with a reasonable degree of assurance for capability to withstand over-voltages and transients, and maintain an acceptable insulation life. 5-97
Turbine-Generator Condition Assessment
The test consists of applying high voltage to the winding (the three phases together, or one at a time, with the other two grounded) for one minute. The recommended test voltage level is [(2 x VLL + 1000) x 1.7] kV DC for new windings (ANSI Standard C50.10). The recommended test voltage level for field-testing and maintenance purposes is 125% to 150% of (VLL x 1.7) kV DC (ANSI Standard C50.10). The value actually chosen for the test voltage is dependent on the age of the machine insulation, knowledge of its general condition, and the specific situation calling for a test. 5.17.4.2.8
AC Hi-Pot
The AC Hi-Pot test is also used to ascertain if the winding is capable of sustaining the required rated voltage levels (without a breakdown of the insulation), with a reasonable degree of assurance for capability to withstand over-voltages and transients, and maintain an acceptable insulation life. The test consists of applying high voltage to the winding (the three phases together, or one at a time, with the other two grounded) for one minute. The recommended test voltage level is (2 x VLL + 1000) kV AC for new windings (ANSI Standard C50.10). The recommended test voltage level for field-testing and maintenance purposes is 125% to 150% of VLL kV AC (ANSI Standard C50.10). The value actually chosen for the test voltage is dependent on the age of the machine insulation, knowledge of its general condition, and the specific reasons for the calling for a test. AC testing is generally done at power frequency of 60 Hz but may also be carried out at a low frequency of 0.1 Hz, which is the accepted industry standard. Generally the AC Hi-Pot is a “pass” or “fail” type of test. However, this is not always the case. There are often times when arcing can be heard and even seen (see Figure 5-29) and the test can be stopped until the problem area is repaired. Then retesting may be carried out to prove the repairs. Also, testing is usually done on water-cooled windings with the system drained and vacuum dried before voltage application. However, there are instances where good DC measurements have been recorded on two phases while one appears grounded, while the winding is technically “wet”. In such an instance, AC testing has been done as a next step, but this is a rare occasion. It is not recommended to proceed with this type of testing unless an expert is present to know how to handle this type of situation. One of the reasons to do a “wet” AC test would be when there appears to be no failure point that can be found and yet the winding will not hold DC voltage and 5-98
Turbine-Generator Condition Assessment
internal contamination of the stator winding hoses is suspected. Under dry conditions, the winding will pass high voltage testing and under wet conditions, the contamination will be conducting. Depending on the type of contamination and its conductivity, the hoses may glow under high voltage AC (see Figure 5-30).
Figure 5-29 Stator Hi-Pot Arcing (Courtesy of IEEE © 2004)
5-99
Turbine-Generator Condition Assessment
Figure 5-30 LKV G5 EE Hoseglow (Courtesy of IEEE © 2004)
5.17.4.2.9
Partial Discharge (PD) Off-line Testing
In principle, PD measurements are based on direct measurement of the pulses of high frequency current discharges created during the occurrence of partial discharges. Some off-line methods are based in a capacitive link between the whole of the winding and the measurement equipment. These set-ups allow the measurement of PD activity in whole windings, or one phase at a time. To measure the PD activity in smaller sections of the winding, methods based on an electromagnetic probe or pickup (which is mounted on a hand held electrically insulated stick) has been developed. One such probe is known as the TVA Probe and is used to traverse the entire length of a slot in the stator bore to search for localized sources of PD. Therefore, each slot is probed over its full length. The partial discharge tests are carried out from voltages below the inception voltage up to rated voltage. On-line partial discharge analysis can be performed by modern instrumentation and methods described in the following sections.
5-100
Turbine-Generator Condition Assessment
5.17.4.2.9.1
PD Monitoring by Capacitive Coupling
Partial discharge monitoring by capacitive coupling is generally an on-line test these days, but off-line measurements are also done on a regular basis. The test setup for off-line capacitive coupling is generally as shown in Figure 5-31.
Figure 5-31 PD Off-Line Capacitive Coupling (Courtesy of IEEE © 2004)
During operation of a generator, the voltage on the stator winding is graded according to the line to neutral connection. Thus, when an on-line test is performer, the bars near the neutral end of the machine are not subjected to high voltage, which represents the actual operating condition. In the off-line test, all stator bars are energized to the level of the test voltage applied and therefore, all may show PD activity. However, in the off-line test, the effects of vibration and bar forces are not in play. These issues are important to be taken into consideration when analyzing the test results and their implication on the condition of the unit. 5.17.4.2.9.2
PD Monitoring by Stator Slot Coupler
The Stator Slot Coupler (SSC) is basically a tuned antenna with two ports. The antenna is approximately 18 inches (46 cm) long and is embedded in an epoxy/glass laminate with no conducting surfaces exposed. SSCs are installed under the stator wedges at the line ends of the stator winding, such that the highest voltage bars are monitored for best PD detection. Since the SSC is also installed lengthwise in the slot at the core end, its two port characteristic gives it inherent directional capability. The problem of noise is virtually eliminated in the SSC. Although the SSC has a very wide frequency response characteristic that allows it to see almost any signal present in the slot where it is installed, it also has the characteristic of showing the true pulse shape of these signals. This gives it a distinct advantage over other methods, which cannot capture the actual nature of the PD pulses. Since PD pulses occur in the 1 to 5 nanoseconds range and are very distinguishable with the SSC, the level of PD activity can be more closely defined. 5-101
Turbine-Generator Condition Assessment
In addition, dedicated monitoring devices have been devised to measure the PD activity detected in the SSC. The capability for PD detection using the SSC and its associated monitoring interface is enhanced to include measurement in terms of the positive and negative characteristic of the pulses, the number of the pulses, the magnitude of the pulses, the phase relation of the pulses and the direction of the pulses (i.e. now from the slot or from the endwinding or actually under the SSC itself at the end of the slot). The other advantage of the SSC is that once it is installed, measurements may be taken at any time without the need for exposing live portions of the generator bus-work, for the purpose of making connections to the test equipment. 5.17.4.2.9.3
Corona Probing for PD
Partial discharge tests in general determine only the relative condition of the stator winding from the generator terminals. They do not locate specific sites of deterioration or damage in the winding. To do this, the winding must be locally scanned with special probes designed to detect localized sources of PD, while the winding is energized to the level of line-to-neutral voltage. There are a couple of variations of probe types, one based on radio frequency noise and the other on acoustical noise. (SSCs do provide some information about the location of the PD activity. The more SSCs installed in a particular winding, the higher the accuracy in determining the location of the offending bar). The “TVA Probe” gets its name from the Tennessee Valley Authority where it was first popularized. It is based on an earlier Westinghouse probe design, sensitive to RF signals produced by PD in the winding. It functions by picking up the RF energy radiated from active PD sites in the winding. The greater the PD, the greater the RF energy produced. The tip of the TVA probe employs a loop antenna similar to that used in an AM radio. The TVA antenna is tuned to about 5 MHz so that it is sensitive to near-field RF discharge. The output of the antenna is directed by a co-axial cable to a tuned RF amplifier and a peak-reading ammeter that is sensitive to peak PD pulses. The closer the antenna is brought to an RF (or PD) source, the higher the output on the meter. The “Ultrasonic Probe” functions based on acoustic noise produced by localized PD sites. The noise is similar to a crackling sound that one might hear when next to a high voltage overhead transmission line on a wet day. This noise is loudest in the ultrasonic frequency range around 40 kHz. A high directional microphone, sensitive to the 40 kHz noise, is used to locate the site of the PD discharges. Given that ultrasonic noise does not easily penetrate insulation, the ultrasonic probe test is primarily sensitive to surface PD, i.e. the sites of slot discharge and surface endwinding PD. 5.17.4.2.10
Capacitance Measurements
Capacitance measurements are also a method of measurement by which the quality of the insulation can be indicated. The measurements are, of course, done with AC voltage, and generally on a per phase basis. 5-102
Turbine-Generator Condition Assessment
Each phase of the stator winding is energized to line-to-neutral voltage, while the other two phases are grounded. The power factor of the winding is measured with a capacitance bridge to determine the value of the per-phase winding capacitance. Comparison of the measured capacitance to the factory measured values, and then successive capacitance readings, can aid in showing deterioration of the ground-wall insulation over time. 5.17.4.2.11
Dissipation/Power Factor Testing
The dissipation factor (or tan δ) is an AC test used to measure the bulk quality of the groundwall insulation, by measuring the dielectric loss (primarily due to partial discharges) per unit of volume of the insulation. Note:
(Dissipation Factor) DF = tan δ (Insulation Power Factor) IPF =
DF
. = sin δ
√ 1 + ( DF )2
Results are generally dependent on the type of the dielectric material in the insulation system. An increase in DF over the life of the winding can be attributed to an increase in internal voids, delaminations, and/or increased slot-coil contact resistance (i.e. deterioration of the semiconducting paint in the slot). The readings are a dimensionless quantity expressed in percent. The absolute values obtained are, again, a function of the type of insulation system being measured and are also directly affected by the temperature of the winding. Therefore, it is important that insulation power factor readings be taken at similar temperatures. The results are even more useful however, in relative terms by comparison of present readings to past readings. Successive measurements provide a scale of the deterioration rate of the insulation system over time. Therefore, when using dissipation factor as a function of time, it is important to maintain constant conditions during testing. DF readings are directly affected by the temperature of the winding and are also a function of the applied voltage. Therefore, comparisons with previous readings should be made on tests done at similar temperatures and the same voltage levels. Since dissipation factor readings are somewhat void dependent, the dissipation/insulation power factor ratio will increase with an increase in the amount of voids present in the insulation. This phenomenon is the base of the DF/IPF Tip-Up test. 5.17.4.2.12
Dissipation/Power Factor Tip-Up Test
The Dissipation Factor Tip-Up or ∆tanδ test looks at the void content in the insulation. That is to say, the dissipation factor will increase with an increase in the amount of voids or de-lamination present in the insulation. In addition, it also provides information on other ionizing losses in the form of partial and slot discharges.
5-103
Turbine-Generator Condition Assessment
The test is done by taking DF (or insulation power factor) measurements at different voltages. A set of readings is therefore obtained, which forms an ascending curve. A fast change of insulation power factor with increasing voltage tends to indicate a coil with many voids. The test is based on the fact that ionization, both internal and external to the insulation is voltage dependent. The test is done generally at 25 and 100 percent of the rated phase to neutral voltage. The Tip-Up value is the DF measurement at the higher voltage, minus the DF measurement at the lower voltage (IEEE Standard 286). Good readings for an epoxy/mica system, indicating minimal void content in the insulation, are typically less than 1%. Good readings for an asphalt system are generally in the 3% range (see Figure 5-32). This test will give a good evaluation of the winding as a group, however any bad coil that deviates greatly from the rest will not be discerned by this test. To ferret out individual bars, which may exhibit higher discharges, a Partial Discharge test can be done with the addition of manual probing for the location of the discharges if high levels are found to exist.
Figure 5-32 Dissipation Factor Tip-Up (Courtesy of IEEE © 2004)
5-104
Turbine-Generator Condition Assessment
5.17.4.3
Generator Rotor Electrical Testing
5.17.4.3.1
Winding Resistance
The field-winding series-resistance is measured to determine the ohms resistance of the total copper winding in the rotor. Given the relatively low DC series resistance of windings of large machines, the measurement accuracy requires significance to a minimum of 4 decimal places. The purpose of the test is to detect shorted turns, bad connections, wrong connections and open circuits. The machine should be at room temperature when the test is performed. As with most other electrical test, the results should be compared with original factory data, if available. 5.17.4.3.2
Insulation Resistance (IR)
The purpose of the IR test is to measure the ohmic resistance between the total rotor winding insulation and ground (i.e. the rotor forging). This test is generally regarded as an initial test to look for gross problems with the insulation system, and to ensure further high voltage electrical testing may (relatively) safely continue, in terms of danger of failing the insulation. Normally, the measurements of IR will be in the mega-ohm range for good insulation, after the winding is subjected to a DC test voltage usually done anywhere from 500 to 1000 V, for one minute. The minimum acceptable reading by IEEE Standard 43 is (Vf in kV + 1) MΩ. The test is carried out with a “Megger” device. The DC test voltage level is usually specified based on: the operating and field forcing voltage of the rotor, utility policy and previous experience, and knowledge of the present condition of the insulation in the rotor. It is essential that the rotor winding be completely dried before any testing, so that any poor readings will be due to a “real” problem and not residual moisture. The readings are also sensitive to factors like humidity, surface contamination of the coils, and temperature. Readings should be corrected to a base temperature of 40oC (see Figure 5-26). All of the above also applies to the rotor bore copper and collector rings. 5.17.4.3.3
Polarization Index (PI)
Insulation resistance is time dependent as well as being a function of dryness for rotor insulation, just as in the stator. The amount of change in the IR measured during the first few minutes depends on the insulation condition, and the amount of contamination and moisture present. 5-105
Turbine-Generator Condition Assessment
Therefore, when the insulation system is clean and dry, the IR value tends to increase as the dielectric material in the insulation absorbs the charge. When the insulation is dirty, wet or a gross insulation problem is present, the charge does not hold and the IR value will not increase, due to constant leakage current at the problem area. Thus, the ratio between the resistance reading at 10 minutes and the reading at 1 minute produces a number or “Polarization Index” which is essentially used to determine how clean and dry the winding is. The recommended minimum PI values are as follows: •
Class B insulation:
2.0
•
Class F insulation:
2.0
The same Megger used for the IR readings should be used to determine the PI. The PI readings should be done at the same voltage as the IR test and can be used as a go/no-go test before subjecting the rotor to subsequent high voltage tests, either AC or DC. The IR readings for the PI test should also be corrected to 40oC as in the IR test (see Figure 5-25). Performing the high voltage tests on wet insulation may result in unnecessary failure of the insulation. 5.17.4.3.4
DC Hi-Pot
The DC Hi-Pot test is used to ascertain if the winding is capable of sustaining the required rated voltage levels (without a breakdown of the insulation), with a reasonable degree of assurance for capability to withstand over-voltages and transients, and maintain an acceptable insulation life. The test consists of applying high voltage to the rotor winding for one minute. DC Hi-Pot testing on rotor windings is normally done between 1500 V up to approximately 10 times the rated field voltage. 5.17.4.3.5
AC Hi-Pot
The AC Hi-Pot test is also used to ascertain if the winding is capable of sustaining the required rated voltage levels (without a breakdown of the insulation) with a reasonable degree of assurance for capability to withstand over-voltages and transients, and maintain an acceptable insulation life. The test consists of applying AC high voltage to the rotor winding for one minute. AC Hi-Pot testing on rotor windings is also normally done at to 10 times the rated field voltage at line frequency of 60 Hz. 5.17.4.3.6
Shorted Turns Detection - General
Shorted turns in rotor windings are associated with turn-to-turn shorts on the copper winding, as opposed to turn to ground faults. 5-106
Turbine-Generator Condition Assessment
Rotor winding shorted turns, or inter-turn shorts can occur from: an electrical break down of the inter-turn insulation, mechanical damage to the inter-turn insulation allowing adjacent turn to turn contact, or contamination in the slot which allows leakage currents between turns. A project was implemented by EPRI that developed a traveling wave monitoring technique that can both detect and determine the location of shorted windings on-line without the installation of a sensing coil. The report On-Line Detection of Shorts in Generator Field Windings, TR-114016, [45] describes this project. When shorted turns occur, the total ampere-turns produced by the rotor are reduced, since the effective number of turns has been reduced by the number of turns shorted. The result is an increase in required field current input to the rotor to maintain the same load point, and an increase in rotor winding temperature. At the location of the short, there is a high probability of localized heating of the copper winding and arcing damage to the insulation between the turns. This type of damage can propagate and worsen the fault, such that more turns are affected, or the ground-wall insulation becomes damaged and a rotor winding ground occurs. One of the most noticeable effects of shorted turns is increased rotor vibration due to thermal effects. When a short on one pole of the rotor occurs, a condition of unequal heating in the rotor winding will exist between poles. The unequal heating may cause bowing of the rotor, and hence vibration. The extent and location of the shorted turns and the heating produced will govern the magnitude of the vibrations produced. One general relationship between the location of the shorted turn/turns and vibration is: •
Lower vibration is generally experienced when the short is on the Q-axis.
•
Higher vibration is generally experienced when the short is nearer the pole or D-axis.
Stated differently, the rotor is more prone to vibration due to shorted turns, if the shorts are located in the “small coils” rather than in the “large coils”. The “small coils” being those located closer to the pole-faces. The reasoning for the above is the lack of symmetry with faults nearer the pole face. There is an inherent unbalance in the geometry and heating effect on the rotor forging. Off-line methods for detecting shorted turns include winding impedance measurements as the rotor speed is varied from zero to rated speed, and RSO (Recurrent Surge Oscillation) tests, based on the principle of time domain reflectometry. In addition, a short of significant magnitude may be identified by producing an Open Circuit Saturation Curve, and comparing it to the design OC Saturation Curve. If the field current required to produce rated terminal voltage has increased from the original design curve, then a short would be likely present. The number of shorted turns may be identified by the ratio of the new field current value over the design field current value. All of the above methods of identifying shorted turns are prone to error and only indicate that a short exists. They do little to help locate which slot the short is in and require special conditions 5-107
Turbine-Generator Condition Assessment
for collecting the data or for testing. To better identify shorted turns, and to employ a method that works on-line, the search coil method has been perfected. Each OEM has their own version of a search coil method, but all work essentially in the same manner. 5.17.4.3.7
Shorted Turns Detection by Recurrent Surge Oscillation (RSO)
In the RSO method, a low voltage (a few volts) high-frequency (kHz range) surge wave is injected at each one of the collector rings. The two signals are then compared to determine if the same waveform is observed at each collector ring. If the waveform is identical, then no shorts are present. Variations in the two waveforms would indicate shorts to be present. This method is based on the principle of time domain reflectometry. This also has the advantage of allowing the rotor to be spun as well, while doing the measurements, to determine if the shorts are also speed sensitive. This test has the advantage of taking the mechanical loading effects into consideration. In the spinning RSO, there may be shorts that reveal themselves, which are not seen when the rotor is at rest, because at rest there is no mechanical load on the winding turns, other than their own weight. Because the RSO also works on a time of flight principle, the location of the coil number where the shorts are, as well as which pole, are also somewhat discernable by this method. Shorts nearer the sliprings show up as blips in the RSO pulse nearer the left side of the traces. And for the number of turns shorted at the particular location (i.e. the particular coil), the magnitude of the blip increases as more turns are shorted. In the “at-rest” test, the RSO is connected directly to the winding via the collector rings. Thus, only the winding impedance is seen by the high frequency, low voltage pulses sent by the RSO. In the “spinning” RSO test, to accommodate the moving rotor, the leads of the RSO must be connected to the brush rigging, and the connection to the winding is then implemented via the brushes-collector-rings. However, with this connection also anything connected towards the excitation equipment is “seen” by the pulses (e.g.: leads, contacts, field breaker, field resistor, excitation equipment). The principle of operation of the RSO is comparing the pulses inserted in each polarity terminal of the winding, and their reflections. The test is extremely sensitive to any asymmetry on the path of the pulses. From a point of view of the wave-impedance seen by the high-frequency pulses, the field winding is by nature very symmetrical, but the excitation system is anything but that. Therefore, in order to obtain any significant signature on the condition of the field-winding, the “noise” originating in the path towards the excitation must be reduced as much as possible. This is achieved by opening the excitation leads at a convenient location between the excitation system and the brush rigging. After the leads are open, only cables of almost exactly equal length are left connected to the brush rigging. The effect introduced by these cables is generally negligible. Figures 5-33 to 5-39 depict samples of RSO test-readings taken on a 2-pole turbo-generator rotor.
5-108
Turbine-Generator Condition Assessment
Figure 5-33 NO Shorted Turns Traces – Superimposed (Courtesy of IEEE © 2004)
Figure 5-34 NO Shorted Turns Traces – Separated (Courtesy of IEEE © 2004)
Figure 5-35 NO Shorted Turns Traces – Summed (Courtesy of IEEE © 2004)
5-109
Turbine-Generator Condition Assessment
Figure 5-36 RSO Single-Shorted Turn – Dual Superimposed Trace (Courtesy of IEEE © 2004)
Figure 5-37 RSO Single-Shorted Turn – Difference Trace (Courtesy of IEEE © 2004)
Figure 5-38 RSO Dual-Trace – Multi-Shorts (Courtesy of IEEE © 2004)
5-110
Turbine-Generator Condition Assessment
Figure 5-39 RSO Difference Trace – Multi-Shorts (Courtesy of IEEE © 2004)
5.17.4.3.8
Shorted Turns Detection by Open Circuit Test
Producing an Open Circuit Saturation Curve, and comparing it to the design Open Circuit Saturation Curve may identify a shorted turn condition of significant magnitude. If the field current required to produce rated terminal voltage has increased from the original design curve, then a short would be likely present (see Figure 5-40). The number of shorted turns may be identified by the ratio of the new field current value over the design field current value. However, due to the many number of turns in a typical rotor winding, the changes in open circuit voltage due to a single shorted turn in the field winding may go unnoticed since the measurement is too small for a positive identification. The open circuit stator voltage versus field current characteristics can be measured in all synchronous machines. This curve, taken with the machine spinning at synchronous speed, is unique for each machine. In principle, this test allows detecting shorted turns in brushless machines, where RSO techniques are too difficult to perform, and always entails partial disconnection of the rotor leads.
5-111
Turbine-Generator Condition Assessment
Figure 5-40 STD by Open Circuit (Courtesy of IEEE © 2004)
5.17.4.3.9
Shorted Turns Detection by Winding Impedance
Impedance measurements while the machine is decelerating or accelerating can also be used to detect a speed dependent shorted turn. Any sudden change in the readings may indicate a shorted turn being activated at that speed. A gradual change of impedance of more than 10% may also indicate a solid short (Figure 5-41).
5-112
Turbine-Generator Condition Assessment
Figure 5-41 STD by Impedance (Courtesy of IEEE © 2004)
5.17.4.3.10
Shorted Turns Detection by Low Voltage DC or Volt Drop
This test is designed to determine the existence of shorted turns in the rotor winding. The test is entirely different when performed on salient pole rotors than in cylindrical (round) rotors. In salient pole machines, a “pole drop” test is done. In this test, the resistance across each individual pole is measured by the V/I method, i.e., applying a voltage of around 100 to 120 volts, 60 Hz, to the entire winding, and then measuring the voltage drop across each pole. A pole with lower voltage drop will indicate a shorted turn or a number of shorted turns. In either salient pole or round rotor machines, the shorted turns are often speed dependent (i.e. they might disappear at standstill). To partially offset this phenomenon, it is recommended to repeat the pole drop test a few times with the rotor at several angles. The gravity forces exerted on the vertically located poles may activate some short circuits between turns, which might not show up when in, or close to, the horizontal position. In round rotors the individual windings are generally not accessible, unless the retaining-rings are removed. Therefore, detection of shorted turns in not always possible by this method. 5.17.4.3.11
Shorted Turns Detection by Low Voltage AC or ‘C’ Core Test
A “C” shaped, wound core is required to carry out this test, together with a voltmeter, wattmeter and single phase power supply (see Figure 5-42). Shorted turns are detected by sharp changes in the direction of wattmeter readings (see Figure 543). 5-113
Turbine-Generator Condition Assessment
In rotors with damper windings, or with the wedges short-circuited at the ends to form a damper winding, these have to be disconnected at the ends. This operation requires removal of the retaining-rings.
Figure 5-42 C-Core 1 (Courtesy of IEEE © 2004)
5-114
Turbine-Generator Condition Assessment
Figure 5-43 C-Core 2 (Courtesy of IEEE © 2004)
5.17.4.3.12
Shorted Turns Detection by Shorted Turns Detector (Flux Probe)
The flux probe is actually a search coil mounted on the stator core by various methods, but located strategically in the air gap. The search coil looks at the variation in magnetic field produced in the air gap by the rotor as it spins. The energized rotor winding and the slotted effect of the winding arc cause a sinusoidal signal to be produced in the winding face of the rotor. The pole face on the other hand has no winding and the signal is more flat since the variation in magnetic field is minimal. The magnitude of the sinusoidal peaks in the winding face is dependent on the ampere-turns produced by the winding in the various slots. If there is a short in a slot, then the peak of the signal for that affected slot will be reduced. The reduction will be dependent on the magnitude of the short. Therefore, as well as knowing which slot the short is in, an estimate of the number of shorted turns can be made fairly accurately. Problems due to saturation effects at full load can occur in analyzing the data and most OEM’s now have a dedicated monitor connected to the flux probe to automate the analysis process. This allows the flux probe and monitor to act as stand-alone sensor to alarm when a short turn is detected and notify the operator for investigation. 5-115
Turbine-Generator Condition Assessment
Probably the most effective method for the detection of shorted turns in solid rotors is the flux probe method. This device maps the flux of the machine as it rotates, indicating possible shorts as changes in the measured waveform. Its main advantage is that it works with the rotor on-line, capturing the speed dependent shorts. Its main disadvantages are the expertise required in analyzing the recorded waveforms, and the fact that the machine has to be de-energized and degassed for the installation of both core-mounted and wedge-mounted types of probes. New commercially available units intended for on-line continuous operation, include software, which analyses the waveform and alerts to a possible shorted-turn condition. 5.17.4.3.13
Field Winding Ground Detection by Split Voltage Test
The “split voltage” test is used locate rotor grounds as a percentage through the field winding. For this test to be effective, the resistance to ground of the fault must be less than 5% of the balance of the rotor insulation, and the voltmeter must have high input impedance, when compared to the ground fault. The retaining-rings should also be left on in case the ground is to one of the rings. The test is done by applying up to 150 Volts DC, ungrounded, across the sliprings. A measurement of DC voltage is then taken from the rotor coupling at the turbine end of the forging to one of the collector rings. The measurement is then made from the other collector ring and the same location on the rotor coupling at the turbine end. In this way, the two voltage measurements can be compared to estimate how far into the winding the ground has occurred. If the two measurements are equal, the rotor ground fault should be found in the middle of the winding. If there is less than 2% difference between the two readings, then the ground could possibly be at the collector rings. This test is very useful in helping to determine how much dismantling is required to find the ground. Depending on where the ground is located, it can obviously make a big difference in the time expended to find the fault (see Figure 5-44).
5-116
Turbine-Generator Condition Assessment
Figure 5-44 Rotor Ground – Split Voltage (Courtesy of IEEE © 2004)
5.17.4.3.14
Field Ground Detection by Current Through Forging Test
The “current through forging” test is another test used to locate rotor-winding grounds. In this particular application, the test is used to locate the actual “axial” position of the ground. The retaining-rings should be left on the rotor in case the ground is at one of the rings. For this test, a DC current of about 500 amps is put through the forging from the tip of the forging at the slipring end to the coupling at the other end. A DC ammeter is used to look for the ground position. This is done by attaching one lead of the ammeter to the most outboard slipring, and then using the other lead to probe along the axial length of the rotor forging. At the point where the ground is, the current should reduce to zero or if the current is not zero but only very low, then there will be a polarity change at the ground location (see Figure 5-45).
5-117
Turbine-Generator Condition Assessment
Figure 5-45 Rotor Ground – Current Through Forging (Courtesy of IEEE © 2004)
5.17.4.3.15
Shaft Voltage and Grounding
During operation, voltage may rise on the generator rotor shaft, unless the shaft is grounded. The sources of shaft voltage are well established and identified as: voltage from the excitation system due to unbalanced capacitive coupling, electrostatic voltage from the turbine due to charged water droplets impacting the blades, asymmetric voltage from unsymmetrical stator core stacking, and homopolar voltage from shaft magnetization. If these voltages are not drained to ground they will rise and break down the various oil films at the bearings, hydrogen seals, turning gear, thrust bearing, etc. The result will be current discharges and electrical pitting of the critical running surfaces of these components. Mechanical failure may then follow. Inadequate grounding of the rotor will also allow voltage to build on the generator rotor shaft. Inadequate grounding may be due to: a problem with the shaft grounding brushes from wear (requiring replacement brushes) or a problem with the associated shaft grounding circuitry if a monitoring circuit is provided. High shaft voltages can also be caused by severe local core faults of large magnitude, which impress voltages back on the shaft from long shorts across the core. Protection against shaft voltage buildup and current discharges is provided in the form of a shaftgrounding device, generally located on the turbine end of the generator rotor shaft. The most common grounding devices consist of a carbon brush or copper braid, with one end riding on the rotor shaft and the other connected to ground. 5-118
Turbine-Generator Condition Assessment
Shaft voltage and current monitoring schemes are also provided in many cases to detect the actual shaft voltage level and current flow through the shaft grounding brushes. This has the advantage of providing warning when the shaft grounding system is no longer functioning properly and requires maintenance. There are numerous monitoring schemes available and each OEM generally has its own system provided with the TG set when purchased. For older machines with only grounding and no monitoring, a monitoring system can usually be retrofit to the existing ground brushes. The OEM should be consulted when upgrading the shaft monitoring.
5.18 Excitation System The primary purpose of the excitation system is to provide excitation to the main generator. The main components of any main generator excitation system are: a power supply, a voltage regulator, and a generator field. The power supply, which can be electrical or mechanical, provides energy to drive the excitation. The electrical supply can be either a transformer taking power from the generator bus or a station auxiliary bus feeding a motor generator set exciter. The main generator field will convert the current supplied by the excitation system into a magnetic field. The lines of magnetic flux from the rotating generator field cut through the stator-mounted armature winding and induce an ac voltage in the windings. To control the current supplied to the generator field, a voltage regulator is used. The control of the field may be direct or indirect. An indirect approach controls the field current, and a direct approach controls the actual generator field current. There are other components in addition to the three mentioned above; they are discussed in the EPRI report Tools to Optimize Maintenance of Generator Excitation System, Voltage Regulator, and Field Ground Protection, 1004556 [46]. This report was developed because the cost of lost generation can greatly exceed the cost of repairing the excitation system. A low-cost high-benefit ratio is needed to maintain and improve reliability and availability with maintenance budgets the way they are in a plant provided will help plants benefit economically by using appropriate levels of PdM and PM tasks through equipment upgrades. For more information on generators, see the EPRI Power Plant Electrical Reference Series: Volume 1, Electrical Generators, EL-5036-V1 [47]. Each of the volumes in this series provides comprehensive and practical information regarding electric power apparatus and electrical phenomena. Volume 1 presents various excitation systems and their effects on generation operation for overexcited and underexcited field conditions. It also describes the basic construction of generators and information concerning the units. An EPRI report to be issued in January 2006, Excitation System Retrofit and ReplacementLessons Learned, 1011675, [60] will provide guidelines on avoiding technical and project management mistakes in the procurement and retrofit process. This guide will also document the best practices in replacing excitation systems in hydro, fossil, and nuclear plants. A generic specification will be published as part of the guide.
5-119
Turbine-Generator Condition Assessment
5.19 Using Data on Condition Assessment to Assess Risk of In-Service Failure As maintenance intervals increase, parts remain in service for longer periods, and the chances for inspection to reveal forms of damage increase. The data obtained on the condition of parts in the turbine or generator may be assessed either deterministically or probabilistically. In either approach, the basic objective is to calibrate the significance of wear or damage, as characterized by the type of NDE test, the strengths and limitations of which have been previously reviewed. A deterministic approach produces a result (in terms of time or cycles) that is subjective to the user’s selection of discrete input or factors. A probabilistic approach relies on statistical distributions to describe these same factors and, by their random combination, estimates the chances of a failure over the same scale of time or cycles. Examples of each are illustrated in Volumes 6 and 7 to guide the inspection and replacement of HP, IP, and LP blades Each approach has merits. A deterministic assessment using NDE data may weigh how much conservatism or factor of safety exists for a given component for a consistently applied operating scenario. It can identify where the weakest point occurs in a structure, which will have the least amount of tolerance to damage. For components where there is a vibratory stress imposed on a steady load (like rotating blades), it can be used to predict the point at which high-cycle fatigue will assume control of crack growth, leading to fracture. A probabilistic approach recognizes that the real world involves uncertainties: for example, variation between parts, the accuracy of different NDE techniques, the inherent variability in tolerances, material properties, and the operating stresses that occur within a part. It is generally a more valid method for weighing the options of run, repair, or replace that are faced by a maintenance engineer when inspection reveals damage that can be tolerated for a limited period. The fundamentals in a risk assessment are summarized as follows: •
Every component has inherent strength, selected by design to withstand an expected range of stress.
•
Conventional remaining life formulas use discrete terms or values to represent properties. Design formulas often include factors of safety.
•
In practice, both are random quantities. If applied in the life prediction formulation, the results generally produce a probability distribution as shown in Figure 5-46, where failure versus no failure is defined by the limit state function.
•
The population to the right of the limit state function has a low risk of failure.
5-120
Turbine-Generator Condition Assessment
Figure 5-46 Example of a Probability Distribution and Limit State Function
Component life is evaluated in two ways; (1) the number of cycles or time to initiate cracks, and/or (2) the remaining cycles or time for cracks to propagate to their critical size. As noted, if there is a dynamic stress present, then this needs to be factored into the assessment. In a probabilistic treatment of NDE data, each mechanism is evaluated using the same basic approach. As shown in Figure 5-47, results are produced by first coupling material properties (obtained from published specimen tests results) with field measurements (obtained during the outage). The operating history is represented in terms of start-stop cycles, hours in service or years of service, depending upon the type of damage mechanism that is being evaluated. For example, low-cycle fatigue or stress corrosion cracking would be evaluated on a scale of startstop cycles or years of service, whereas creep would be based on hours in service at high temperature. High-cycle fatigue life is generally measured in hours or days, particularly when dealing with turbine blades.
5-121
Turbine-Generator Condition Assessment
Figure 5-47 Basic Elements of a Probabilistic Analysis
In any assessment, stress is typically the most critical factor that is unknown to the operator. Unfortunately, until faced with a problem, many operators do not learn that it is often difficult or impossible to obtain stress from the designer and that independent analysis requires dimensional details from the component and time to perform the calculations. It is therefore becoming standard practice among many plants to reverse-engineer parts, such as blades, that they expect to replace or to maintain spares even though they may still rely on the OEM as a supplier. With this data in hand, the simulation is performed in advance of the outage so that the distribution of stress throughout the component is available prior to the outage. When each of the factors has been described, a Monte Carlo simulation is performed, which essentially means that the input factors are randomly combined within the selected life consumption formula. NDE data may be processed directly when it is available. Prior to an outage, incremental ranges of indication/crack sizes can be processed. These results are plotted either on a log scale or linear scale as a family of curves to show the relative change in risk that occurs as the size of an indication is increased and refined as necessary when actual data are taken.
5-122
Turbine-Generator Condition Assessment
To illustrate this approach and its application in condition assessment, assume that the previously mentioned technique in remote optical examination of the first row HP buckets was performed during a boiler inspection. A noticeable degree of solid particle erosion on the leading and/or trailing edges of the blades was identified. To make a quick determination of the risks involved in running the row in its present condition, a family of probability curves was produced in advance of the inspection As shown in Figure 5-48, these individual curves relate the risk of a failure to incremental notch sizes of 50–350 mils (1.27–8.89 mm). They are based on stress results obtained from a finite element analysis, using an erosion rate that was determined from the inspection records prior to when notches became visibly apparent. A hypothetical distribution of NDE results was applied to produce the initial set of curves that was easy to update if necessary when the actual results became available.
Figure 5-48 Example of SPE Inspection Criteria Using Series of Probability of Failure Curves
Inspection revealed no cracks, but random patterns of notch wear ranged from 90 to 225 mils (2.3 to 5.7 mm) on the trailing edge. The maximum size notch of 225 mils (5.7 mm) measured after six years is plotted to show the present status of the row. The change in probability is reflected for each subsequent year of continued operation, extended to the next six-year interval. In the example, the probabilities are also categorized using a hazard risk assessment classification matrix originally produced for the Department of Defense (MIL-STD-882C 5-123
Turbine-Generator Condition Assessment
“Military Standard System Safety Program Requirements,”) [48]. This classification can further facilitate the interpretation of results, beyond a comparison in the relative change in odds. When reflected in this manner, it quickly becomes apparent that the operator has a powerful tool by which the NDE data reflecting the present condition of the component can be assessed and extrapolated to support a run, repair, or replacement strategy. If combined with the costs of doing the maintenance versus deferring the maintenance, the financial outcome can further assist in identifying what the optimum time would be to replace the damaged blades. The following is noted regarding risk assessment as it relates to turbine-generators: 1. Many commercially available programs have evolved to the point where it is not the mathematics that limits the effectiveness or value of probabilistic applications. 2. The most important ingredient in any assessment is a fundamental understanding of the damage and/or failure process that is being evaluated. The probabilistic model needs to focus only on those factors that are relevant. 3. Any risk assessment should focus on individual types or forms of damage, rather than trying to treat all potential contributors/mechanisms together. The types of damage that affect turbine-generator components are varied, may occur at different times within the start-stop cycle, and affect different locations on the component. 4. A well-planned model should be capable of running millions of studies in a matter of minutes. This can allow the operator to consider “what if” scenarios in which the input parameters can be varied to test the sensitivity of the projected risk with regard to key assumptions. 5. Since the approach is meant to be component specific, valid stress results must be available as input to the probabilistic model. These results should be derived from finite element analysis, not design formulas that tend to approximate and/or include factors that do not reflect the stress field in regions of concentration (where damage naturally tends to form). A competently performed analysis should be able to produce very reasonable values of stress as input to the model, that is, with a minimum amount of uncertainty. 6. Most turbine-generator components rely on conventional materials that are well understood in terms of their mechanical behavior and for which there is ample published information to develop a statistical basis for input to the model. As more data are used in the model, the degree of uncertainty associated with this key parameter is reduced. It is important, however, not to mix data taken from different alloys or test conditions, but to segregate the information appropriately. 7. In terms of affecting the projected risks, the greatest source of uncertainty is often associated with the NDE data and how it is characterized. As a rule of thumb, the uncertainty increases and confidence decreases as the size of the indication approaches the detection limits of the sensor.
5-124
Turbine-Generator Condition Assessment
8. At present, most sensors cannot accurately characterize indications with any degree of certainty below 30–50 mils (0.762–1.27 mm). To compensate for this, many risk analyses presume damage exists of a magnitude just below this detectable threshold. It should be noted, however, that for some mechanisms (SCC and LCF), this assumption might represent a significant acceleration in the actual rate of crack formation and growth and thereby artificially increase the projected risk of failure. This is particularly true if the damage is occurring in regions of low stress. The last point is worth some further discussion. Figure 5-49 represents an example published by an OEM [49] in which test indications obtained by UT were compared via destructive analysis. The plot shows the ratio of measured sizes, based on UT, versus actual sizes. The range of scatter reflects the increased statistical uncertainty that would be introduced if used as input to a probabilistic analysis. The statistical uncertainty becomes more notable as the cracks approach the detection limits of the NDE technology. For application in a risk assessment, it is therefore important to scrutinize data obtained on the condition of a part to ensure that it is a valid representation of its actual condition.
Figure 5-49 Ratio of Actual Crack Sizes to Measured Crack Sizes
5-125
Turbine-Generator Condition Assessment
Additional information is provided in Volumes 6 and 7, the EPRI report Steam Turbine Disk Brittle Failure – Influencing Parameters and Probabilistic Analysis Demonstration, 1003264 [50]. The report describes a methodology for assessing the probability of disk brittle failure due to stress corrosion cracking. The probabilistic method features a finite-element-based approach to calculate stress intensity as a function of crack length for arbitrary crack geometries. The importance of key factors governing the probability of failure is demonstrated through a parametric study.
5-126
6
OIL FLUSHING
Oil flushing often represents a significant period of critical path time. The information presented within this guideline is primarily designed to identify and describe methods or approaches that have been used to perform these operations effectively. Methods, tools, and practices that can accelerate the flushing procedure are discussed for various systems found on different types of units. Practical approaches or modifications to present systems that would allow bearings to be isolated and drained on an individual basis are also discussed. This section of the guidelines identifies, reviews, and compiles practices and techniques that are normally involved or should be undertaken during the flushing and replacement of the turbinegenerator lubricants.
6.1
Preparations and Precautions for Flushing the System
Because the turbine lube oil is often contaminated during maintenance inspections, the oil system must normally be flushed. To flush the turbine lube oil system, the oil velocity must be increased, and the oil must be heated, cooled, and filtered so that contaminants can be removed from the system. The following information describes the necessary changes that must be made before the start of a turbine lube oil flush. 1. At the lube oil reservoir, perform the following activities before flushing: a. Drain and clean the lube oil reservoir. b. Remove and clean the oil cooler tube bundles. c. Install one clean oil cooler tube bundle in the cooler. d. Make provisions to heat and cool the oil in the oil cooler with the tube bundle installed. e. Set the oil cooler transfer valve to the cooler with the bundle installed. f. Install a fine mesh screen on the bayonet screens in the return oil sump in the reservoir. g. Fill the lube oil reservoir with clean oil. h. Make provisions to monitor the electrical loads on the lube oil pump motor during the flush.
6-1
Oil Flushing
i. On units with bearing oil relief valves, increase their setting so that they do not open during the flush. j. Clean and check all reservoir door seals. k. On units with temporary oil filtering skids, connect the skid to the reservoir, install new filtration cartridges, and make the appropriate electrical connections. 2. At all steam control valves for mechanical hydraulic control (MHC) units: (stop, throttle, control, governor, reheat stop, and intercept), perform the following activities before flushing: a. Disconnect the oil feed and drain lines to the servomotors. b. Install bypass jumpers from the oil feed line to the oil drain line. c. Install blanks on the servomotors where the oil lines were disconnected. 3. At the front pedestals, perform the following activities before flushing: a. Clean the interior of the front pedestal. b. Install a bypass jumper on the journal bearing (feed to drain) with the valve in the line. c. Install bypass jumpers on any mechanical components in the front pedestal, for example, gear feed lines, zero speed switches, and safety bearing feed lines. d. Install the pedestal cover. 4. At the journal and thrust bearings, perform the following activities before flushing: a. Install bypass jumpers on the bearing (feed to drain) with the valve in the line. b. Remove the strainer/orifices from the oil feed lines. c. Label the strainer/orifices and seal in plastic bags. d. Clean all bearing pedestals and install the pedestal covers. 5. At the thrust bearing wear detector, install bypass jumpers on the thrust bearing wear detector (feed to drain). 6. At the turning gear, install a bypass jumper on the turning gear (feed to drain). 7. At the generator, perform the following activities: a. Install blanks in the hydrogen seal feed and drain the line connections to the end bells. b. Install jumpers in the hydrogen seal (feed to drain) lines with the valve in line. 6-2
Oil Flushing
c. Remove the bearing oil feed line to the end bell, and cover the ends with plastic. d. Install a bypass jumper on the bearing (feed to drain) with the valve in line. 8. At the exciter, perform the following activities: a. Remove the bearing oil feed and drain lines, and cover the ends with plastic. b. Install a bypass jumper on the bearing (feed to drain) with the valve in line. The bypass jumpers allow the oil to bypass the bearings and mechanical devices during the oil flush. The valves in the bypass lines allow the oil flows to be maximized in each section during the flush. All ac oil pumps are operated during the flush to increase the oil velocities. The bypass jumper line valves are used to keep from overloading the oil pump motors. During the oil flush preparation period, all blanks that have been installed are recorded for future reference. All strainers, orifices, and piping that have been removed are to be kept in a central location to be installed at the completion of the flush.
6.2
Resources That Should Be Available While Flushing
During a turbine lube oil flush, resources are needed to ensure that the flush is completed in the shortest time possible while performing the flush in a safe manner. Lube oil is highly flammable, and the turbine should be kept as clean as possible during the flush. The system should be constantly monitored during the oil flush to verify that oil does not leak at any of the penetrations at the reservoir, pedestals, end bells, valve servos, or any of the temporary connections. Lube oil flushes are performed around the clock until the flush is completed. Continuous flushing requires resources to be available at all hours. During the flush, mechanics are required to: •
Monitor the oil system
•
Check motor loads
•
Make valve changes
•
Clean screens
•
Change filter elements
•
Stop oil leaks if found
The number of mechanics that are needed will depend on the size of the unit being flushed. Generally, two mechanics per shift are sufficient for a small turbine, and four mechanics per shift are sufficient for a large turbine.
6-3
Oil Flushing
A chemist is required during the oil flush to perform oil sampling and oil cleanliness tests. The chemist can be one of the station chemists, or an outside lab can perform the testing. The oil samples must be taken at regular intervals to determine the progress of the flush. One chemist must be available on each shift. The chemists should coordinate the testing so there is no need for their continuous involvement in the flush. Station operators are needed to •
Start and stop the lube oil pumps
•
Coordinate the heating and cooling of the oil
•
Give advice on the operation of the system
The operator might also choose to monitor the oil pump motor loads. One operator must be available on each shift. The control room operator might choose to have an auxiliary operator also available to perform some of the flushing duties. A field engineer (manufacturer’s representative) is required on each shift to direct the mechanics, chemists, and operators. The field engineer (technical director) determines the period for each segment of the flush, monitors the oil cleanliness, and coordinates the craft. The field engineer is usually the person responsible for the coordination of the entire oil flush. The field engineer is familiar with the turbine lube oil, control oil, and generator hydrogen oil systems. The field engineer knows the proper methods for setting up the flush, performing the flush, and returning the oil system to its proper operating condition after flushing is complete. If a union hall supplies the craft performing the flush, a supervisor to work with the craft will be required on each shift. Depending on the station, the supervisor may be able to perform the duties of one of the other craft involved with the flush.
6.3
Precautions While Flushing
During the oil flush, special precautions must be taken to ensure the safety of the personnel and the cleanliness of the oil system. Serious damage can occur if the lube oil catches on fire. Many plants have had serious oil fires that have resulted in total destruction of the turbine-generator set. The following precautions are to help the maintenance personnel keep the site safe while cleaning the turbine oil systems: •
Avoid any burning, welding, smoking, or open flame in the turbine areas during the flush. Hot oil vapors will ignite when exposed to open flame. Post signs around the turbine to this effect to alert others who might not be involved with the flush.
•
Do not exceed the nameplate rating on totally enclosed fan-cooled or explosion-proof motors. Open drip-proof motors may be operated continuously at 15% above rated amps for flushing.
•
Verify that all temporary hoses are secured before starting the pumps.
6-4
Oil Flushing
•
Do not remove the plastic pedestal covers during the flush unless necessary. These covers keep oil vapors in and dust and dirt out.
•
Make certain that extra fire extinguishers are readily available around the turbine, valves, generator, and oil reservoir.
•
Establish the proper emergency communications with the control room in the event that an emergency shutdown of the pumps is required.
•
When flushing the generator seal oil system, make sure that no oil backs up into the defoaming tanks.
•
Do not allow water to leak into the oil system during the flush. Water can cause oxidation in piping and pedestals and will contaminate the oil system.
•
Be careful while working around hot oil and hot water. During the flush, the oil must be heated to loosen contaminants from the system.
•
Limit the hot water to 190°F (87.7°C) for units with cast-iron water box heads and to 200°F (93.3ºC) for units with steel water box heads.
•
Ensure that all connections, flanges, etc., are thoroughly sealed to keep oil from leaking from the system during the flush.
•
Check for oil spills each time the oil pumps are started.
•
Have operators and crew communicate by using two-way radios during the flush. This will allow the pumps to be shut off quickly in case of an oil spill or an emergency.
•
Monitor oil filter differential pressure to keep the filter cartridges from collapsing. High differential pressures can develop quickly during an oil flush.
•
Monitor the return screens in the oil reservoir to keep oil from overflowing during the flush. Monitor oil levels in the pedestals to keep oil from overflowing at oil seals. Throttle the oil flow valves as needed in bearing bypass jumpers to keep the oil from overflowing at the pedestals.
•
Monitor the oil contamination level at regular intervals to complete the flush in the shortest time possible.
•
Flush the components in the correct sequence. Do not allow dirty oil to enter areas that have been previously flushed.
•
Always use lint-free rags when cleaning components. Lint from the rags can contaminate previously cleaned areas.
6.4
Oil Cleanliness Criteria
The cleanliness of lube oil systems can be measured using different methods. The approved method for Siemens-Westinghouse turbines is to circulate oil through a filter for 30 minutes. The filter is then taken to a chemistry lab and analyzed for contaminant levels. The approved method for General Electric turbines is to collect a 100-ml sample (grab sample) in a bottle from various locations. The samples are taken to a chemistry lab and analyzed for contamination levels. 6-5
Oil Flushing
The method for collecting samples using the filter method is as follows: 1. Protect the sample location from surrounding contaminants. 2. Isolate the strainer by closing the ball valve. 3. Rinse the strainer housing with solvent. 4. Open the strainer petcock, and drain the strainer housing. 5. Open the strainer housing. 6. Rinse down the strainer housing, and cap the interior with solvent. 7. Insert a clean sample screen (150 mesh). 8. Close the container. 9. Close the drain petcock. 10. Open the isolation ball valve. 11. Run the oil sample flow through the strainer for 30 minutes. 12. Close the isolation ball valve. 13. Open the drain petcock. 14. Rinse down the strainer housing with solvent before opening. 15. Open the housing, and remove the sampling strainer. 16. Place the sampling strainer into a clean container. 17. Close the strainer drain petcock. 18. Close the strainer housing. 19. Open the isolation valve to allow continuous flow. 20. Transport the containers to the designated location for sample counting and analysis. The method for collecting samples using the grab sampling method is: 1. Remove the cap and plastic film from a sample bottle. 2. Dip the container into the oil volume, or hold it under a flowing stream of oil. 3. The sample locations are at the pump level in the oil tank, in the detraining section of the oil tank, at the oil purification discharge in the main oil tank, and at the bearing header at the front standard. 6-6
Oil Flushing
4. Replace the film and cap on the bottle. 5. Wipe the outside of the container, and transport the containers to the lab for analysis. Determining whether the oil meets the cleanliness criteria after using the filter sample method is: 1. Wash the sampling strainer with a clean fluid, collecting all residue on a 200-mesh filter into a vacuum flask. 2. Remove the filter membrane and, using a 10X-scaled magnifier, scan the filter to determine the particle size and the number of particles in the 0.005–.010" (0.127–0.254 mm) range. 3. Do not attempt to move or rotate the particles while scanning. 4. Acceptable criteria are: no hard particles above 0.010" (0.254 mm) are allowed and less than five hard particles in the 0.005–0.010" (0.127–0.254 mm) range. 5. Soft particles may exceed the above criteria and are not considered harmful. 6. Soft particles can be readily crushed between the fingers and include lint, paper, sawdust, asbestos or other insulation, and tobacco. 7. Label and date all samples removed from the system to be compared with samples taken at a later date. The cleanliness criteria for the grab sampling method are summarized in Table 6-1. Table 6-1 Recommended Cleanliness Criteria Preferred
Maximum Allowed (Acceptable)
0.005–0.010 mm – less than 32,000
0.005–0.010 mm – less than 128,000
0.010–0.025 mm – less then 10.700
0.010–0.025 mm – less then 42,000
0.025–0.050 mm – less than 1,510
0.025–0.050 mm – less than 6,500
0.050–0.100 mm – less than 225
0.050–0.100 mm – less than 1,000
0.100–0.250 mm – less than 21
0.100–0.250 mm – less than 92
Greater than 0.250 mm – None
Greater than 0.250 mm – None
Shown are number of particles per 100 ml sample of oil
The procedure for counting the number of particles in the oil sample is outlined in the Society of Automotive Engineers, ARP-598A [51]. This procedure says to filter a given volume of oil through a membrane and to count and size the particles deposited on the membrane by using a microscope. Many independent testing labs perform oil analyses, and most large oil companies perform oil particle counts. The acceptable range for the maximum number of particles is similar to the SAE Class 6, and the preferred range for the maximum number of particles is similar to SAE Class 4 contamination levels. 6-7
Oil Flushing
A comparison of the cleanliness levels of turbine oil is provided by the International Organization for Standardization (ISO). The Solid Contamination Code used by ISO is assigned based on the number of particles greater than 5 microns per unit volume and greater than 15 microns per unit volume. ISO Code 16/13 is approximately equivalent to in service oil cleanliness recommendations provided by turbine manufacturers.
6.5
Heating and Cooling the Oil Without Damaging the Bearing
The lube oil should be heated during the flushing operation to increase the flow of the oil and to thermally shock the lube oil piping. The higher oil flow will allow turbulent flow to occur, which will help to dislodge debris that is lodged in the lube oil piping. The heating and cooling of the lube oil piping thermally shocks debris from the piping and allows it to be carried downstream into screens and filters where the debris can be removed. The heating of the oil must be done carefully to keep from damaging the oil or the oil system. During the flush, the turbine lube oil is usually heated in one of the lube oil coolers located at the oil reservoir. The oil coolers use cold water to cool the lube oil during operation of the turbine. One of the coolers should have cold water flowing through it and the other should have hot water flowing through it. This will allow the use of the cooler transfer valve to transfer from hot oil to cool oil during the flush. The methods used for heating the lube oil are: 1. Use hot water from an auxiliary supply. 2. Mix steam and water in a closed system, and then pipe the hot water to the cooler. 3. Mix steam and water in an open system, and pipe the hot water to the cooler. 4. Immerse cal rods in the lube oil reservoir. 5. Immerse a coil in the oil reservoir, and pass hot water through it. Some power stations have the ability to make hot water to use for heating. This hot water can be piped into the lube oil cooler heads to heat the lube oil. Care should be taken to ensure that the oil is heated in the prescribed manner to keep from damaging the oil. Steam and water can be mixed in closed or open systems to supply hot water to the oil cooler. If steam is used to heat water, care must be taken not to overpressure the system or to alloy steam into the oil cooler. Cal rods can be used to heat the oil directly by immersing them in the oil reservoir. These heaters must be the correct types to heat the oil properly. The main problem with the cal rods is that a suitable power supply is not always available.
6-8
Oil Flushing
The last method for heating the lube oil is to immerse a coil into the reservoir and pass hot water or steam through the coil. The size of a coil that is large enough to heat the oil is usually too big to fit into the oil reservoir through the openings at the top of the reservoir. The oil temperature should be brought up to 180°F (82.2°C) when heated. When using water to heat the oil, the water to oil temperature differential should be limited to 100°F (37.7°C) maximum. For coolers with steel water box heads, the water temperature should not exceed 200°F (93.3°C). For coolers with cast-iron water box heads, the water temperature should not exceed 190°F (87.7°C). The preferred method of heating oil is to pipe hot water to the oil cooler and use the cooler to heat the oil. The oil may be heated in one cooler while the system is being flushed using the other heater.
6.6
Minimizing the Use of External Piping While Flushing
Lube oil system cleanliness is essential after a turbine-generator outage and before the turbine is put on turning gear. The ideal lube oil supply has full flow filters installed that filter the lube oil going directly to the bearings. The only unprotected piping that must be cleaned is located between the filter and the bearing. A modified flush can be completed using the turning gear oil pump (TGOP), orifice strainer toggle blanks, valve, connecting pipe, hose, and 100 mesh filter bags. A modified TGOP flush can be completed as bearing standard assemblies are being completed and readied for closure. Orifice strainer toggle blanks are installed in all bearing locations. The orifice strainer toggle blank will prevent oil flow to a bearing and standard when the bearing is unassembled but will allow flushing in another area. The oil piping to the completed standard is flushed using the TGOP with either the orifice strainer in place or the toggle blank installed and unseated. The assembly shown in Figure 6-1 is also installed to allow oil flow through the piping and bearing lower half. A 100-mesh bag, acting as a tell-tale, is added to the end of the drain hose. The 100-mesh bag is the indicator for a clean line.
6-9
Oil Flushing
Figure 6-1 Oil Flushing Piping
The example provided references a system with full-flow filtration before the bearings. The bearings have temporary piping coming out at the horizontal joint to which is attached a 100mesh filter. A modified oil flush can be done without having this full-flow filtration before the oil inlet to the bearings.
6.7
Flushing Without an External Filter
During maintenance inspections, turbine lube oil flushes are often performed without the use of external filters. Flushing the turbine without external filters requires more time to clean the oil system, but it can be less expensive than with external filters. If external filter canisters are not used, some way to remove contaminants must be provided. In the past, fine mesh screen (100 mesh) has been used to catch the contaminants from the lube oil as the oil passes through the bearing strainers. This method of cleaning the oil is not effective because it reduces the volume of oil flowing through the piping. A more effective way to remove the contaminants from the oil is to install a bypass jumper around the bearing to increase the volume flow rate and to install a fine mesh filtration bag on the jumper where it exits the pipe in the bearing cavity. This method allows the flushing of a few bearings at a time, which allows the flows to be turbulent in the piping section that is being flushed. The filtration bag will have sufficient surface area to allow high volumes to flow while removing the contaminants from the oil. These filter bags can be made of nylon or cloth. The detraining section of the oil reservoir is also an area where contaminants can be recovered during an oil flush. The oil in the detraining section must pass through bayonet screens on its way back to the oil pumps. Fine (100 mesh) mesh should be placed on these bayonet screens during the oil flush to capture the large contaminants in the oil. The screens should be checked periodically to ensure that the oil does not flow over the screens as they become clogged with 6-10
Oil Flushing
particulate matter. It is recommended that the oil reservoir have two bayonet screens at the detraining section. This allows the oil to always have one fine screen to pass through while the other screen is being cleaned. One of the most important steps in the oil flush is to drain the oil from the reservoir and thoroughly clean the reservoir before beginning the flush. The oil reservoir has many areas where contaminants settle. During the flush, oil flows are increased, and these contaminants become suspended in the oil. As these contaminants travel throughout the oil system, they allow clean areas to become contaminated. Oil coolers often trap contaminants and should be cleaned before flushing. After the oil reservoir is cleaned, it should be filled with clean, filtered oil. The turbine oil that comes from the vendor frequently does not meet the cleanliness criteria and must be filtered prior to filling the reservoir. It is important that the oil purification system be placed in service during the flush to clean the oil in the reservoir. Some owners have permanently removed the bowser from the unit and have installed a kidney loop oil filtration system, which provides superior filtration of the oil. These kidney loop systems also remove more water from the oil than the bowser. The oil reservoir has doors with rubber gaskets that seal the doors to keep contaminants and water out of the oil. These gaskets should be cleaned and checked to ensure that they are working properly. The vapor extractor should be in operation during the oil flush to allow the oil system to work properly. Some turbines require the vapor extractor to be in operation when the oil pumps are on to keep from overflowing the bearing pedestals. Vapor extractors are used to keep moist air from contaminating the lube oil.
6.8
Techniques to Get Maximum Flow Through Piping
In Section 6.1, there was much discussion about the preparations that must be performed for a lube oil flush. An emphasis on the preparation for the oil flush discussed bypassing the components to enable getting a higher oil velocity in the pipe being flushed. The components that were bypassed had bypass jumpers installed from the feed lines to the drain lines (oil sump in the pedestal). The reason to bypass the components was to keep contaminants out of the clean component and to increase the flow at that location. The components that had bypass jumpers installed were: •
Bearings
•
Thrust bearing wear detector
•
Turning gear
•
Valve servos
•
Front pedestal mechanical components
•
Generator hydrogen seals
6-11
Oil Flushing
The bypass jumpers allow the oil to bypass the bearings and mechanical devices during the oil flush. The sequential valving of the bypass lines allows the oil flows to be maximized in each section during the flush. During the flush, the ac oil pumps on the reservoir are operated to increase the oil velocities. The bypass jumper line valves are also used to keep from overloading the oil pump motors. Another way to increase the flow of oil in the lube oil system is to install a supplemental high velocity flushing pump on the system. The supplemental oil pump is sized according to the main bearing feed header diameter. These pumps can deliver 2,000–5,000 gallons (7.57–18.93 kl) of oil per minute to allow the oil flow to become turbulent in the piping. The supplementary oil flushing pumps are necessary when the auxiliary oil pump motor is not large enough to flush the main bearing feed header. If the auxiliary oil pump does not have enough power to deliver the required flow at the proper pressure, the supplemental oil pump will produce a moderate reduction of the time required to perform the oil flush. The engineer must determine when it is economical to use a supplemental oil flushing pump. If a supplemental pump is required, the desirable type of pump is an electric-driven centrifugal pump. These pumps get their suction from the bottom of the oil reservoir and discharge into the top of the reservoir. There is a gate valve on the pump suction and discharge line and a bypass line around the pump to control the flow rate. On General Electric turbines, when the oil-driven booster pump is removed, the pump discharge can be connected to the plate where the booster pump is mounted. Two bypass lines with throttle valves are also connected to the booster pump plate. When using a supplemental pump, all connection piping must be cleaned before being connected to the reservoir. Flexible stainless steel connection piping is often used to connect supplemental pumps. The oil cooler maximum pressure must not be exceeded when using a supplemental oil pump. This pressure can be found on the cooler nameplate. One of the oil coolers will need to have the internals removed and blanks installed to allow for high oil flow through the cooler. Care must be taken to ensure that there are no leaks at the pump, reservoir, cooler, or piping when using a supplemental pump. Filtration skids are often employed when using supplemental pumps to allow for faster cleaning of the oil. These skids have two canisters that have differential pressure gauges installed to monitor the pressure differential across the filter. Many spare filter cartridges are needed when using filter skids.
6-12
7
ROTOR ALIGNMENT AND BALANCING
During reassembly of the turbine-generator, significant delays can be avoided by the way in which the engineer treats the massive amount of information needed to align bearings and couplings and to rebalance the rotor. Lack of semi-automated or automated means to collect, process, and relate clearance or alignment measurements to the original design specifications can cause problems to be missed until the process of turbine assembly begins. Inefficient management and processing of tight wire information can further contribute to lost time finding internal alignment and balancing solutions. This section of the guidelines presents a generic discussion of the advantages and limitations of different alignment techniques and practices currently applied within the industry. Included in this discussion is a review of techniques for automating the alignment process and the requirements for their application. The discussion proceeds to issues and techniques for automating the alignment process and the requirements for their application. The discussion continues with issues and techniques for balancing the rotor and criteria that can be used to decide allowable vibration limits. An alignment and a balancing primer are available in Volume 3 of this series. The alignment primer covers both coupling and tight wire alignment problems in detail. The balancing primer includes a discussion of issues associated with turbine-generator vibration diagnostics related to balancing. Also included in this series is guidance in the TGAlign (English and SI units versions) software developed by EPRI that optimizes the coupling alignment with a minimum number of bearing moves or no bearing moves. The EPRI report Shaft Alignment Guide, TR-112449, [52] has been developed to provide information on the fundamental causes and effects of misalignment on machinery and the fundamentals of shaft alignment.
7.1
Different Tight Wire Techniques
When performing internal alignment on turbine-generators there are four methods used to record alignment data: •
Tight wire
•
Arbors
•
Precision optical scopes
•
Helium Neon (He Ne) lasers 7-1
Rotor Alignment and Balancing
The tight wire method has been used for many years and is a method that is often used for turbine alignment. Many millwrights are familiar with this process, and many power stations have the special brackets and tools needed to perform tight wire alignment. When the tight wire method is used, a wire is set to special set points at the ends of the turbine casings. The wire is fixed at one end and has a weight suspended from the other end. Sag charts have been developed for specific wire diameters and wire weights. Most turbine maintenance engineers are trained in the tight wire alignment method and can supervise the data collection with ease. The limitations to using a tight wire for turbine component alignment are: •
It is easy to accidentally move the set points.
•
The wire sag must be compensated for.
•
The wire must be moved when installing lower components.
•
It is difficult to get tops-on readings.
The readings obtained with a tight wire are accurate within 0.001" (0.0254 mm), which is within the tolerance for alignment. When the wire is bumped, the set points must be checked to ensure that the wire has not moved. It is very easy to accidentally hit the wire and move the set point. During the tight wire alignment process, much time is spent setting the wire to the set points. The wire sag must be compensated for and can cause errors in the alignment process. Any time a lower component is installed, the wire must be removed and set up again after installing the component. If there are many component moves, the wire setup time can be a considerable amount of the alignment time. When it is necessary to take tops-on readings, it is difficult to get inside the casings to take the wire readings without moving the wire set points. There is also a limit to the number of internal components that can be installed during the topson readings when using a wire. The person taking the data must have enough room to take the wire readings, which is difficult with small diameter components. Another method of internal alignment data collecting is arbor alignment. This method requires an arbor be built specifically for the turbine casing that is being aligned. The arbor can have dial indicators mounted to it, or it can have proximity probes mounted to it to collect the alignment data. The preferred method for collecting data when using an arbor is to use proximity probes. The arbor with proximity probes allows for easy data collection when performing tops-on alignment. Limitations to using an arbor are: •
The arbor sag must be compensated for.
•
The arbor must be moved when installing lower components.
•
The arbor is heavy if not made from tubing or pipe.
7-2
Rotor Alignment and Balancing
•
Fabrication of an arbor requires a significant amount of time and may impact the outage critical path time if done during an outage.
•
Arbors are expensive.
•
An arbor may require modification for each turbine casing.
Data collection goes very quickly when using an arbor and proximity probes. Arbors can be saved for future use and, in some cases, can be modified to be used on more than one turbine casing. Precision optical scopes have been used for internal alignment with limited success. These scopes require special targets be made to collect the alignment data. There are limited resources available when using optical scopes, and this method is not very popular for large steam turbines. There is no need to compensate for sag when using an optical scope. The limitations for using an optical scope are: •
Limited resources are available.
•
They need a highly trained operator.
•
Special targets are required.
•
They are expensive.
•
They are affected by heat.
•
They are delicate and can be damaged easily.
Helium neon lasers are constantly being developed and improved for use in internal turbine alignment. Lasers are becoming the preferred method for internal alignment because they are very accurate and can shorten the time for performing the alignment. Lasers can be used to take flatness and perpendicularity measurements that are not possible with other methods. The laser can take joint flatness readings, casing fit readings, and radial position readings. Lasers can be set up for short or long distances for internal alignment purposes. The readings are easy to take once the laser is set up. There is no need to compensate for sag when using a laser. There is no need to move the laser when installing lower components, so alignment moves can be performed much faster. The limitations for using a laser are: •
Limited resources are available.
•
They need a highly trained operator.
•
Special targets are required.
•
The laser light can be deflected by heat and smoke.
7-3
Rotor Alignment and Balancing
7.2
Information Collected from the Unit
A comprehensive slow-speed balancing package should contain a complete vibration history and a record of the vibration readings before and after each in-service balance shot. Also, the type of shot—static or couple—should be recorded. In-service balance shot data provides specific information for each rotor balanced. The sensitivity and high spot numbers can be developed for each rotor. Rotor critical speeds should also be logged during startups. Rotor run-outs should be maintained, especially for rotors that have increasing bows. The run-outs can be plotted to provide a time-effect change of the run-out and assist in planning for refurbishment or replacement. Figure 7-1 shows a 10-year tracking of rotor bowing.
Figure 7-1 Ten-Year Record of Rotor Bowing
Rotor work done during the outage should be included in the balancing reference plan. The following is a typical slow-speed balancing process: 1. Determine the rotor weight, location of balancing planes, unbalance tolerances, etc., before slow-speed balancing. 2. For each balance plane, record the balance weight mass, angular location, and balance plane location for all (both factory and field) previously installed balance weights. Leave the factory weights in place, but remove and resolve the field balance weights. 3. Prepare the couplings for slow speed balance by measuring the coupling bolt hole fit diameters and, if the hole fits are not within 0.003" (.0762 mm) of the smallest diameter fit area, install appropriately sized shim stock the length of the fit area in the oversize bolt hole(s). For example, if the fit diameter of a single bolt hole was 10 mils (0.254 mm) larger than the remaining holes in the coupling, a piece of 5-mil (0.127-mm) stainless shim stock would be rolled to the fit diameter and cut to the length of the hole fit and installed in the hole. 7-4
Rotor Alignment and Balancing
4. Remove any coupling spacers before balancing except those that are bolted and doweled. 5. Rotate the rotor at a slow speed (typically around 200–300 rpm) to remove any temporary static bow. 6. Record the total indicator run-out (TIR) of the rotor at several planes including the mid-span of the rotor. Typically, run-out is recorded between each turbine stage, at the oil seal areas, the journals, and the couplings. It is not unusual to see rotor bow of less than 0.003" (0.0762 mm). 7. Use the “factory” grooves to balance the rotor during the slow-speed balancing process. Typically, the mid-span balance location is not used for LP rotor during slow-speed balancing. 8. After balancing to the specified criteria, compare the required weights and locations with the removed weights. 9. Install the appropriate weights. Use the mid-span location for HP and IP rotors, especially if they are bowed. (A bowed rotor is one with greater than 0.003" (0.0762 mm) TIR.) The slow-speed balance weight distribution for a bowed rotor is determined by the location (end planes or mid-span) of the bow in the rotor.
7.3
Automated and Semi-Automated Alignment Processes
The steam turbine alignment process consists of internal alignment (stationary components) and shaft alignment (couplings). The internal alignment consists of the alignment performed on casings, stationary blades and diaphragms, oil pump, and governor stand. Shaft alignment consists of the alignment of each turbine shaft (rotor or spindle) to its adjacent shaft. Computer programs can assist in the internal alignment process. These programs can determine the optimum placement of each component and can perform these calculations very quickly after the data have been entered into the program. The turbine manufacturers use computer programs to perform internal alignment on their equipment. This alignment software is available for purchase by turbine owners if they choose to perform the alignment themselves. The collection of alignment data has also been automated, and there are electronic devices that record and store data using computers. These data can be downloaded into automated alignment programs to speed up the alignment process. In addition, computer programs such as TGAlign can assist with the shaft alignment process and save considerable outage time when used on large multi-rotor machines as previously described. The program calculates the optimum bearing moves to achieve an expected alignment within user-specified alignment and move limits. Output reports specify radial positions at oil or gland bores along with shim changes for each bearing that is moved. Changes can be calculated very quickly for different alignment requirements. Previous methods for calculating bearing moves were performed by hand and could take hours. After the data are entered into TGAlign, the program can calculate alignment moves in seconds with no math errors. Lasers can be used to 7-5
Rotor Alignment and Balancing
assist with the collection of coupling alignment data. Before the use of lasers, these data were taken with micrometers, sliding parallels, and a dial indicator. During the shaft alignment process, the shafts must be turned to collect data. This process is complicated due to the weight of the turbine shafts and the need for the shafts to be stopped at a precise location with no binding in the coupling. A device called the Hutter Pin has been developed that assists with this process, and an ac inverter can be used to power the turbine turning gear. (See Volume 3, Section 3.2 for more details on the Hutter Pin). The Hutter Pin allows for concurrent rotation of both shafts without binding in the coupling. The ac inverter allows the turning gear to be operated as a variable speed motor. By operating the turning gear motor using the ac inverter, the shaft speed can be slowed to obtain better control, which aids in the data collection process. A shaft alignment method called strain gauge alignment allows the coupling alignment to be determined without separating the coupling halves. This method also is the only way to check shaft alignment while the unit is still hot. Strain gauge alignment is performed by measuring the strain in the turbine shaft near the coupling. The strain on the periphery of the shaft is directly related to the concentricity of the two turbine rotors. Strain gauge alignment requires modeling of the turbine rotors, the installation of strain gages on the shaft on each side of the coupling, electronic strain gauge data collecting, and software that analyses the data. This method for shaft alignment is very accurate and results in very low shaft vibration amplitudes. The limitations of the strain gauge alignment method are: •
Modeling of the rotors takes considerable time and effort.
•
Strain gauges must be designed and fit to each rotor.
•
The cost of modeling each turbine is expensive.
Modifications must be made at the pedestals to enable the recording of data. These modifications are required to provide access to the strain gauges when the unit is on turning gear. The cost of the installation is often offset by the benefits of getting more accurate coupling alignment and getting alignment data without separating the turbine couplings. This method allows the owner to check turbine shaft alignment if the unit is off-line for only 8–12 hours.
7.4
Slow-Speed Versus High-Speed Balancing
Slow-speed balancing should be performed on turbine rotors when significant work has been performed to the rotor. A slow-speed balancing procedure is found in Volume 3 that can assist a plant to direct or monitor this process. Significant work would include replacement of buckets/blades, bucket cover replacements, and tie wire replacements. Major work to the steam path does not include any local hand dressing of the buckets or covers. Balancing would also be required if any non-cylindrical machining were done to the rotor. The objective of slow-speed balancing is to get the rotor through the first rotor critical speed during startup after a major outage. Slow-speed balancing is easily completed at the utility 7-6
Rotor Alignment and Balancing
facility using portable equipment, but high-speed balancing is not often practical for turbine rotors. The time to ship the rotor to a high-speed balance location is often greater than the time required to high-speed balance the rotor during turbine startup. The probability of at-speed onsite balancing still exists even if the rotor was originally high-speed balanced off-site. The effect of the off-site high-speed balance may easily be offset by factors that did not exist where the rotor is balanced, such as assembly, bearing stiffness, loading, dampening factors, and cross effect. High-speed balance is practical for small, single-rotor turbines where field balancing is difficult and for rotors that cannot be field balanced to acceptable levels. The positive impact of a turbine rotor slow-speed balance is not realized in generator fields. Unlike turbine rotors, generator fields do not have a mid-span balance location. Slow-speed balance corrects for the first critical mode (graphically exaggerated in Figure 7-2) by placing weight in the appropriate opposite location to offset the “bow” or “loop.” The most effective spot for the weights is at the location of the bow, which is typically the mid-span location or in close proximity to it. The grooves for weight insertion into a generator field are located out toward the ends, outboard of the retaining rings and inboard of the bearings, generally in the fan ring or similar location. This location for balancing a rotor is commonly called the end plane. It can be seen that the mode shape of the second critical speed (in Figure 7-2) could be excited by the incorrect placement of weights that are attempting to compensate for the first critical mode. The weight placement to compensate for the first critical speed would place one weight opposite the second critical unbalance and opposite the first critical unbalance and may seem to be visually effective. But the second weight on the opposite end could end up being placed on top of the second critical unbalance location, exciting the rotor to greater unbalance and increased vibration. Most 3600-rpm generator fields operate above their second critical speed and below their third critical speed. They experience going through the third critical speed during overspeed conditions. Most 1800-rpm machines operate between the first and second critical speeds will run above second critical speed during an overspeed condition. The third critical speed may also be excited by weights placed in an end plane that attempt to correct for the first critical unbalance. Therefore, slow-speed balancing is not normally done to generator fields because of the reduced possibility of positive impact and the possibility of exciting the higher vibration modes.
7-7
Rotor Alignment and Balancing
Figure 7-2 Exaggerated Rotor Motion for the First Three Field Critical Speeds
7.4.1 Slow-Speed Balance Requirements/Considerations The manufacturer of the portable balance machine should provide certification of the machine’s balance capability for rotor speed and maximum weight. The machine should be sized to carry the largest rotor in both diameter and length. The vendor that supplies the portable machine should provide machine “footprint” and setup requirements as shown in Figure 7-3. The balancing machine must be lightweight and easy to install without special foundation requirements. Setup should be quick and easy without the need for extended calibration. The machine should be capable of sensitive low-speed balancing, which enhances safety and reduces power requirements. Rotor pedestals must be easy to move for various rotor configurations. The rollers should be self-aligning, which reduces setup time.
7-8
Rotor Alignment and Balancing
Figure 7-3 Low-Speed Portable Balance Machine
7.5
When Spin Balancing Is Required
Spin balancing of turbine and generator rotors is performed when the turbine is disassembled and the rotors have been removed from the machine. Spin balancing is necessary if any of the following are true: •
The turbine rotor has had bucket or blade work.
•
A rotor is discovered to be bowed.
•
The generator field has been rewound or has had the retaining rings removed.
•
The rotor balance has changed without any definite cause.
It is important to note that generator fields may require a compromise balance between the mechanical and electrical effects on the windings. All fields by nature have some degree of thermal sensitivity, but a field may bow excessively as loading increases (increase in field current); if the unbalance exceeds acceptable vibration limits, the field is identified as being thermally sensitive. The cause of the thermally induced bowing may be uneven temperature distribution within the field or axial forces associated with differential thermal growth between the copper windings and steel body not being evenly distributed. 7-9
Rotor Alignment and Balancing
Thermal sensitivity may be reversible or irreversible in operation. Reversible thermal sensitivity follows field current (load) both decreasing and increasing. A compromised balance may be required to offset the thermal vector and maintain the vibration within acceptable limits. Corrections for reversible thermal sensitivity may be done either off-site or in the stator if the causes and required balance information are known. Thermally sensitive rotors may be the result of a field rewind. If strict care is not taken during the field rewind, the windings may not be uniformly wound, insulation thickness, binding, uneven friction forces in the slots or under the retaining rings may all be causes of thermal sensitivity. Re-wedging to clean and repair wedges with improper reassembly or after a generator field rewind may also result in a thermally sensitive rotor if wedge tightness is not uniform. Irreversible thermal sensitivity follows an increasing field current but does not reduce with decreasing field current, or it may partially reduce and then “lock in.” This condition typically will limit loading and unit operation. Often, the unit must be taken off-line to turning gear operation to unlock the restrained forces. This condition typically cannot be compensated for with a balance program but requires field disassembly, rewinding, and then balancing. Most spin balancing is performed at low speeds and can be performed at the power plant using a portable balance machine. In some cases, the portable lathe that is used to machine the rotor can also perform the spin balance of the rotor. Low-speed spin balancing is done at speeds below 500 rpm, and the speed varies depending on the size of the rotor and the length of the blades. High-speed balancing of turbine and generator rotors is performed off-site in special balance chambers (or “pits”) or on-site during startup. Balance chambers are designed to spin the rotor at speeds up to the design overspeed of the rotor. For turbines, these speeds are up to 112% of rated speed and, for generator fields, up to 120% of rated speed. It is possible for some high-speed balance chambers to run the balance test at design conditions with heating of the generator field to determine if the field has any shorted windings. All high-speed balancing pits have the ability to pull vacuum to be able to spin rotors up to rated speed without overheating the blades. High-speed off-site balancing may not always be practical for these limiting factors: •
Travel time
•
Rotor size
•
Spin pit capacity
•
Spin pit schedule
•
Balancing duration
•
Rotor availability during the outage
The adverse effect of these factors may force high-speed balancing, which can be done only during the startup period of an outage.
7-10
Rotor Alignment and Balancing
High-speed balancing may be necessary any time a generator field is rewound. Balancing of the generator field in the stator is very difficult and can take a considerable amount of time and effort. Off-site high-speed balancing of turbine rotors is suggested if the unit is a large baseloaded unit, and the cost of lost generation due to downtime is high. Performing a high-speed balance is expensive, and it is difficult to transport rotors to the balance chambers. If the cost of lost generation is greater than the cost of the high-speed balance, a high-speed balance is recommended. Balance programs in the power plant after a maintenance inspection can take a considerable amount of time, and they are always on the critical path of the outage. A high-speed balance specification that can be used for a turbine, generator, or exciter rotor is provided in Volume 3. The specification includes guidance and acceptability requirements for unbalanced vibration of rotors balanced in a high-speed spin pit. Off-site high-speed balancing of small turbines or turbines that have low capacity factors is generally not recommended. It is usually more cost effective to perform a low-speed balance on the rotor and trim balance the unit when it is started. There may be special cases that require a high-speed balance of a small turbine, but normally, the cost prohibits this work. If many rows of blades are replaced or if the rotor is bowed and requires a lot of machining work, the investment for a high-speed balance may be justified. A high-speed balance should follow all generator field rewinds.
7.6
On-Line Balancing Devices
On-line balancing devices are widely used on smaller non-turbine applications, such as reactor coolant pumps, industrial fans, grinding spindles, machining centers, etc. This technology is being adapted to turbine-generators. For example, technology is currently available for an active balancing system for turbine application that corrects for imbalance while in operation. The obvious advantage of an active on-line balancing system is the time saved during startup and while on-line to correct unbalance conditions. An active on-line balancing system would also allow correction above and below rotor critical speeds, optimizing the critical speed transient vibration. Table 7-1 presents selected specifications of a high-speed turbomachinery active balancing system.
7-11
Rotor Alignment and Balancing Table 7-1 Specifications for an On-Line Active Balancing System Parameters
Range
RPM speed
500–15,000
Balancer capacity
Up to 200 oz-in. (14.40 kg/cm)
Temperature range
-67° to 302°F (- 55 to 150°C)
Humidity
10% to 90% non-condensing
Balance functions
Automatic multi-plane balancing Operator-initiated auto sequence Operator-confirmed individual step On-line system identification Automatic single-plane balancing Manual balance weight positioning
7.7
Potential Consequences of Not Balancing the Rotor
The objective for slow-speed balancing an individual rotor during an overhaul is to increase the probability that the rotor will pass through its critical speeds during startup and achieve running speed without tripping due to high vibration. The time necessary for high-speed balancing of the rotor after the overhaul has been completed should be reduced. When work has been done on a rotor during an overhaul, it is estimated that 50% of them do not make operating speed if no slow-speed balancing has been performed. It has also been estimated that rotors that have been slow-speed balanced during an outage have a 95% probability of reaching operating speed without a balance shot and that more than 80% of rotors do not need high-speed trim balancing during startup after the outage has been completed. In the machine, balance shots are normally done at rated speed conditions; therefore, the data to calculate a shot is at speed. It is much more difficult to calculate a balance shot before getting to rated speed. The reference information used to calculate a balance shot is: 1. Sensitivity in oz./mil (g/mm) is used to calculate the amount of weight to offset the imbalance. 2. The “high spot” number is used for the angular placement of the weights. The amount of weight required and the calculated angular component is obtained from plotting and resolving the imbalance vectors on polar graph paper. The imbalance vectors are the amplitude and angle reference of the rotor vibration readings. This information is not normally available at all the speed variations from turning gear to at speed and may be very difficult to obtain if the unit is vibrating severely at an off speed condition. The rotor must be held at that speed, and the vibration readings must be taken to calculate a balance shot with best guess sensitivity and high spot number. Slow-speed balancing helps the probability of obtaining rated speed after an outage. 7-12
Rotor Alignment and Balancing
7.8
Selecting Vibration Limits
There are many causes for turbine vibrations; many of which are discussed at greater length in the balance primer contained in Volume 3. Vibration limits are necessary to determine the proper time for turbine balancing. Vibration limits differ depending on whether the unit is starting up after a long maintenance inspection, or if the vibration amplitudes are increasing during operation. The turbine manufacturers give limits for steady-state conditions, critical speeds, maximum levels, and levels for well-balanced units. The following vibration limits are for mechanical unbalance and are measured peak to peak: •
For steady state conditions at high loads on 3600-rpm turbines, the turbine manufacturers recommend that the shaft vibration levels remain at 4.0 mils (0.10 mm) or below.
•
At critical speeds, the vibration levels should remain at 8 mils (0.20 mm) or below.
•
The maximum vibration level for a turbine shaft is 6 mils (0.15 mm).
•
A well-balanced unit should have vibration levels of 2 mils (0.05 mm) or less.
The vibration trip limits depend upon speed, reason for vibration, and length of time at the vibration level. The following trip limits are for 3600-rpm turbines: •
The turbine should be tripped if the speed is 800 rpm or less and the vibration level reaches 5 mils (0.13 mm).
•
For speeds between 800–2000 rpm, the turbine should be tripped if the vibration level reaches 10 mils (0.25 mm) or if the vibration level reaches 7 mils (0.18 mm) for two minutes.
•
For speeds between 2000–3600 rpm, the turbine should be tripped if the vibration level reaches 10 mils (0.25 mm) or if the vibration level reaches 7 mils (0.18 mm) for 15 minutes.
•
For 1800-rpm turbines, the trip limits are 2 mils (0.05 mm) higher than for 3600-rpm turbines due to their larger mass.
Steady-state vibration limits for 3600-rpm turbines vary by manufacturer. The vibration levels used most often for high loads for large steam turbines are as follows: •
Satisfactory operation – 3 mils (0.08 mm) or less
•
Alarm – 5 mils (0.13 mm)
•
Trip – 10 mils (0.25 mm)
When starting a turbine after a long maintenance inspection, it is normal to perform a balance program. Continuous recording equipment is installed on the unit that records vibration amplitudes and phase angles for each bearing and possibly for some of the couplings. Vibration levels are measured at varying loads, and the cause of any unbalance can be determined using the recorded data. The change in vibration amplitude and phase angle is used to determine what corrective actions are necessary. Some problems that cause high vibration levels during start up are: •
Rubbing
•
Misalignment 7-13
Rotor Alignment and Balancing
•
Water induction
•
Steam temperature variation
•
Faulty steam seal operation
•
Out-of-round bearing journals
7.9
Balance Limits
Out-of-machine slow-speed balance tolerances are a function of rotor weight, operating speed, rotor type, and applicable or chosen equipment standard. Rotors are classified as either rigid or flexible. A rotor is classified as rigid if it operates below any resonant frequency. A rule of thumb states that rotors that operate below 70% of their critical speed are considered rigid. Flexible rotors are those that operate above 70% of the first critical speed or, in general, operate above at least one resonant frequency. Therefore, slow-speed balancing is a rigid rotor mode activity. Table 7-2 provides a listing of some of the available standards that provide information regarding the balancing of equipment and the balancing of rotors.
7-14
Rotor Alignment and Balancing Table 7-2 Sources for Equipment and Rotor Balancing Standards International Organization for Standardization (ISO) Specification
Subject
Content
ISO 1925:1990
Mechanical vibration -- balancing -Vocabulary
Contains definitions of most balancing and balancing equipment terms
Mechanical vibration -- balancing quality requirements of rigid rotors
Classifies rotating work pieces and recommends balance tolerances
ISO 19401:1986
Part 1: determination of permissible residual unbalance ISO 19402:1997
Mechanical vibration – balance quality requirements of rigid rotors Part 2: balance errors
ISO 2953:1999
Mechanical vibration – balancing machines -- description and evaluation
ISO 2954: 1975
Mechanical vibration of rotating and reciprocating machinery -- requirements for instruments for measuring vibration severity
ISO 10814:1996
Mechanical vibration -- susceptibility and sensitivity of machines to unbalance
ISO 11342:1998
Mechanical vibration -- methods and criteria for the mechanical balancing of flexible rotors
Describes for a prospective balancing machine user how to specify requirements to a balancing machine manufacturer, including proposal requirements, and identifies how to test a machine to ensure compliance to the specification
American National Standard Institute ANSI S2.191989
Mechanical vibration -- balance quality requirements of rigid rotors
U.S. counterpart of ISO 1940/1-1986
Part 1, determination of permissible residual unbalance American Petroleum Institute ANSI/API Std 610-1995 Std 611
Centrifugal pumps for petroleum, heavy duty chemical and gas industry services
Contains vibration and balance limits
General purpose steam turbines for petroleum, chemical, and gas industry services
Contains vibration and balance limits
Society of Automotive Engineers ARP1134
Adapter interface - turbine engine blade moment weighing scale
Military Standards Mil Std-167-1
Mechanical vibration of shipboard equipment
Includes basis of acceptability criteria, methods, and limits
7-15
Rotor Alignment and Balancing
Each standard provides a different method to calculate the required balancing tolerance or the amount of residual unbalance in a rotor. Figure 7-4 is a plot of the impact of various standards on the amount of remaining unbalance.
Figure 7-4 Various Standards for Residual Unbalance
The following are examples of tolerances applied for slow-speed balancing: •
High-pressure rotor
18 ounce-inches/plane (12.96 gram-meters/plane)
•
Intermediate-pressure/reheat rotor
36 ounce-inches/plane (25.92 gram-meters/plane)
•
Low-pressure rotor
70 ounce-inches/plane (50.4 gram-meters/plane)
These tolerances are a slight modification from Mil. Std. 167-1 that states: oz-in = 4 W/N Where:
W= rotor weight in pounds for the plane
Example, for a two-bearing symmetrical rotor, W = Half of the total weight of the rotor N = The operation speed of the rotor, for greater than 1,000 rpm A balancing tolerance is applied because it is not necessarily reasonable to take a balance level down to loss of phase signal to ensure smooth rotor operation. Modern sensitive state-of-the-art balance equipment with microprocessor controls is capable of achieving residual unbalance levels below what is normally required for smooth rotor operation. Therefore, slow-speed balance tolerance is a combination of choosing the right residual unbalance level and the right balancing equipment. 7-16
Rotor Alignment and Balancing
7.10 Access to Turbine-Generator Rotors Access ports through the HP and IP shells for mid-span and end planes and access to LP rotor end plane locations through the condenser are provided for field balancing. Mid-span balance locations may be located through a multi-use location such as rotor pre-warming access. End plane locations in HP or IP section will usually just penetrate the lowest possible pressure location of the section. End plane access through shells is usually radial and axially angular. The balance access and the rotor balance location may not always line up in both cold and hot positions. Therefore, it is advantageous to observe rotor/shell locations during an outage and modify the access if required. Figure 7-5 shows a sketch of an offset modification to a shell bore. The modification is axial and provides sufficient angularity to catch the rotor balance plane in both the hot and cold positions.
Figure 7-5 Offset Modification to a Shell Bore
The generator field balancing locations are the most difficult to access of all turbine-generator locations. Access to the HP and IP section is from outside the shells and through the shells. The difficulty in accessing the HP and IP balance locations may be reduced by removal of the balance port fasteners or alignment of the balance location on the rotor through the shells. The LP rotor balancing location is typically accessed through an opening in the LP hood outer shell. The safety hazards in accessing the LP section may be air quality or elevated humid temperature, especially if the unit has just been shut down. The hazard in entering a generator 7-17
Rotor Alignment and Balancing
that has just been shut down to access balance locations is the H2 and CO2 that may be present. The generator casing must be cleared of H2 and purged of CO2 prior to entry. But the major difficulty in accessing the generator balance locations is logistical, as seen in Figure 7-6.
Figure 7-6 Access to Balance Grooves
The balance groove located on the fan ring must be accessed by entering the space between the outer and inner end shields. At least one segmented fan nozzle ring is then removed to access the balance groove. The space between the outer and inner end shields is very small. The person inserting the balance weights must be “not claustrophobic,” thin, and flexible. These are legitimate criteria because it is quite possible to get caught on interior fastener lock tabs while sliding through the limited space. Given the space, it can become extremely difficult to release caught clothing. All tools must be tied off to a retrieval string in case they are dropped. Therefore, the generator shot program is planned very carefully to minimize the number of balance shots required.
7.11 Turbine-Generator Balance Support When turbine-generator balancing program is performed, all of the necessary tools and equipment must be on hand before the unit startup. Equipment needed for balancing a turbinegenerator set includes: •
Continuous data collection equipment
•
Power cables, connector cables and rope
•
Atmosphere tester for both LP and generator balancing
•
Balance weights for each rotor and coupling
•
Weight installation tools
7-18
Rotor Alignment and Balancing
•
Lights and strobe or phase reference
•
Balance shot calculation software or a manual means of determining an appropriate balance shot
•
Safety harness and appropriate extraction equipment for LP balancing (if access to the LP balance location is considered a “confined space” or requires a “confined space” permit)
The equipment used to collect the vibration data varies for each machine. This equipment can gather data continuously or by manual means. If the data are collected manually many pieces of equipment must be checked out before startup. Most systems that are used to gather vibration data on large steam turbines are automatic. After they are set up, they can collect data continuously until they are disconnected. These systems use non-contact proximity probes to transmit the vibration amplitudes and a key phase reference on the shaft to determine the phase angle of the imbalance. It is important to make sure the cables are connected to the proper proximity probe and that they are calibrated to read correctly. This is done during the turbine outage to ensure that they will work during startup of the turbine. Balance weights for each rotor and coupling should be made before the balance program and should include all locking hardware. The balance planes on the rotors should have the weights consolidated during the outage to ensure that there will be sufficient room available if a balance weight needs to be installed. Balance weight installation tools should be checked to make sure that it is possible to install a weight in any location along the shaft if necessary. When installing balance weights, the engineer should be familiar with each rotor and should know how to install the weights. Rope will be needed to tie off weights and tools so they do not be lost if dropped. Lights should also be tied off so they will not be lost if dropped. The turning gear is used to rotate the rotor when installing balance weights. The engineer should be familiar with the operation of the turning gear and the allowable time limits for removing a turbine from turning gear operation when the unit is hot. After the balance weight is installed, the turbine must be rolled on the turning gear for a sufficient amount of time to allow the rotor eccentricity to return to an acceptable level prior to starting the turbine. The turbine engineer should be familiar with these limits before beginning the balancing program. The calculation of a balance shot is required if the rotor is determined to have mechanical unbalance. This can be done manually or by balance software. Most manufacturers use computer programs to plot balance shots. This software is programmed using high spot numbers, sensitivity to weight, pickup angles, equipment phase angle, and influence coefficients for other rotors that are coupled along the turbine train. All of these influences are necessary to plot a balance weight. If the balance weight is plotted manually, the balance engineer should be familiar with each of the influences and should know how to use them to plot a shot.
7-19
Rotor Alignment and Balancing
7.12 Turbine-Generator Balance Weights 7.12.1 Split-Weight Design Dovetail Weights When attempting to install dovetail-style weights during field balancing situations, you may experience the problem where access to the specific location on a rotor to install a dovetail weight is not available. Removal of previously installed balance weights to gain access to an access slot would be necessary. Figure 7-7 shows a typical split-weight dovetail design that can be used for fast access to the dovetail weight groove at any location, thus not requiring the removal of previously installed weights. Installing this type of weight can save significant time and effort in any individual balancing situation. The weight can be replaced with a standard onepiece weight when the rotor is removed during the next outage where the weights are consolidated during the slow-speed balance. Typical material for balance weights is 12-chrome stainless steel (AISI 403). This split-weight design balance weight uses a long set screw to go through a clearance hole in the top half of the weight assembly and screw into the threaded bottom half of the weight assembly. (See Figure 7-7 for details). By installing a nut (to function as a lock nut) on the exposed portion of the long set screw above the upper half weight, the upper half is fastened relative to the bottom half, and at the same time, the set screw is tightened in the bottom half weight against the rotor, locking the two pieces in the balance groove. Securing this style weight in the balance groove is somewhat of a “trial and error’ process, requiring several iterations of tightening the outer lock nut and then the overall set screw/nut assembly in place. When calculating the required weight size, remember to include the set screw and nut as part of the entire assembly. As an added precaution when installing the balance weight, the nut may be tack welded to the set screw and balance weight. Be sure to install Part B first in the balance weight groove before part A. See Figure 7-7.
7-20
Rotor Alignment and Balancing
Figure 7-7 Split-Weight Dovetail Weight
7-21
Rotor Alignment and Balancing
7.12.2 Tungsten-Style Weights Utilities have used balance weights made of tungsten, a material much heavier than the 12chrome stainless steel balance weight material (AISI 403) normally used. However, tungsten is not as resistant to erosion as 12-chrome stainless material and is more likely to lose mass during operation due to exposure to the harsh operating environment inside the turbine; therefore, some precautions should be taken when using this type of weight material. For the LP sections of a unit where moisture is present, dovetail-style tungsten weights can be employed to get more weight in a smaller location; however, the weight should be coated with cadmium to prevent water erosion of the tungsten material. There are companies in the United States that are capable of performing this coating process, which obviously would require some lead time prior to needing to install the weights in the unit. Use of tungsten material for plug-type weights in the HP and IP sections (and any LP sections that may use plug weights) of the unit is usually accomplished by drilling out (counter-boring) the standard field balance plug weight from the bottom side of the plug and then inserting the tungsten material into this counter bore with a size-to-size fit. The plug is then circumferentially seal welded at the tungsten/12-chrome interface to prevent the tungsten from falling out of the standard plug either during installation or removal. This process captures the tungsten material in an erosion-free atmosphere, while also giving the effect of more weight in the same size standard plug.
7-22
8
PRE–STARTUP CHECKS
Currently, the procedure for realigning hydraulic controls is slow, and relies heavily on the availability of an experienced controls engineer and staff of mechanics to relay valve strokes and hydraulic pressures. While the resetting of controls is still somewhat of an art, it is beneficial to establish the practices and procedures used to guide the performance of resetting the turbine controls and prepare a checklist of activities that should precede the return of the unit into service. This section of the guidelines provides a compilation of practices, procedures, and experiences relating to unit startup after overhaul. A review of advances in neural networking techniques that have been used to set boiler controls is included as an emerging technology that could possibly be adopted to automate the procedure by which control line-up problems are solved.
8.1
Steps to Minimize Startup Time
Before the turbine startup, the operator should review the starting procedure and become familiar with each item so that the operation of the turbine will go smoothly. The units in the power station seldom are identical, and each turbine will probably have variations in the starting procedure. It is helpful to create a startup document that identifies and summarizes critical turbine-generator elements to be observed during each startup process. The document should be organized to reflect the flow of startup activities and include: •
The applicable unit
•
Topic of concern
•
Source reference (for additional information or future verification)
•
Process variables
•
Limits
•
Definitions
•
Actions
The information presented in Table 8-1 shows examples of startup topics and relevant details that could be included in a startup document. Note that the applicable unit, primary reference, and other information are not included in this example.
8-1
Pre–Startup Checks Table 8-1 Recommended Outline for a Startup Document Topic
Details
Hydrogen – Air Leak Test
(a) The amount of hydrogen consumed during operation or air during testing is either absorbed into the seal oil or lost through leakage at the seals or other locations. (b) The test should be run with the gas temperature as steady as possible. (c) The generator should be allowed to thermally stabilize for 1–2 hours after filling with gas.
Seal Oil Flow (Absorption)
Gas Loss (Leakage)
Acceptable
Unacceptable
Acceptable
Unacceptable
5–15 gpm (18.9–56.78 Lpm)
28.317 m per min). Measure seal oil flow and calculate leak rate. Typical leak rates 3 are 400–500 ft (11.327– 3 14.158 m ) per day 3
3
>2,000 ft (>56.634 m ) per day. Locate and seal loss areas. Design Gas Loss (Leakage) 3
3
300 ft (8.495 m ) per day at 32 psi (220.6 kPa) 3 3 400 ft (11.327 m ) per day at 47 psi (324.05 kPa) 3 3 500 ft (14.158 m ) per day at 62 psi (427.5 kPa) 3 3 600 ft (16.990 m ) per day at 77 psi (530.9 kPa) Turning Gear Operation
10:1 ratio of time off to time on up to eight hours of turning gear run time.
H2O Induction
50°F (10°C) temperature mismatch is considered H2O induction.
T/G Operation After H2O Induction Incident
For temporary rotor bow, 2–6 hours on gear to remove bow. Humped shell may need >24 hours.
8-2
Pre–Startup Checks Table 8-1 (cont.) Recommended Outline for a Startup Document Topic
Details Reduce set point to 75–80°F (24–27°C) 75–80°F (24–27°C) Raise set point to 90°F (32°C) 90°F (32°C) o 95 F° (35°C) (Off wobbulator at 3000 rpm) 100°F (38°C) 110–120°F (43–49°C); 5 mils (0.13 mm) 7 mils (0.18 mm) for 2 min, or >10 mils (0.25 mm) 7 mils (0.38 mm), for 15 min or >10 mils (0.25 mm) 12 mils (0.30 mm) max (at critical speeds)
Vibration Limits (Roll Up)
< 800 rpm 800–2000 rpm > 2000 rpm
Bearing Metal Temperature
Normal Temperature Ranges: °
°33
Tilt pad 180–220 F (82–104 C) Elliptical 170–190°F (77–88°C) Short elliptical 190–210°F (77–99°c) Alarm: Tilt pad at 225°F (107°C) Elliptical at 210°F (99°C) Maximum temperature: 250°F (121°C) Vacuum Breaking
It is recommended that vacuum not be broken until the unit has reached 2/3 (2400 rpm) of rated speed unless an emergency condition, such as high vibration, requires the unit to be slowed down as fast as possible.
8-3
Pre–Startup Checks Table 8-1 (cont.) Recommended Outline for a Startup Document Topic
Details
Wobbulator
Used to prevent the turbine from running at a constant speed (3000 rpm) when bucket critical speed might be experienced.
Oil Trip Test
