Electrical Motor Diagnostics [2 ed.] 9780986347733

Citation preview

Electrical Motor Diagnostics 2nd Edition

Penrose_FM.indd i

4/22/08 1:57:04 AM

©2008 SUCCESS by DESIGN Publishing All Rights Reserved. No part of this book may be reproduced in any form or by any electronic or mechanical means including information storage and retrieval systems – except in the case of brief quotations embodied in critical articles or reviews – without permission in writing from its publisher, SUCCESS by DESIGN Publishing. Published By: SUCCESS by DESIGN Publishing Old Saybrook, CT 06475 Email: [email protected] http://www.motordoc.com This publication is designed to provide accurate and authoritative information in regard to the subject matter covered. It is sold with the understanding that the use of the information contained within this book does not imply or infer warranty or guaranties in any form. ISBN: 0-9712450-7-X ISBN-13: 978-0-9712450-7-5

Penrose_FM.indd ii

4/22/08 1:57:04 AM

Table of Contents

Table of Figures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiii Table of Tables. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xix Table of Equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xxiii Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .xxvii Chapter 1: The Motor Diagnostics and Motor Health Study . . . . . 1 Electric Motor Repair Study, 1995 . . . . . . . . . . . . . . . . . . . . . 2 In Service Motor Testing, 1999 . . . . . . . . . . . . . . . . . . . . . . . 4 Success by Design Best Practice Survey . . . . . . . . . . . . . . . . . 6 Dealing with the State of the Motor Testing Industry . . . . . . . 7 Chapter 2: Basic Electricity and Electro-Magnetism . . . . . . . . . . . 9 Current,Voltage and Resistance . . . . . . . . . . . . . . . . . . . . . . 10 Electro-Magnetic Fields . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 Alternating Current Electricity (Hz). . . . . . . . . . . . . . . . . . . 12 Chapter 2 Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 Chapter 3: Electric Motor Theory . . . . . . . . . . . . . . . . . . . . . . 17 AC Induction Motor Theory . . . . . . . . . . . . . . . . . . . . . . . . 17 Wound Rotor Motors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 iii

Penrose_FM.indd iii

4/22/08 1:57:04 AM

iv

Electrical Motor Diagnostics: 2nd Edition

Synchronous Motors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 Direct Current Electric Motors . . . . . . . . . . . . . . . . . . . . . . 25 Machine Tool Motors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 Traction Motors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 Transformers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 Chapter 3 Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35 Chapter 4: Electrical Insulation Systems and Theory . . . . . . . . . 37 The Motor Circuit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37 Insulation and Magnetic Field Effects . . . . . . . . . . . . . . . . . . 39 Winding Faults . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45 Traditional Test Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . 45 Insulation to Ground Testing (Meg-Ohm Meters) . . . . . . . 46 Polarization Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46 Resistance Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46 Surge Comparison Testing . . . . . . . . . . . . . . . . . . . . . . . . 46 Modern Low-Voltage Testing: Motor Circuit Analysis . . . . . . 47 Detection of Winding Contamination . . . . . . . . . . . . . . . 48 Overheated Windings . . . . . . . . . . . . . . . . . . . . . . . . . . . 48 Winding Shorts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49 Additional Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49 Rotor Testing and Back Iron Effects on MCA . . . . . . . . . . 50 Armature and Commutator Contamination Detection . . . 51 Summary of MCA Theory . . . . . . . . . . . . . . . . . . . . . . . . . 52 Development of Experiments . . . . . . . . . . . . . . . . . . . . . . . 52 Test Tolerances . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54

Penrose_FM.indd iv

4/22/08 1:57:04 AM

Table of Contents

v

MCA Tolerances . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54 Surge Testing Tolerances . . . . . . . . . . . . . . . . . . . . . . . . . . 55 Experiment Test Results . . . . . . . . . . . . . . . . . . . . . . . . . . . 55 Experiment 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55 Experiment 2a . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58 Experiment 2b . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59 Experiment 2c . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60 Experiment 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62 Experiment Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . 64 Experiment 1 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . 65 Experiment 2a Conclusions . . . . . . . . . . . . . . . . . . . . . . . 65 Experiment 2b Conclusions . . . . . . . . . . . . . . . . . . . . . . . 65 Experiment 2c Conclusions . . . . . . . . . . . . . . . . . . . . . . . 65 Experiment 3 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . 66 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66 Chapter 4 Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67 Chapter 5: Common Electric Motor Testing . . . . . . . . . . . . . . . 69 Electrical Motor Diagnostics Defined . . . . . . . . . . . . . . . . . . 69 Condition-Based Technologies . . . . . . . . . . . . . . . . . . . . . . . 70 Voltmeter Troubleshooting of an AC Motor . . . . . . . . . . . . . 73 The Power of a Volt Meter . . . . . . . . . . . . . . . . . . . . . . . . 74 Checking Contacts with a Volt Meter . . . . . . . . . . . . . . . . 76 Checking Fuses with a Volt Meter . . . . . . . . . . . . . . . . . . 77 Ammeter Testing of an AC Motor . . . . . . . . . . . . . . . . . . . . 77 Analog Ammeters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77

Penrose_FM.indd v

4/22/08 1:57:05 AM

vi

Electrical Motor Diagnostics: 2nd Edition

Digital Ammeters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78 Load Measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78 Current Unbalance Conditions . . . . . . . . . . . . . . . . . . . . 79 Other Current Conditions . . . . . . . . . . . . . . . . . . . . . . . . 80 DC Resistance Testing of AC Machines . . . . . . . . . . . . . . . . 81 Selecting the Right Measurement Tool . . . . . . . . . . . . . . . 81 Considerations When Using an Ohm Meter . . . . . . . . . . . 85 Troubleshooting Motors with Ohm Meters . . . . . . . . . . . 87 Classical Insulation Resistance Testing. . . . . . . . . . . . . . . . . . 87 The Basic Insulation Resistance Test . . . . . . . . . . . . . . . . . 88 Utilizing Hi-Pot Testing for Insulation to Ground Stress Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . 94 Evaluating Motor Condition with Low Voltage Advanced Diagnostics . . . . . . . . . . . . . . . . . . . . . . . 95 Motor Circuit Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . 95 Surge Comparison Testing . . . . . . . . . . . . . . . . . . . . . . . 102 Comparison of the ‘Big Three’ . . . . . . . . . . . . . . . . . . . . 104 Introduction to Electrical Signature Analysis . . . . . . . . . . . . 105 Multi-Technology Approach to CBM . . . . . . . . . . . . . . . . 111 Major Components and Failure Modes . . . . . . . . . . . . . . 113 Common Approaches to Multi-Technology . . . . . . . . . . 117 Application Opportunities . . . . . . . . . . . . . . . . . . . . . . . 118 Chapter 5 Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121 Chapter 6: Time to Failure Estimation . . . . . . . . . . . . . . . . . . . 123 The Tools for TTFE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123

Penrose_FM.indd vi

4/22/08 1:57:05 AM

Table of Contents

vii

The Severity Modifier . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125 Understanding Series and Parallel Availability . . . . . . . . . . . 125 Time To Failure Estimation with MCA . . . . . . . . . . . . . . . 129 Electric Motor Winding Failure . . . . . . . . . . . . . . . . . . . 129 Insulation Breakdown . . . . . . . . . . . . . . . . . . . . . . . . . . 130 Stages of Winding Failure . . . . . . . . . . . . . . . . . . . . . . . 131 Introduction to Motor Circuit Analysis (MCA) . . . . . . . . 134 Trending Fault Descriptions . . . . . . . . . . . . . . . . . . . . . . 134 AC Rotating Machine Testing . . . . . . . . . . . . . . . . . . . . 136 Time to Failure Estimation . . . . . . . . . . . . . . . . . . . . . . 136 Nuisance Tripping Situations . . . . . . . . . . . . . . . . . . . . . 137 Winding Contamination . . . . . . . . . . . . . . . . . . . . . . . . 138 Winding Shorts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139 Test Cases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141 Chapter 6 Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143 Chapter 7: Motor Circuit Analysis Testing AC Machines. . . . . . 145 Fault Detected Procedure . . . . . . . . . . . . . . . . . . . . . . . . . 146 Data Collection and Test Result Issues . . . . . . . . . . . . . . . . 146 Rotor Test Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149 Synchronous Machine Testing with MCA. . . . . . . . . . . . . . 150 Most Common Synchronous Motor Faults . . . . . . . . . . . 152 Basic Steps for the Analysis of Synchronous Machines with MCA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152 Generator Analysis with MCA . . . . . . . . . . . . . . . . . . . . . . 153 Evaluating Generators with MCA . . . . . . . . . . . . . . . . . 155

Penrose_FM.indd vii

4/22/08 1:57:05 AM

viii

Electrical Motor Diagnostics: 2nd Edition

MCA Instrumentation: The ALL-TEST PRO 31 . . . . . . . . 156 Analysis of Pin-Hole Shorts (VFD-Related Motor Failure) . . . . . . . . . . . . . . . . . . . . 157 AC Machine Case Studies . . . . . . . . . . . . . . . . . . . . . . . . . 159 Case 1: New Motor Phase Unbalance . . . . . . . . . . . . . . . 159 Case 2: Casting Void 200 HP Motor . . . . . . . . . . . . . . . . 163 Case 3: ID Fan Motor Rebuild Defect . . . . . . . . . . . . . . 164 Case 4: Cable Faults . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165 Case 5: Evaluation of an 8000 HP Synchronous Motor . . 167 Case 6: AC Traction Motor – Good . . . . . . . . . . . . . . . . 168 Case 7: Spindle Motor – Good . . . . . . . . . . . . . . . . . . . . 169 Case 8: Brushless DC Servo – Coil-to-Coil Short . . . . . . 169 Case 9: 100 Horsepower, 1800 RPM with Dirty Windings . . . . . . . . . . . . . . . . . . . . . . . . . . . 170 Case 10: Servo Motor – Turn-to-Turn Fault . . . . . . . . . . 171 Case 11: Rotor Position Related Unbalance . . . . . . . . . . 171 Case 12: Motor Tested from MCC – Shorted Windings . . 172 Case 13: Brand New 300 Horsepower Motor with Shorted Phase . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173 Case 14: 3000 Horsepower with Fractured Rotor Bar . . . 174 Case 15: New 50 Horsepower, 2-pole Motor Combined Faults . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175 Case 16: 300 HP, 2300 Volt, 3600 RPM Broken Rotor Bars . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176 Case 17: 1000 HP, 3600 RPM – Special Signature on Good Motor . . . . . . . . . . . . . . . . . . . . . . . 177

Penrose_FM.indd viii

4/22/08 1:57:05 AM

Table of Contents

ix

Case 18: 700 HP, 3600 RPM, Broken Rotor Bars and Cracked Shaft . . . . . . . . . . . . . . . . . . . . . . . . . 178 Chapter 7 Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179 Chapter 8: Motor Circuit Analysis Testing DC Motors . . . . . . . 181 The Commutator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181 Brush and Commutator Operation. . . . . . . . . . . . . . . . . . . 182 Seating, Tensioning and Cleaning . . . . . . . . . . . . . . . . . . . . 184 Basics of DC Machines for MCA Testing . . . . . . . . . . . . . . 186 Common DC Motor Electrical Faults . . . . . . . . . . . . . . . . 187 Armature Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 188 Series Motor Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189 Shunt Motor Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 190 Compound Motor Testing . . . . . . . . . . . . . . . . . . . . . . . . . 190 General DC MCA Testing Notes . . . . . . . . . . . . . . . . . . . . 191 Chapter 8 Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193 Chapter 9: Electrical Signature Analysis Theory . . . . . . . . . . . . 195 Alternating Current Induction Motors . . . . . . . . . . . . . . . . 197 Induction Motor Troubleshooting Procedure . . . . . . . . . 197 Induction Motor Troubleshooting Analysis Overview . . . 198 Power Quality and Electrical Analysis . . . . . . . . . . . . . . . 199 Rotor Analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 201 AC Motor and Stator Analysis . . . . . . . . . . . . . . . . . . . . 204 Driven Equipment Evaluation . . . . . . . . . . . . . . . . . . . . . . 206 Belts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 206 Gear Mesh . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207

Penrose_FM.indd ix

4/22/08 1:57:05 AM

x

Electrical Motor Diagnostics: 2nd Edition

Driven Equipment – Blade Pass Frequencies . . . . . . . . . . 207 Generator Analysis with ESA . . . . . . . . . . . . . . . . . . . . . . . 208 DC Electric Motors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 208 DC Motor Testing Procedure . . . . . . . . . . . . . . . . . . . . . 209 Basic DC Motor Analysis . . . . . . . . . . . . . . . . . . . . . . . . 210 Chapter 9 Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 212 Chapter 10: Electrical Signature Analysis Pattern Recognition AC Machines . . . . . . . . . . . . . . . . . . . . . . . . . . . 213 The Elusive Stator Slots and Rotor Bars . . . . . . . . . . . . . . . 213 Stator Slots . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 214 Rotor Bars . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 217 Use of Databases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 217 Alarm Levels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 217 Evaluating Data: Basic Steps for Pattern Recognition. . . . . . 218 Common AC Induction Motor Signatures . . . . . . . . . . . . . 219 Unbalance and Misalignment . . . . . . . . . . . . . . . . . . . . . 219 Gearbox Fault . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219 Classical Rotor Bar Fault . . . . . . . . . . . . . . . . . . . . . . . . 220 Punch Press: Driven Equipment Effects. . . . . . . . . . . . . . 221 Fan and Belt Faults . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221 Coil Movement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 222 Static Eccentricity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 223 Variable Frequency Drives . . . . . . . . . . . . . . . . . . . . . . . . . 223 Chapter 10 Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . 226 Chapter 11: Electrical Signature Analysis Pattern Recognition DC Machines . . . . . . . . . . . . . . . . . . . . . . . . . . 227

Penrose_FM.indd x

4/22/08 1:57:05 AM

Table of Contents

xi

DC Troubleshooting Patterns . . . . . . . . . . . . . . . . . . . . . . . 227 Considerations in DC Analysis . . . . . . . . . . . . . . . . . . . . . . 230 Chapter 11 Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . 234 Chapter 12: Developing an EMD Program . . . . . . . . . . . . . . . 235 Concepts of the Program . . . . . . . . . . . . . . . . . . . . . . . . . . 236 Key Performance Indicators . . . . . . . . . . . . . . . . . . . . . . . . 238 RCMM Team . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 239 Considerations in Selecting Equipment . . . . . . . . . . . . . . . 240 Defining Motor Management Programs . . . . . . . . . . . . . 240 Company Mission . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 242 The RCM-Based Approach for Motor Management . . . . 242 Selecting CBM Equipment . . . . . . . . . . . . . . . . . . . . . . 245 Appendix 1: Advanced Transformer Analysis with the ALL-TEST IV PRO 2000 . . . . . . . . . . . . . . . . . . . . . . . . . . . 247 Single Phase Transformer Testing . . . . . . . . . . . . . . . . . . . . 247 Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 247 Test Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 248 Three Phase Transformer Testing . . . . . . . . . . . . . . . . . . . . 249 Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 249 Testing Notes and Limits . . . . . . . . . . . . . . . . . . . . . . . . 250 Appendix 2: Evaluation of Capacitance in Motor Circuit Analysis Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251 Appendix 3: MCA Tables for TTFE Baselines . . . . . . . . . . . . . 259 Appendix 4: Determining Testing Frequency . . . . . . . . . . . . . . 261 Appendix 5: Stator Slot and Rotor Bar Tables . . . . . . . . . . . . . 263

Penrose_FM.indd xi

4/22/08 1:57:06 AM

Penrose_FM.indd xii

4/22/08 1:57:06 AM

Table of Figures

Figure 1: Cost Impact of Maintenance Programs $/HP/YR . . xxviii Figure 2: Modifications to Windings Through Rewind Process . . . 3 Figure 3: Testing Through the Repair Process . . . . . . . . . . . . . . . 4 Figure 4: Reasons for Motor Testing . . . . . . . . . . . . . . . . . . . . . . 5 Figure 5: Technologies Used . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 Figure 6: $/Hour Downtime . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 Figure 7: Shifts and Operation . . . . . . . . . . . . . . . . . . . . . . . . . . 7 Figure 8: Primary Driver for Motor Testing Program . . . . . . . . . . 7 Figure 9: Bohr’s Model of the Atom . . . . . . . . . . . . . . . . . . . . . . 9 Figure 10: Basic Motor Circuit . . . . . . . . . . . . . . . . . . . . . . . . . 17 Figure 11: Three Phase Stator and Windings . . . . . . . . . . . . . . . 19 Figure 12: Three Phase Rotor . . . . . . . . . . . . . . . . . . . . . . . . . . 20 Figure 13: Rotating Field and Rotor Cage . . . . . . . . . . . . . . . . 21 Figure 14: Series Motor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 Figure 15: Shunt Wound Motor . . . . . . . . . . . . . . . . . . . . . . . . 27 Figure 16: Compound Wound Motor . . . . . . . . . . . . . . . . . . . . 28 Figure 17: Power Transformer . . . . . . . . . . . . . . . . . . . . . . . . . . 31 xiii

Penrose_FM.indd xiii

4/22/08 1:57:06 AM

xiv

Electrical Motor Diagnostics: 2nd Edition

Figure 18: Delta-Delta Transformer . . . . . . . . . . . . . . . . . . . . . . 31 Figure 19: Delta-Wye Transformer . . . . . . . . . . . . . . . . . . . . . . 32 Figure 20: Wye-Delta Transformer . . . . . . . . . . . . . . . . . . . . . . . 32 Figure 21: Wye-Wye Transformer . . . . . . . . . . . . . . . . . . . . . . . 33 Figure 22: Delta-Wye Transformer Connection . . . . . . . . . . . . . 33 Figure 23: Delta-Delta Transformer Connection . . . . . . . . . . . . 34 Figure 24: Single-Phase Transformer Connection . . . . . . . . . . . . 34 Figure 25: Single-Phase Equivalent Circuit . . . . . . . . . . . . . . . . 38 Figure 26: Insulation Model of Motor Winding System . . . . . . . 39 Figure 27: Generation of a Dipole. . . . . . . . . . . . . . . . . . . . . . . 40 Figure 28: Dipolar Effect of Insulation. . . . . . . . . . . . . . . . . . . . 40 Figure 29: Magnetic Dipoles . . . . . . . . . . . . . . . . . . . . . . . . . . . 41 Figure 30: Balanced Wye System . . . . . . . . . . . . . . . . . . . . . . . . 44 Figure 31: Surge Results T1-T2 . . . . . . . . . . . . . . . . . . . . . . . . 55 Figure 32: Surge Results T1-T3 . . . . . . . . . . . . . . . . . . . . . . . . 56 Figure 33: Surge Results T2-T3 . . . . . . . . . . . . . . . . . . . . . . . . 56 Figure 34: Fault Detection (Night Vision Video). . . . . . . . . . . . . 57 Figure 35: Arc Damage (Night Vision Video) . . . . . . . . . . . . . . . 57 Figure 36: Effect of Surge Comparison Testing on Winding Contamination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61 Figure 37:Virtual Motor with Spacers to Hold Airgap . . . . . . . . 62 Figure 38: Surge Values on Each Test . . . . . . . . . . . . . . . . . . . . . 63 Figure 39: Arc Damage to Winding and Insulation . . . . . . . . . . . 64 Figure 40: Impact of Voltage Deviation . . . . . . . . . . . . . . . . . . . 75 Figure 41:Voltage Unbalance Multiplier . . . . . . . . . . . . . . . . . . 76

Penrose_FM.indd xiv

4/22/08 1:57:06 AM

Table of Figures

xv

Figure 42: Wheatstone bridge . . . . . . . . . . . . . . . . . . . . . . . . . . 82 Figure 43: Double Kelvin Bridge . . . . . . . . . . . . . . . . . . . . . . . 83 Figure 44: Basic 4-Wire Kelvin Bridge . . . . . . . . . . . . . . . . . . . 84 Figure 45: 4-Wire Kelvin Bridge . . . . . . . . . . . . . . . . . . . . . . . 85 Figure 46: Insulation Resistance Temperature Correction . . . . . . 90 Figure 47: Dielectric Absorption . . . . . . . . . . . . . . . . . . . . . . . . 91 Figure 48: Polarization Index . . . . . . . . . . . . . . . . . . . . . . . . . . 93 Figure 49: Good Impedance and Inductance Pattern . . . . . . . . . 96 Figure 50: Bad Impedance and Inductance Pattern . . . . . . . . . . . 97 Figure 51: PI Curve and Capacitive Discharges . . . . . . . . . . . . . 98 Figure 52: Broken Rotor Bar Signature . . . . . . . . . . . . . . . . . . 107 Figure 53: Coil Movement Signature . . . . . . . . . . . . . . . . . . . 109 Figure 54: Louis Allis 800 HP Stator . . . . . . . . . . . . . . . . . . . . 109 Figure 55: Copper from Louis Allis Coil Failure . . . . . . . . . . . . 110 Figure 56: Inherent Availability of Parallel Pump Components . 127 Figure 57: Failure Curve of Primary A for Bearing Detection Scenario . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128 Figure 58: Winding Contamination Rating to Hours . . . . . . . . 138 Figure 59: Rotor and Stator Relationship . . . . . . . . . . . . . . . . 155 Figure 60: ALL-TEST PRO 31 Kit . . . . . . . . . . . . . . . . . . . . . 157 Figure 61: Inductive Unbalance/Rotor Test . . . . . . . . . . . . . . . 160 Figure 62: Impedance Balance Test Using Alternate Impedance Tool at 1200 Hz . . . . . . . . . . . . . . . . . . . 161 Figure 63: Inductive Unbalance/Rotor Test, Alternate Motor . . 162 Figure 64: Impedance Unbalance, Alternate Motor . . . . . . . . . 162 Figure 65: Rotor Defect . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163

Penrose_FM.indd xv

4/22/08 1:57:06 AM

xvi

Electrical Motor Diagnostics: 2nd Edition

Figure 66: Good Rotor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164 Figure 67:Visual Representation of Faulted Brushless DC Servo . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169 Figure 68: 100 HP Visual Data . . . . . . . . . . . . . . . . . . . . . . . . 170 Figure 69: Rotor Position Visual Representation . . . . . . . . . . . 172 Figure 70:Visual Representation of Shorted Windings . . . . . . . 173 Figure 71: Broken Rotor Bar Data . . . . . . . . . . . . . . . . . . . . . 174 Figure 72: Combined Rotor and Stator Faults, Inductance . . . . 175 Figure 73: Combined Rotor and Stator Faults, Impedance . . . . 176 Figure 74: Broken Rotor Bars, 350 HP, 3600 RPM, 2300 Volt . 177 Figure 75: 1000 HP, 3600 RPM, Batman Signature . . . . . . . . . 177 Figure 76: 700 HP, 3600 RPM, Broken Rotor Bars and Cracked Shaft . . . . . . . . . . . . . . . . . . . . . . . . 178 Figure 77: Coefficient of Friction . . . . . . . . . . . . . . . . . . . . . . 183 Figure 78: Line Frequency . . . . . . . . . . . . . . . . . . . . . . . . . . . 196 Figure 79: Line Frequency with Harmonic Content . . . . . . . . 197 Figure 80: FFT Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197 Figure 81: Over and Under Voltage Impact on Motor Operation. . . . . . . . . . . . . . . . . . . . . . . . . . 200 Figure 82:Voltage Unbalance (Derating Factor) . . . . . . . . . . . . 200 Figure 83: Power Factor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 201 Figure 84: Example of Broken Rotor Bar Sidebands . . . . . . . . 202 Figure 85: Static Eccentricity . . . . . . . . . . . . . . . . . . . . . . . . . 203 Figure 86: Dynamic Eccentricity. . . . . . . . . . . . . . . . . . . . . . . 203 Figure 87: Stator Passing Frequencies . . . . . . . . . . . . . . . . . . . 205 Figure 88: Mechanical Imbalance . . . . . . . . . . . . . . . . . . . . . . 205 Figure 89: Belts and Sheaves . . . . . . . . . . . . . . . . . . . . . . . . . . 206

Penrose_FM.indd xvi

4/22/08 1:57:06 AM

Table of Figures

xvii

Figure 90: Gear Mesh. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207 Figure 91: Rotating Field Signature, High Frequency (Good) . . 209 Figure 92: Rotating Field Signature, Low Frequency (Poor Condition) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 209 Figure 93:Verified Running Speed . . . . . . . . . . . . . . . . . . . . . 215 Figure 94: Stator Slot Identifier . . . . . . . . . . . . . . . . . . . . . . . . 216 Figure 95: Unbalance or Misalignment . . . . . . . . . . . . . . . . . . 219 Figure 96: Gearbox Fault . . . . . . . . . . . . . . . . . . . . . . . . . . . . 220 Figure 97: Classical Rotor Bar Fault . . . . . . . . . . . . . . . . . . . . 220 Figure 98: Punch Press . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221 Figure 99: Fan and Belt Faults. . . . . . . . . . . . . . . . . . . . . . . . . 222 Figure 100: Coil Movement . . . . . . . . . . . . . . . . . . . . . . . . . . 222 Figure 101: Static Eccentricity . . . . . . . . . . . . . . . . . . . . . . . . 223 Figure 102: Low Frequency VFD Data . . . . . . . . . . . . . . . . . . 224 Figure 103:Voltage and Current at 0.05 Seconds . . . . . . . . . . . 224 Figure 104: Floating Noise Floor (Voltage) . . . . . . . . . . . . . . . 225 Figure 105: Lower Frequency Spectra of a DC Drive Fault. . . . 230 Figure 106: Bad Drive Current . . . . . . . . . . . . . . . . . . . . . . . . 231 Figure 107: Bad Drive Voltage . . . . . . . . . . . . . . . . . . . . . . . . . 231 Figure 108: Good Low Frequency Spectra . . . . . . . . . . . . . . . . 232 Figure 109: Good Drive Current . . . . . . . . . . . . . . . . . . . . . . 232 Figure 110: Good Drive Voltage . . . . . . . . . . . . . . . . . . . . . . . 233 Figure 111: Development of the Program (RCMM Map) . . . . 236 Figure 112: Dielectric Model . . . . . . . . . . . . . . . . . . . . . . . . . 252 Figure 113: Dissipation Factor (Fi) . . . . . . . . . . . . . . . . . . . . . 252 Figure 114: Fault Detection Frequency . . . . . . . . . . . . . . . . . . 261

Penrose_FM.indd xvii

4/22/08 1:57:07 AM

Penrose_FM.indd xviii

4/22/08 1:57:07 AM

Table of Tables

Table 1: Lost Potential in the USA Compared to Top Ten Economies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .xxvii Table 2: Average Impact of CBM Monitoring Programs . . . . . . xxix Table 3: Motor Synchronous and Operating Speed (60 Hz, NEMA Design B) . . . . . . . . . . . . . . . . . . . . . . . 22 Table 4: Assembled Motor Tolerances . . . . . . . . . . . . . . . . . . . . 54 Table 5: Disassembled Motor Tolerances . . . . . . . . . . . . . . . . . . 54 Table 6: Pre-Test Results with MCA . . . . . . . . . . . . . . . . . . . . . 55 Table 7: Before and After AT4 Test Results with Differences . . . . 58 Table 8: Before and After AT31 Test Results and Differences at 200 Hz . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58 Table 9: AT4 Test Results on Motor . . . . . . . . . . . . . . . . . . . . . 60 Table 10: AT31 Test Results on Motor (60Hz) . . . . . . . . . . . . . . 60 Table 11: AT4 Test Results on Stator . . . . . . . . . . . . . . . . . . . . . 60 Table 12: AT31 Test Results on Stator (60Hz) . . . . . . . . . . . . . . 61 Table 13: AT4 Test Results on Stator Post Test . . . . . . . . . . . . . . 61 Table 14: AT31 Test Results on Stator Post Test . . . . . . . . . . . . . 61 Table 15: AT4 Test Results on Virtual Motor . . . . . . . . . . . . . . . 62 xix

Penrose_FM.indd Sec18:xix

4/22/08 1:57:07 AM

xx

Electrical Motor Diagnostics: 2nd Edition

Table 16: AT4 Test Results After Test 6 and Before Test 7 on Virtual Motor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63 Table 17: AT4 Test Results on Virtual Motor After High Voltage Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63 Table 18: Phase to Phase Voltage Test Results . . . . . . . . . . . . . . . 74 Table 19: kVA Code AC Induction Motors . . . . . . . . . . . . . . . . 80 Table 20: Insulation Resistance Test Voltage . . . . . . . . . . . . . . . . 89 Table 21: Insulation Resistance Minimum Values . . . . . . . . . . . . 89 Table 22: Dielectric Absorption Chart . . . . . . . . . . . . . . . . . . . . 92 Table 23: Polarization Index . . . . . . . . . . . . . . . . . . . . . . . . . . . 93 Table 24: Rotor Bar Failure Levels . . . . . . . . . . . . . . . . . . . . . 107 Table 25: Electrical and Mechanical Faults . . . . . . . . . . . . . . . . 108 Table 26: Motor System Diagnostic Technology Comparison . . 118 Table 27: Management Considerations . . . . . . . . . . . . . . . . . . 119 Table 28: Common Approaches . . . . . . . . . . . . . . . . . . . . . . . 119 Table 29: Additional Considerations . . . . . . . . . . . . . . . . . . . . 120 Table 30: Example Number 1 . . . . . . . . . . . . . . . . . . . . . . . . . 124 Table 31: Reading Change Table for AC Rotating Machinery . 136 Table 32: Ambient Conditions . . . . . . . . . . . . . . . . . . . . . . . . 139 Table 33: Load Conditions (Condition Variables) . . . . . . . . . . . 140 Table 34: Start/Stop . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140 Table 35: MCA Tolerances (Assembled Motor – All sizes) . . . . . 145 Table 36: MCA Tolerances (Disassembled Motor – All sizes) . . . 146 Table 37: MCA Interpretation (1 of 2) . . . . . . . . . . . . . . . . . . 147 Table 38: MCA Interpretation (2 of 2) . . . . . . . . . . . . . . . . . . 148 Table 39: Sample Test Results . . . . . . . . . . . . . . . . . . . . . . . . . 158

Penrose_FM.indd Sec18:xx

4/22/08 1:57:07 AM

Table of Tables

xxi

Table 40: Adjusted Frequency Sample, Good . . . . . . . . . . . . . . 158 Table 41: Adjusted Frequency Sample, Bad . . . . . . . . . . . . . . . 158 Table 42: Phase Unbalance Test Data . . . . . . . . . . . . . . . . . . . . 159 Table 43: Alternate Motor . . . . . . . . . . . . . . . . . . . . . . . . . . . 161 Table 44: Modification to Rotor . . . . . . . . . . . . . . . . . . . . . . . 165 Table 45: Good 150 HP Motor. . . . . . . . . . . . . . . . . . . . . . . . 166 Table 46: Initial Synchronous Motor Tests (8,000 hp) . . . . . . . . 167 Table 47: Synchronous Motor, Second Reading . . . . . . . . . . . . 167 Table 48: Traction Motor Test Data . . . . . . . . . . . . . . . . . . . . . 168 Table 49: Good Spindle Motor . . . . . . . . . . . . . . . . . . . . . . . . 169 Table 50: Shorted DC Brushless Servo . . . . . . . . . . . . . . . . . . 169 Table 51: Winding Condition - Dirty Windings . . . . . . . . . . . . 170 Table 52: Shorted Servo Data . . . . . . . . . . . . . . . . . . . . . . . . . 171 Table 53: Rotor Position Readings . . . . . . . . . . . . . . . . . . . . . 172 Table 54: Shorted Windings Measured from MCC . . . . . . . . . . 173 Table 55: New 300 HP Motor with Winding Fault . . . . . . . . . 174 Table 56: Recommended Brush Tension . . . . . . . . . . . . . . . . . 185 Table 57: DC Motor Test Frequency . . . . . . . . . . . . . . . . . . . . 192 Table 58: Rotor Analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 204 Table 59: Motor Fault Signatures . . . . . . . . . . . . . . . . . . . . . . 205

Penrose_FM.indd Sec18:xxi

4/22/08 1:57:07 AM

Penrose_FM.indd Sec18:xxii

4/22/08 1:57:07 AM

Table of Equations

Equation 1: Energy Stored in the Magnetic Field . . . . . . . . . . . . 12 Equation 2: Transformer Voltage . . . . . . . . . . . . . . . . . . . . . . . . 13 Equation 3: Transformer Current . . . . . . . . . . . . . . . . . . . . . . . 13 Equation 4: Inductive Reactance . . . . . . . . . . . . . . . . . . . . . . . 13 Equation 5: Capacitive Reactance . . . . . . . . . . . . . . . . . . . . . . . 14 Equation 6: Impedance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 Equation 7: Synchronous Speed . . . . . . . . . . . . . . . . . . . . . . . . 21 Equation 8: Number of Poles . . . . . . . . . . . . . . . . . . . . . . . . . . 21 Equation 9: Percent Slip (s) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 Equation 10: Slip Frequency . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 Equation 11: Horsepower and Torque Relationship . . . . . . . . . . 23 Equation 12: Efficiency Equation . . . . . . . . . . . . . . . . . . . . . . . 24 Equation 13:Voltage Turns Ratio . . . . . . . . . . . . . . . . . . . . . . . 29 Equation 14: Current Turns Ratio . . . . . . . . . . . . . . . . . . . . . . 29 Equation 15: Load Impedance . . . . . . . . . . . . . . . . . . . . . . . . . 30 Equation 16: Equivalent Primary Impedance . . . . . . . . . . . . . . . 30

xxiii

Penrose_FM.indd xxiii

4/22/08 1:57:07 AM

xxiv

Electrical Motor Diagnostics: 2nd Edition

Equation 17: Internal Impedance . . . . . . . . . . . . . . . . . . . . . . . 30 Equation 18: Capacitance of Each Portion of Circuit . . . . . . . . . 42 Equation 19: Insulation Boundary . . . . . . . . . . . . . . . . . . . . . . . 42 Equation 20: Phase Capacitance . . . . . . . . . . . . . . . . . . . . . . . . 43 Equation 21: Circuit Inductance . . . . . . . . . . . . . . . . . . . . . . . . 43 Equation 22: Inductance with ‘N’ Coils . . . . . . . . . . . . . . . . . . . 43 Equation 23: Per Phase Impedance . . . . . . . . . . . . . . . . . . . . . . 43 Equation 24: Inductive Reactance. . . . . . . . . . . . . . . . . . . . . . . 44 Equation 25: Capacitive Reactance . . . . . . . . . . . . . . . . . . . . . . 44 Equation 26: Simple Impedance (Complex AC Resistance) . . . . 44 Equation 27: Simplified Circuit Impedance Example . . . . . . . . . 44 Equation 28: Mutual Inductance . . . . . . . . . . . . . . . . . . . . . . . . 48 Equation 29: Primary emf . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50 Equation 30: Secondary emf . . . . . . . . . . . . . . . . . . . . . . . . . . . 50 Equation 31: Turns Ratio . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50 Equation 32: Ratio of Primary and Secondary Impedance . . . . . 50 Equation 33: Rotor Influence . . . . . . . . . . . . . . . . . . . . . . . . . . 50 Equation 34:Voltage Deviation (460 Volt Motor) . . . . . . . . . . . . 75 Equation 35:Voltage Unbalance (1) Determine the Average Voltage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75 Equation 36:Voltage Unbalance (2) Determine the Unbalance . . 75 Equation 37: Average Voltage (Va) . . . . . . . . . . . . . . . . . . . . . . . 78 Equation 38: Average Current (Aa) . . . . . . . . . . . . . . . . . . . . . . 79 Equation 39: Percentage of Motor Load . . . . . . . . . . . . . . . . . . 79 Equation 40: Calculating the kVA/HP (3 Phase Motors) . . . . . . 80

Penrose_FM.indd xxiv

4/22/08 1:57:08 AM

Table of Equations

xxv

Equation 41: Calculating the Locked Rotor Amps (LRA) . . . . . 81 Equation 42: Balance Equation Double Kelvin Bridge . . . . . . . . 84 Equation 43: Resistance of Cold Winding . . . . . . . . . . . . . . . . . 86 Equation 44: Resistance of Hot Winding . . . . . . . . . . . . . . . . . . 86 Equation 45: Series Resistances (Rs) . . . . . . . . . . . . . . . . . . . . . 86 Equation 46: Parallel Resistances (Rp) . . . . . . . . . . . . . . . . . . . . 87 Equation 47: Resistive Losses . . . . . . . . . . . . . . . . . . . . . . . . . . 87 Equation 48: New Winding Surge Voltage Maximum . . . . . . . . 103 Equation 49: Used Winding Surge Voltage . . . . . . . . . . . . . . . . 103 Equation 50: Pole Pass Frequency . . . . . . . . . . . . . . . . . . . . . . 107 Equation 51: Mean Time To Failure . . . . . . . . . . . . . . . . . . . . 124 Equation 52: Mean Time for Corrective Maintenance . . . . . . . 124 Equation 53: The Failure Rate . . . . . . . . . . . . . . . . . . . . . . . . 125 Equation 54: Modified Inherent Unavailability . . . . . . . . . . . . 125 Equation 55: The Reliability Function (Inherent Availability) . . 125 Equation 56: Series Availability . . . . . . . . . . . . . . . . . . . . . . . . 126 Equation 57: Parallel Availability (2 systems) . . . . . . . . . . . . . . 126 Equation 58: Parallel Availability (3 or more identical systems) . 126 Equation 59: Winding Contamination Multiplier . . . . . . . . . . . 139 Equation 60: Log Scale Used . . . . . . . . . . . . . . . . . . . . . . . . . 139 Equation 61: Winding Short Formula . . . . . . . . . . . . . . . . . . . 141 Equation 62: Winding Contamination . . . . . . . . . . . . . . . . . . . 141 Equation 63: Winding Short . . . . . . . . . . . . . . . . . . . . . . . . . . 142 Equation 64: Winding Contamination . . . . . . . . . . . . . . . . . . . 142 Equation 65: Static Eccentricity . . . . . . . . . . . . . . . . . . . . . . . 202

Penrose_FM.indd xxv

4/22/08 1:57:08 AM

xxvi

Electrical Motor Diagnostics: 2nd Edition

Equation 66: Number of Poles . . . . . . . . . . . . . . . . . . . . . . . . 214 Equation 67: Slot Multiplier . . . . . . . . . . . . . . . . . . . . . . . . . . 215 Equation 68: Lower Frequency Value. . . . . . . . . . . . . . . . . . . . 216 Equation 69: Upper Frequency Value . . . . . . . . . . . . . . . . . . . 216 Equation 70: Parallel Capacitors . . . . . . . . . . . . . . . . . . . . . . . 257

Penrose_FM.indd xxvi

4/22/08 1:57:08 AM

Introduction

In 1979, an MIT study on tribology estimated $200 Billion dollars were spent on the direct costs of Reliability and Maintenance (R&M). At the time, it was also estimated that over 14% of the 1979 USA Gross Domestic Product (GDP) was lost opportunity due to improper R&M practices. This level increased to approximately 20% of the $12.5 Trillion USA GDP in 2005, or approximately $2.5 Trillion in lost business opportunity. This is greater than all of the national GDP’s of all but the top three economies of the United States, Japan and Germany (Table 1). At the present time, it is estimated that the size of the R&M industry is ~$1 Trillion with $500 – 750 Billion related to breakdown (reactive) maintenance or generally poor, incorrect or excessive practices. Table 1: Lost Potential in the USA Compared to Top Ten Economies Country

2005 GDP in USD

United States Japan Germany

$12.5 T $4.5 T $2.8 T Lost Opportunity USA: ~$2.5 T

China United Kingdom France Italy Spain Canada Brazil

$2.2 T $2.2 T $2.1 T $1.7 T $1.1 T $1.1 T $0.8 T

The primary cause of the loss is that over 60% of maintenance programs are reactive, which includes those programs that were initiated and later failed due to ‘maintenance entropy.’ Over 90% of maintenance initiatives fail, 57% of CMMS applications fail and over 93% of motor management programs fail.The primary reason is that the present xxvii

Penrose_FM.indd Sec19:xxvii

4/22/08 1:57:08 AM

xxviii

Electrical Motor Diagnostics: 2nd Edition

business mindset calls for immediate improvements whereas it normally takes 12 to 24 months for a supported program to take hold and begin to show results. This ‘rule of thumb’ does not just apply to R&M, but to all business practices.

Figure 1: Cost Impact of Maintenance Programs $/HP/YR

The opportunities through properly implemented, planned and sustained reliability and maintenance programs are inspiring. The USA has the potential of improving electrical energy consumption by over $26.5 Billion (2006) and resulting greenhouse gas emissions by over 3,000 Mega-Tons per year. Individual maintenance departments can see troubleshooting and evaluation times improved by over 50%, repair costs by over 30% and general reactive maintenance-related labor costs by up to 50%. Corporate opportunities, as outlined in Table 2, can be significant. However, it must be identified that apparent direct cost and impacts may actually increase depending on the existing program and

Penrose_FM.indd Sec19:xxviii

4/22/08 1:57:08 AM

Introduction

xxix

the implementation of a program over a non-existing program while the benefits are seen business-wide. Table 2: Average Impact of CBM Monitoring Programs Opportunity Reduction in maintenance costs Elimination of unplanned outages Reduction in downtime Increase in throughput Reduction of PM’s Man-hour utilization improvements

Improvement 24–30% 70–75% 35–40% 20–25% 33–66% 40–50%

The R&M function represents the single greatest opportunity for corporate profitability through operations should a program be approached with a logical process. It also represents the least understood opportunity as most companies approach R&M with ‘programs of the month’ and promote ‘hero maintenance.’ In effect, companies will attempt to start a program, not see any immediate and direct cost or production improvements and will abandon the program. Additionally, most programs will congratulate personnel who are involved in returning emergency breakdowns back into service while ignoring successful breakdown avoidance in planned maintenance programs. The correct approach would be to investigate the reasons why a breakdown occurs while supporting and recognizing the avoidance of other such situations. The general mindset of most executives has turned away from the opportunities presented in the R&M function and have, instead, focused on obtaining significant returns on asset investments without investing in their upkeep. In programs where the assets are new or a robust program existed, there will be some level of inertia before significant reductions in throughput and increases in direct maintenance costs occur. At this point, the response tends to be a cut in R&M budget. Once this situation starts, it becomes a self-feeding recipe for corporate failure. Electric motor systems provide one of the most significant common opportunities across all industries. The DC electric motor initiated the second industrial revolution in the early 1800’s, followed later with the introduction of AC motors, generators and distribution systems in 1888. In 1997, electric motor systems consumed 20% of all energy used in North America, 57% of all electric power generated and 70% to 90% of process industry energy. The population of electric motors estimated by the US Department of Energy in 1997, was 1.2 Billion in the USA with

Penrose_FM.indd Sec19:xxix

4/22/08 1:57:09 AM

xxx

Electrical Motor Diagnostics: 2nd Edition

a projected 200,000 new motors added per year, 96% were estimated less than 5 horsepower, 2.5% were between 5 and 25 horsepower and the remaining 1.5% were over 25 horsepower. The motors over 25 horsepower consume over 60% of all electric motor energy with a majority being fans, pumps and compressors. These populations and numbers represent electric motors of all types, both AC and DC. With their importance to the operation of modern industrial and manufacturing, as well as power generation, transportation, distribution and other applications, the maintenance of the electric motor system is critical. The electric motor system consists of the distribution system, controls, electric motor, coupling, load and process. For the most part, technologies have been introduced which evaluate the mechanical condition of the motor and driven equipment and power analysis systems that evaluate the distribution system. As far as the actual energy converter, or motor, most systems were limited to testing the groundwall insulation system of the motor windings with few technologies or techniques that could accurately evaluate the remainder of the winding and motor systems with the equipment in-place. Since the mid-1980’s, new technologies have been introduced in both the high and low voltage testing (method of test) realms that allow the technician or motor manager to evaluate the condition of the complete insulation system. These systems varied from field-capable surge comparison testers, which are used to apply high voltage impulses in order to evaluate the turn insulation strength of winding systems and low voltage motor circuit analysis devices, which use applied voltage less than 10Volts AC and higher frequencies to evaluate the turn insulation condition. Another new technology allows for evaluating the condition of the motor system under operating conditions using voltage and current analysis. The purpose of this textbook is to provide the student with information starting from basic electricity through the development of an electric motor system maintenance and management program utilizing the Multi-Technology Approach™ to Electrical Motor Diagnostics. It is assumed that the student will have a basic understanding of industrial applications, some level of electrical or mechanical knowledge as well as electric motor applications. The test technology portion of the book will focus on motor circuit analysis in addition to electrical and current signature analysis.

Penrose_FM.indd Sec19:xxx

4/22/08 1:57:09 AM

Chapter

1 The Motor Diagnostics and Motor Health Study

Mechanical faults in electric motors comprise of approximately 53% of failure while winding and rotor faults make up the remaining 47% of faults, according to EPRI and EASA post-mortem studies. Of the 47% of motor rotor and winding faults, depending on the study, 5–10% are related to electric motor rotors. The remainder are electrical winding faults which normally start as a short between conductors. Prior to 1980, the primary methods for evaluating the condition of electric motors consisted of: Resistance, including milli-Ohm testing; Insulation resistance to ground testing; Hi-Potential testing; Surge comparison testing; Vibration analysis; and, Voltage and Current testing. Ultrasonics and infrared technologies were added to the motor system testing arsenal. Each method had its strengths and weaknesses, specific levels of training required and intrusiveness for testing. In the 1980’s, a number of companies introduced a variety of new technologies that viewed the electric motor windings. Although each technology provided a different basic set of test results, that varied in degrees of accuracy, they were combined under the heading of motor circuit analyzers (MCA). In the 1980’s and 1990’s, motor current signature analysis (MCSA) and electrical signature analyzers (ESA) instruments were introduced to the market. By the end of the 1990’s, the combined technologies fell under the umbrella of the term ‘Electrical Motor Diagnostics,’ or EMD.

1

Penrose_BMI.indd 1

4/22/08 5:41:30 AM

2

Electrical Motor Diagnostics: 2nd Edition

The purpose of the Motor Diagnostics and Motor Health (MDMH) study, published in 2003, was to review motor diagnostic technologies, studies and a specific maintenance and reliability survey. The purpose was to provide a comprehensive overview which encompassed: 1. An understanding of motor management and motor diagnostic needs by industry at a global level. 2. An understanding of the perception of technology capabilities by reliability and maintenance. 3. An understanding or potential improvements to competitiveness of companies through the application of motor diagnostic technologies. 4. A roadmap for motor diagnostic companies and users alike. In this chapter, we will cover the highlights of the study.

Electric Motor Repair Study, 1995 The “Industrial Motor Repair in the United States” was a third party study funded by the Bonneville Power Administration (BPA) and performed by the Washington State Energy Office (now the Washington State University Energy Extension Center). The purpose of the report was to:  Characterize the motor repair industry in the United States;  Summarize current motor repair and testing practice; and,  Identify barriers to energy efficient motor repair practice and recommend strategies for overcoming these barriers. The particular areas selected by the MDMH from the BPA study were in the area of testing performed and potential impact on post-repair reliability. According to the report, “The shops surveyed had a strong craftsman ethic and a desire to do good work despite customer requirements for fast turnaround.”1 As such, it was assumed that the results of the study were due to responses from above average quality electric motor repair shops, of which the report estimates there are over 4,100 in the United States. Over half, 2,700 at the time of the report, were Electrical Apparatus Service Association (EASA) members. 1

Schueler, Leistner and Douglass, Industrial Motor Repair in the United States, Bonneville Power Administration, 1995.

Penrose_BMI.indd 2

4/22/08 5:41:33 AM

The Motor Diagnostics and Motor Health Study

3

“Only one-third of the shops used written quality assurance standards of any type and were familiar with quality assurance procedures. Testing practices vary widely from shop to shop. Testing was most often used as a diagnostic tool for troubleshooting. Although insulation, winding resistance, vibration and core loss testing should be done routinely as part of a quality repair, only insulation testing was done regularly.”2

Figure 2: Modifications to Windings Through Rewind Process

One of the key findings of the BPA/WSU report was that 81% of motor repair shops modified windings through the rewind-repair process. 37% modified the windings due to shop preference; 36% for ease of winding, especially in the tight slots of energy and premium efficient electric motors; 13% for other reasons. Only 10% were modified at the customers’ request and 4% for improved reliability or energy. The problem, in this case, is that some modifications are performed using an engineering approach so the motor runs effectively the same, while most are reductions in wire size to reduce the circular mils to an ‘acceptable’ level while allowing more room in the slot. This increases one of the more significant motor losses, the I2R, which both reduces the motor efficiency and increases the operating temperature,reducing the reliability/ life of the insulation system. On average, fewer than 50% of motor repair shops reported, in the survey, that testing of any kind was performed on motors through the 2

Schueler, et.al.

Penrose_BMI.indd 3

4/22/08 5:41:34 AM

4

Electrical Motor Diagnostics: 2nd Edition

Figure 3: Testing Through the Repair Process

repair process, including simple Meg Ohm testing. Most repair shops were also found not to run the motors, before or after repair, or to only run them for a short time on reduced voltage.

In Service Motor Testing, 1999 Performed in 1999 by the Washington State University Energy Extension Center (WSU), the focus of the study was to determine if, and what, companies were doing to test their electric motors. While the focus was originally supposed to be related to the energy efficiency of electric motors, it did discuss testing the reliability of the electric motors, the primary focus of most of the motor owners. According to the study, of the sites visited, 73% of sites performed some type of motor testing in-house. The general response was that highly invasive test methods are not desired by the companies as a majority of them operated 24/7. Their preference for technologies were that they:    

Penrose_BMI.indd 4

Were non-invasive; Simple and portable; Reasonable and accurate; Cost effective

4/22/08 5:41:36 AM

The Motor Diagnostics and Motor Health Study

5

Figure 4: Reasons for Motor Testing

Figure 5: Technologies Used

Penrose_BMI.indd 5

4/22/08 5:41:36 AM

6

Electrical Motor Diagnostics: 2nd Edition

A majority of the companies involved in the study performed some level of condition-testing for their electric motors with the exception of any testing used to detect winding defects, even when troubleshooting and none that can be used to estimate (Predictive Maintenance) the possibility of a developing winding short.

Success by Design Best Practice Survey One of the keys to the report was a general survey of ReliabilityWeb.com users. A survey of questions was placed and the response was provided by approximately 2% of those contacted. The respondents were evenly split between USA domestic and outside of the USA. The results of the study were interesting as the average respondent stated that their average cost per hour downtime was $10,000 and almost all claimed 24/7 facilities. Within this group, 68% claimed they had a motor program in place of which a majority said they did not see a return on investment. However, they counted insulation resistance testing, vibration, current and visual inspections as motor testing. Of the 38% of that 68% who were actually performing motor diagnostics (MCA or ESA/MCSA), 91% saw a high return on investment.

Figure 6: $/Hour Downtime

Penrose_BMI.indd 6

4/22/08 5:41:37 AM

The Motor Diagnostics and Motor Health Study

7

Figure 7: Shifts and Operation

Figure 8: Primary Driver for Motor Testing Program

Dealing with the State of the Motor Testing Industry The electric motor system can impact the profitability, throughput, operating costs and inventory level requirements of a company. Some specific measures can be put in place to estimate failure, reduce troubleshooting times and control quality of new or repaired electric motors coming into the plant. These include: 1. The development of a motor management program; 2. Spare motor storage program; 3. Commissioning program; and, 4. Time to Failure Estimation (TTFE™)

Penrose_BMI.indd 7

4/22/08 5:41:38 AM

8

Electrical Motor Diagnostics: 2nd Edition

The use of Electrical Motor Diagnostics (EMD) technologies, in combination with a multi-technology approach to motor diagnostics, is a key component to achieving these goals. While the scope of this book is not to discuss these topics in detail, they will be considered important in how we approach the discussion of testing electric machines from the standpoint of commissioning, maintaining and estimating (PdM) equipment faults. The complete Motor Diagnostics and Motor Health Study can be downloaded from http://www.motordiagnostics.com.

Penrose_BMI.indd 8

4/22/08 5:41:39 AM

Chapter

2 Basic Electricity and Electro-Magnetism

In order to understand the operation of an electric motor and its insulation system, a study of basic electricity and electric motor operation is necessary. For this, we will start with the makeup of the classical atom, or the Bohrs Model. For the purpose of this book, a full study of

Figure 9: Bohrs Model of the Atom

9

Penrose_BMI.indd 9

4/22/08 5:41:39 AM

10

Electrical Motor Diagnostics: 2nd Edition

electricity and electro-magnetism is not necessary, just an understanding of the principles. In this model, the atom is represented as protons (positive charge) and neutrons (no charge) of equal numbers being orbited by electrons (negative charge). Some atoms have more electrons than protons and are considered negatively charged ions while some have fewer electrons than protons and are considered positively charged ions. Electrons can be gained and lost by some types of atoms easily. When an electron is emitted from an atom, a photon of light is also emitted at a specific frequency relating to the atom. When enough are emitted, they can be seen in the visible or invisible light spectrum. The number of protons and neutrons determine the atom’s atomic weight and what element it represents. How the atoms of materials act will determine if the material is a conductor, a dielectric or an insulator:  Conductors: Have free electrons in the material that are easily directed. They are usually metals such as copper, aluminum, gold or platinum.  Semi-Conductors: Also referred to as dielectrics have four valence electrons and will polarize when an electrical field is applied.  Insulators: No free electrons and are inert when an electrical field is applied. Insulators include ceramics, glass and mica.

Current, Voltage and Resistance Whether discussing the topic in terms of direct or alternating current, the basic elements of electricity include current, voltage and resistance. Current (I) is the flow of electricity much like the flow of water in a pipe. It is measured in Amperage as opposed to gallons per minute of water.The measurement of current is standardized as 1 Amp being equal to 6.28 × 1018 electrons passing one point in one second.This value is also termed as 1 Coulomb of electrons being equal to one Amp.The electron charge is therefore equal to 1.60219 × 10–19 Coulombs. For the purpose of this book, we will determine that amperage flows from the negative (excess electrons) charge to positive charge. Voltage (V or E) is the electrical pressure in the system much like water pressure in a pipe. Electrical pressure is measured in Volts as opposed to pounds per square inch and the greater the value of Voltage, the greater the electrical pressure. This can also be considered ‘electrical

Penrose_BMI.indd 10

4/22/08 5:41:39 AM

Basic Electricity and Electro-Magnetism

11

potential,’ with 1 Volt equaling 1 Joule/Coulomb (Q) representing the work to move the electrical charge against a field. If we now consider this as being similar to air pressure in a compressed air system, if both the inside of the compressed air system is equal to the outside, then no current will flow. If the pressure inside the system increases, the potential energy to move current increases. This is an important concept that we will explore later as, in order to determine Voltage, we must have a reference. For example, if you were standing in a room and the floor had a potential equal to 4,160 Volts to the hallway outside and a table in the room had a potential equal to 4,160 Volts, you would be able to touch the table and walk around in the room safely. If, however, you place one foot outside into the hallway while one foot was still touching the floor in the room, the potential of 4,160 Volts would exist and current will flow. Resistance (Ohms – R or Ω) is simply the restriction to current flow in a circuit. Such things as wire size and material used will determine the resistance of an object. Resistance is related to voltage and current in the relationship known as Ohm’s Law where Resistance is equal to Voltage divided by Current (R = V/I). This relationship stands in both AC and DC circuits. Another relationship between Voltage, Current and Resistance is Power which is represented in terms of Watts in electrical systems. The Watt is equal to 1 Joule per Second and represents energy changing from one form to another, such as electrical energy to heat, or magnetic fields to torque. In electrical terms, Watts are equal to the Current squared times Resistance (I2R).

Electro-Magnetic Fields Magnets are materials that have both a north and south pole which relates to the magnetic fields surrounding the material. These fields are often defined in terms of lines of magnetic flux which flows from the South Pole and into the North Pole. Permanent magnets are materials that retain magnetic fields for limited time through an indefinite period. If another material that can be magnetized is brought in range of the material, it will expand the magnetic field around the additional material. Another method of introducing a magnetic field is to pass a current through a conductor. As current flows through the conductor, it generates a field around itself. If the conductors are wound in a coil, the fields add together and the field passes axially through the center of the coil.

Penrose_BMI.indd 11

4/22/08 5:41:39 AM

12

Electrical Motor Diagnostics: 2nd Edition

The direction of the field will be based upon the direction the current is flowing and the direction the coil is wound. As a voltage is applied to an electro-magnetic coil, the magnetic field and current lags behind the building voltage. When voltage is removed, the field collapses relatively slowly and current continues to flow. If an alternating voltage is applied, the current and magnetic field lags behind the increasing and decreasing voltage.

Alternating Current Electricity (Hz) The term Alternating Current, or AC, is applied to both voltage and current to describe an increasing and decreasing value in relation to the number of times it goes from positive to negative to positive again in one second, or Hertz (Hz). The frequency can be in the form of a sinusoidal waveform, square wave or a variation of the two. As we are dealing with electric machines in this textbook, the first two terms we will define are the inductance and capacitance of the windings and insulation system. Most of the work that we will discuss in MCA will relate directly to these two components of the motor circuit. The inductance of a coil is the winding’s capacity to develop a magnetic field per amp, or store electromagnetic energy in its magnetic field. This value is presented in Henries (H) or milli-Henries. Equation 1: Energy Stored in the Magnetic Field W = ½ Li2 Where W is the energy stored in the magnetic field, L is the coefficient of self-inductance and i is the current through the coil. With the above inductance value being related to the ‘self-inductance’ of the coil, or the result of inductance related to a lone coil, another coil brought into proximity with the coil would result in mutual inductance. In effect, when two coils are in close proximity, a changing voltage and current in one coil will induce a changing voltage and current in the other coil. This concept, also known as transformer theory, provides a significant amount of information concerning how an induction motor works and the relationship of a perfect transformer can be see in Equations 2 and 3.

Penrose_BMI.indd 12

4/22/08 5:41:39 AM

Basic Electricity and Electro-Magnetism

13

Equation 2: Transformer Voltage

V2 = V1 ×

N2 N1

Where V1 is the primary coil voltage, V2 is the secondary coil voltage, N1 is the number of turns in the primary, N2 is the number of turns in the secondary The voltage is inversely proportional to current, as shown in Equation 3. Equation 3: Transformer Current

I 2 = I1 ×

N1 N2

Where I1 is the primary coil current, I2 is the secondary coil current, N1 is the number of turns in the primary, N2 is the number of turns in the secondary We will address more advanced application of inductance later in this textbook as it relates to the operation of an electric motor or transformer. However, we will introduce the AC resistive property of inductance known as the Inductive Reactance (XL) of the circuit where: Equation 4: Inductive Reactance

X L = 2π fL Where f is the applied frequency and L is the inductance The capacitance is presented in Farads, micro-Farads and pico-Farads, which we will treat in detail while discussing insulation systems. A capacitor is defined as a device that stores electrical potential in an electric field and is normally constructed of a dielectric between two plates. In effect, when a potential is placed across the capacitor in a specific direction it will maintain that charge until a discharge path is present, or the potential is reversed. Capacitance causes current to lead voltage. The AC

Penrose_BMI.indd 13

4/22/08 5:41:40 AM

14

Electrical Motor Diagnostics: 2nd Edition

resistive property of capacitance is known as the Capacitive Reactance (XC) of the circuit where: Equation 5: Capacitive Reactance 1 XC = 2π fC Where f is the applied frequency and C is the capacitance The relationship of these components can be represented as a simplified complex resistance, or Impedance, formula: Equation 6: Impedance

Z=

R 2 + ( X L − XC )

2

Where R is the simple DC resistance of the circuit With these basic definitions we can discuss the operation of different types of electrical machines.

Penrose_BMI.indd 14

4/22/08 5:41:40 AM

Basic Electricity and Electro-Magnetism

15

Chapter 2 Questions 1. What kind of charge does a neutron have in the Bohr’s Model atom? 2. Define an insulator. 3. Define Current. 4. Define Voltage. 5. How many poles are there in an electro-magnet? 6. What is the effect on current as it relates to voltage in an inductor? 7. What is the complex AC resistance of a circuit?

Penrose_BMI.indd 15

4/22/08 5:41:41 AM

Penrose_BMI.indd 16

4/22/08 5:41:41 AM

Chapter

3 Electric Motor Theory

Most textbooks separate AC and DC motor theory and treat them separately. In this case, we are going to treat all types of AC and DC electric motors working from the basic theories of AC and DC machines. This understanding of operating theory is important for the analysis of an electric machine.

AC Induction Motor Theory The simplified circuit for one phase of a three phase induction motor can be found in Figure 10. As you can see, the circuit consists of resistance, inductance, capacitance, inductive and capacitive reactances and the overall impedance.

Figure 10: Basic Motor Circuit

17

Penrose_BMI.indd 17

4/22/08 5:41:41 AM

18

Electrical Motor Diagnostics: 2nd Edition

The AC induction motor consists of a number of basic components. Those include the following: • The stator frame: This component protects the interior of the electric motor and contains components of the electric motor. For most motors, the surface of the stator frame is set up to expel heat from the stator core. • The end shields: Coupled with the stator frame, the end shields protect the interior of the electric motor and hold the bearings centered through stator. • The bearings: Hold the rotor centered within the motor airgap while reducing the friction of the operation of the motor. • The stator core: The stator core both holds the stator windings and directs the magnetic fields developed in towards the airgap (space between stator and rotor cores). The material used is selected to reduce hysteresis losses (the resistance to a changing magnetic field) and is generally 19 to 49 mils thick and insulated from each other in order to reduce eddy-current losses. Older core steels tended to be a low-carbon annealed steel with energy and premium efficient motors requiring silicone steels. The laminations are stamped or cut (laser) in such a way that they have the same magnetic ‘grain’ as well as slots for the stator windings. The steel on either side of the slots are referred to as the teeth and the outer diameter is referred to as the ‘back iron.’ The back iron is thicker on higher speed motors and gets thinner as the synchronous RPM of the motor is lower. • The stator windings: In three phase motors, there are three separate phases of winding groups. The phases are separated by 120 electrical degrees with incoming voltage and resulting current separated by approximately 120 electrical degrees. The result is a rotating field. The number of slots in a stator core and the number of coils will be multiples of the number of phases by the number of poles, which are always in pairs. For instance, if there 4 poles in a three phase machine, the minimum number of slots is 3 × 4 = 12. However, there are often more than one coil per group.

Penrose_BMI.indd 18

4/22/08 5:41:41 AM

Electric Motor Theory

19

• The rotor core:The rotor core is the exact reverse of the stator core with the rotor windings being close to the outer diameter surface towards the stator core. • The squirrel cage: The first AC induction motor type we will discuss will be the squirrel-cage rotor type. This consists of rotor bars inserted through the rotor slots and shorted together at both ends.

Figure 11: Three Phase Stator and Windings

Penrose_BMI.indd 19

4/22/08 5:41:41 AM

20

Electrical Motor Diagnostics: 2nd Edition

Figure 12: Three Phase Rotor

The purpose of the AC induction motor is to convert electrical energy into mechanical torque. This is accomplished through the interaction of the magnetic fields of the stator and rotor in the air gap. As three phase voltage is supplied to the motor windings separated by 120 electrical degrees, the windings are also separated by 120 electrical degrees. For each magnetic pole pair, as the voltage and resulting current increase one pole becomes more north and one more south. This alternates within the stator core and windings such that a ‘whirlpool’ effect occurs within the air gap of the stator. The magnetic fields cut through the rotor bars generating a current that creates an opposite magnetic pole. Now, this is the part that is difficult to understand: The magnetic field of the rotor follows the stator rotating field by ~90 electrical degrees regardless of the speed of the rotor, or if it is at rest. The rotor core condenses and directs this rotating field back into the air gap. The rotor assembly begins to follow the rotating field.

Penrose_BMI.indd 20

4/22/08 5:41:42 AM

Electric Motor Theory

21

Figure 13: Rotating Field and Rotor Cage

The speed of the rotating magnetic field can be determined as: Equation 7: Synchronous Speed 120 f Ns = p Where f is the line frequency and p is the number of poles found as: Equation 8: Number of Poles p=

Number of Groups of Coils 3

The number of poles is normally expressed as an even number. The actual output speed of the rotor is related to the synchronous speed via the slip, or percent of slip: Equation 9: Percent Slip (s) N −N %s = s × 100% Ns

Penrose_BMI.indd 21

4/22/08 5:41:43 AM

22

Electrical Motor Diagnostics: 2nd Edition

The actual frequency seen within the rotor windings is known as the slip frequency (fs): Equation 10: Slip Frequency fs =

Ns − N Ns

× 60

When the motor is first started and the rotor is stationary, the slip is 100%. The result is that the rotor is seeing full line frequency and the rotor bars and rotor core saturate magnetically.This causes a very high rotor current which is reflected on the stator windings and can be measured as the 4–8 times nameplate ‘inrush’ current. As the rotor speed approaches synchronous speed, the rotor frequency drops and the current falls. If unloaded, the slip is very small and the current is relatively low (25–40% of full load current at no load). As the motor is loaded, it slows until it reaches full load and full load speed. The amount of current measured is a direct result of the rotor current plus the losses of the motor. Table 3: Motor Synchronous and Operating Speed (60 Hz, NEMA Design B) Number of Poles

Synchronous Speed

Common Actual Speed

2

3600

3450

4

1800

1735–1790

6

1200

1150

8

900

885

The operating torques of an electric motor are defined as: • Full Load Torque:The full load torque of a motor is the torque necessary to produce its rated horsepower at full-load speed. In pounds at foot radius, it is equal to the horsepower times 5250 divided by the full-load speed. • Locked Rotor Torque: The locked rotor torque of a motor is the minimum torque which will develop at rest for all angular positions of the rotor, with rated voltage applied at rated frequency. • Pull-Up Torque: The pull-up torque of an alternating current motor is the minimum torque developed by the motor during the period of acceleration from rest to the speed at which breakdown torque occurs. For motors which do not have a definite breakdown torque, the pullup torque is the minimum torque developed up to rated speed.

Penrose_BMI.indd 22

4/22/08 5:41:43 AM

Electric Motor Theory

•

23

Breakdown Torque: The breakdown torque of a motor is the maximum torque which it will develop with rated voltage at rated frequency without an abrupt change in speed.

By varying the resistance within the rotor bars of a squirrel cage rotor, the amount of torque output of the motor varies. If the resistance is increased, starting torque increases and breakdown torque decreases and the slip increases. If resistance is decreased, the starting torque decreases and the breakdown torque increases. The motor horsepower is a direct relation of motor output speed and torque (expressed in lb-ft). Equation 11: Horsepower and Torque Relationship Horsepower =

RPM × Torque 5250

NEMA defines, in NEMA MG 1-1993, four motor designs dependant upon motor torque during various operating stages: Design A: Has a high starting current (not restricted), variable lockedrotor torque, high break down torque, and less than 5% slip. Design B: Known as “general purpose” motors, have medium starting currents (500 to 800% of full load nameplate), a medium locked rotor torque, a medium breakdown torque, and less than 5% slip. Design C: Has a medium starting current, high locked rotor torque (200 to 250% of full load), low breakdown torque (190 to 200% of full load), and less than 5% slip. Design D: Has a medium starting current, the highest locked rotor torque (275% of full load), no defined breakdown torque, and greater than 5% slip. Design A and B motors are characterized by relatively low rotor winding resistance. They are typically used in compressors, pumps, fans, grinders, machine tools, etc. Design C motors are characterized with dual sets of rotor windings. A high resistive rotor winding to introduce a high starting torque and a low resistive winding for a medium breakdown torque. They are typically used on loaded conveyors, pulverizers, piston pumps, etc. Design D motors are characterized by high resistance rotor windings. They are typically used on cranes, punch presses, etc.

Penrose_BMI.indd 23

4/22/08 5:41:44 AM

24

Electrical Motor Diagnostics: 2nd Edition

The electrical energy that is consumed by an electric motor is accounted for in losses. There are two basic types of losses, constant and variable, both of which develop heat. Core Losses (Constant): A combination of eddy-current and hysterisis losses within the stator core. Accounts for 15 to 25% of the overall losses. Eddy-Current Losses: Heating that occurs in steel when an alternating current is passed close to it, causing a transformer action. The transformed currents generate heat. Reduced in electric motors by stacking steel plates ranging from about 0.019 inches to 0.049 inches thick. Hysterisis Losses: Caused by the opposition to the changing magnetic field within the makeup of the core steels themselves. As the current changes direction, the molecules within the steel must realign, which generates heat. Reduced in electric motors by using low-hysterisis and silicone steels in manufacturing laminations. Friction and Windage Losses (Constant): Mechanical losses which occur due to air movement over rotor surfaces and bearings. Accounts for 5 to 15% of the overall losses. Stator Losses (Variable): The I2R (resistive) losses within the stator windings. Accounts for 25 to 40% of the overall losses. Rotor Losses (Variable): The I2R losses within the rotor windings. Accounts for 15 to 25% of the overall losses. Stray Load Losses (Variable): All other losses, including leakage (capacitance), not otherwise accounted for. 10 to 20% of the overall losses. These losses impact the overall efficiency of the electric motor. Equation 12: Efficiency Equation WattsIn − WattsLosses %Eff = × 100 WattsIn

Wound Rotor Motors Wound rotor motors are a special design of three-phase induction motors that allow a variable resistance within the rotor. The stator windings are similar to those of a standard induction motor. The rotor, however, is made up of a series of windings connected to a set of three slip rings which are connected to a special control and resistor bank.

Penrose_BMI.indd 24

4/22/08 5:41:44 AM

Electric Motor Theory

25

The resistor bank and control are used to vary the locked rotor, pullup and full-load torque of the motor and can be used for soft-starting and slip-based speed control. These motors are primarily used in cranes and pumps. However, they are limited in accurate speed control.

Synchronous Motors An industrial synchronous motor consists of a three-phase stator winding, a squirrel-cage rotor windings, and a series of rotor fields (field-poles) designed to carry DC power. The stator fields generate a synchronous magnetic field which interacts with the rotor windings in the same way as an induction motor. As the rotor begins to turn, a direct current is introduced into the rotor fields either through a generator attached to the rotor shaft or through a static control. The DC magnetic fields lock with the stator fields causing the rotor to turn at synchronous speed.The rotor windings then act to counter any sudden load changes so that the motor continues to operate at synchronous speed. Industrial synchronous motors are often found in reciprocating compressors. They can also be used to correct power factor by overexciting the rotor fields. Another type of synchronous motor involves the use of permanent magnets. The stator fields interact with the rotor magnets. These motors tend to be smaller in size and used in machining applications as stepper or servo motors.

Direct Current Electric Motors Direct Current electric motors operate under a basic principle of electricity: interaction between two magnetic fields positioned at an angle from each other will attract/repel resulting in movement. In the case of a DC electric motor, power is provided to a stator field and an armature creating magnetic fields that are, electrically, about 90 degrees from each other. The resulting attraction/repulsion of the armature from the field generates a torque and the armature turns. The basic components of a DC electric motor include: Frame – Makes up the outer structure of the machine. It is used to mount most of the other components of the motor. Fields – Are coils mounted on field pole pieces that generate a stationary magnetic field.

Penrose_BMI.indd 25

4/22/08 5:41:45 AM

26

Electrical Motor Diagnostics: 2nd Edition

Interpoles – Are coils that are placed between the field coils that generate a field that is used to prevent excessive sparking of the brushes. Endshields – Also called bearing housings, are used to house the brushes, brush rigging, and to house the shaft bearings, holding the armature centered in the frame. Brush rigging – Holds and positions the brushes above the armature commutator. Usually, a tension device is used to maintain a constant pressure on the brushes. Brushes – Are used to provide DC to the armature. The brushes ride on the commutator. Commutator – Consists of many copper bars that are separated by mica. Each bar is connected to coils in the armature. Armature – Is the rotating portion of the motor that contains coils. Unlike most AC motors, DC motors require separate power to be provided to both the fields and the armature.The DC provided to the stator fields generate a constant North and South set of fields. DC provided to the armature generates North and South fields that are 90 electrical degrees from the stationary field. As the armature generates a torque and moves towards the appropriate North or South pole, the brushes change position on the commutator, energizing another set of coils 90 electrical degrees from the stationary field. This actually makes the armature an Alternating Current component as the current will travel in one direction, based upon brush position, then in another direction as the motor operates. The brushes are set in such a position that they are electrically “neutral” (no induced current from the stator fields) in order to reduce sparking. In most DC motor connections, by varying the armature voltage, the operating speed may be changed. One general danger that is inherent in DC motors is that if field current is lost while armature current is maintained, the motor may take off and the speed increase until the armature self-destructs. The three basic winding types that can be used to identify the type of DC motor include: Series: Normally found in applications that need a high starting torque.They consist of a set of field windings of large wire and relatively few turns, marked S1 and S2, that are connected in series to the interpoles and armature, marked A1 and A2 (See Figure 14). Series connected motors are normally used as traction motors and have a very low basic resistance.

Penrose_BMI.indd 26

4/22/08 5:41:45 AM

Electric Motor Theory

27

Figure 14: Series Motor

Shunt: Normally found in applications that require constant speed. They consist of a set of field windings of smaller wire with many turns, marked F1 and F2 for single voltage and F1, F2, F3 and F4 for dual voltage, and A1 and A2 for the interpoles and armature (See Figure 15). Shunt connected motors are normally used as crane and machine tool motors and have a relatively high basic resistance.

Figure 15: Shunt Wound Motor

Compound: Combine the benefits of both the series and shunt wound motors. They provide a relatively high torque with a basic resistance to a change in operating speed. The connections combine both the series and shunt connections (See Figure 13). Compound motors are the most common and are commonly found in industrial manufacturing.

Penrose_BMI.indd 27

4/22/08 5:41:45 AM

28

Electrical Motor Diagnostics: 2nd Edition

Figure 16: Compound Wound Motor

Machine Tool Motors Machine tool motors can be described as including small pump motors, servo motors, and spindle motors. The description can vary depending on the type of machine tool being used. A simple lathe may have a spindle motor and cutting fluid pump motor while more complex machines will have spindle motors, cutting fluid pumps and axis motors. The purpose of a machine tool is to take an object, place it on a table or into a chuck then remove material with a cutting tool until the final product is completed. Spindle: Can be a special spindle motor, DC motor or AC motor, used to turn a piece (such as on a lathe) or the cutting tool (such as on a milling machine). The spindle may be operated through a DC or variable frequency drive in order to change spindle speeds, or a mechanical method can be used for changing speeds. The spindle motor may have feedback devices and overloads built into the motor and connected through amphenol plugs. Axis Motors: Will have position feedback devices, such as encoders, for positioning and speed feedback. The motors may be DC, AC, Synchronous or ‘brushless DC’. Some Axis motors are considered servo and others stepper motors, depending on the type of application. Machine tools may have as few as one axis to many multiple axis’. Cutting Fluid Pumps: Are normally standard induction motors.

Penrose_BMI.indd 28

4/22/08 5:41:46 AM

Electric Motor Theory

29

Traction Motors Traction motors are used in trains, hybrid motor vehicles, trams and other transport vehicles. They are designed to produce very high torque. DC Traction motors are basically large series connected DC motors with very low resistance windings. AC Traction motors are similar to Design D motors and will have a very low stator and rotor winding simple resistance and impedance. Both AC and DC Traction motors will require a high current capability.

Transformers To understand the basic concepts of a transformer, we shall start with an “ideal transformer,” or a theoretical transformer that has no losses. The purpose of the transformer is to convert one level of voltage and current to another level of voltage and current for distribution and application purposes. This is achieved by having a primary winding located close to secondary winding and allowing for mutual induction to occur between the windings. When a sine-wave voltage is applied to the primary windings a magnetic field is established that expands and contracts based upon the applied frequency. This field interacts with the secondary winding, producing a voltage within the secondary that is directly proportional to the turns ratio, while current is inversely proportional to the turns ratio. Equation 13: Voltage Turns Ratio N1 / N 2 = a Where N1 is the number of turns in the primary and N2 is the number of turns in the secondary Equation 14: Current Turns Ratio N 2 / N1 = 1 / a For example, an ideal transformer with 100 turns in the primary and 50 turns in the secondary, with 480 Volts applied to the primary and a 100 amp load on the secondary would have: a voltage turn ratio of 2;

Penrose_BMI.indd 29

4/22/08 5:41:46 AM

30

Electrical Motor Diagnostics: 2nd Edition

a current turn ratio of 1/2; a 480 V1, 50 A1 load reflected on the primary and a 240 V2, 100 A2 load on the secondary. Equation 15: Load Impedance Z L = V2 / I 2 Equation 16: Equivalent Primary Impedance Z 1L = a 2Z L Equations 14 and 15 can be used to reference the impedance from the secondary to primary. This can also be used inversely. Internal impedance can be matched to load impedance as found in Equation 16. Equation 17: Internal Impedance Z S = a 2Z L = Z 1L In a “real transformer” there are certain losses, including core losses (hysterisis and eddy-currents), the magnetizing current, and leakage. In addition, supply voltage and load currents may have harmonic loads and other issues that would impact the effectiveness of a transformer. The purpose of static MCA is to reduce or eliminate these issues, and to isolate the transformer for testing. Transformers of both single and three phase have a variety of connection types for a variety of loads. In a three-phase circuit, these connections are: Wye-Delta, Delta-Wye, Delta-Delta, and Wye-Wye. Single-phase, pole mounted transformers normally have a single-winding primary with a two-winding or center-tapped secondary.

Penrose_BMI.indd 30

4/22/08 5:41:46 AM

Electric Motor Theory

31

Figure 17: Power Transformer

Three phase transformer connections are developed for a variety of applications, including lighting (less than 230 Volts) and power (motors, etc.) loads: Delta-Delta: Lighting and power applications, normally used when power loads are greater than lighting loads.

Figure 18: Delta-Delta Transformer

Penrose_BMI.indd 31

4/22/08 5:41:47 AM

32

Electrical Motor Diagnostics: 2nd Edition

Note:Three phase transformer connections are labeled H1, H2, and H3 on the primary and X1, X2, X3, with X0 as the neutral, on the secondary. Delta-Wye: Normally provides a 4-wire on the secondary which allows for balanced single-phase loads between neutral and each phase.

Figure 19: Delta-Wye Transformer

Open-Delta: Lighting and power applications, used when lighting loads are greater than power loads. Open Wye-Delta: Will allow 57% capacity if one phase is disabled.

Figure 20: Wye-Delta Transformer

Penrose_BMI.indd 32

4/22/08 5:41:48 AM

Electric Motor Theory

33

Wye-Delta: Lighting and power applications.

Figure 21: Wye-Wye Transformer

Wye-Wye: Power applications, used when stepping power up in voltage (ie: 2400 to 4160 Volts).

Figure 22: Delta-Wye Transformer Connection

Penrose_BMI.indd 33

4/22/08 5:41:49 AM

34

Electrical Motor Diagnostics: 2nd Edition

Figure 23: Delta-Delta Transformer Connection

Note: Single-phase pole mounted transformers are often connected and labeled H1 and H2 on the primary and X1, X2 (center tap), and X3.

Figure 24: Single-Phase Transformer Connection

Penrose_BMI.indd 34

4/22/08 5:41:50 AM

Electric Motor Theory

35

Chapter 3 Questions 1. What is the impact on torque with a Design B motor if the rotor resistance is increased? 2. What are the losses associated with AC induction motors? Define. 3. What is the primary difference between a synchronous motor and an AC induction motor? 4. What is the benefit of an open wye-delta transformer and what types of applications are they used for? 5. What are the three types of DC motors and what are they characteristics of each? 6. In an ideal transformer, if the primary voltage is 12,400 Volts and the current seen on the primary is 400 Amps, what is the secondary voltage and current if the primary has 500 turns and the secondary has 20 turns?

Penrose_BMI.indd 35

4/22/08 5:41:50 AM

Penrose_BMI.indd 36

4/22/08 5:41:50 AM

Chapter

4 Electrical Insulation Systems and Theory

Motor Circuit Analysis (MCA) techniques utilizing variations Resistance (R), Impedance (Z), Inductance (L), Phase Angle (Fi), current/ frequency response (I/F), capacitance (C) and insulation resistance have been in practice since 1985. The technique has been successfully applied for the detection of winding defects (shorts, resistive unbalances and insulation to ground), cable defects and rotor defects. It has also been found to be able to trend and estimate winding failures with a high degree of accuracy.3,4 In this chapter, we will discuss the physical properties that allow for the detection of these motor electrical circuit faults using MCA. We will then support the concepts by presenting a series of experiments performed using MCA techniques and comparing to existing technologies.

The Motor Circuit The three phases of a three phase induction electric motor are separated by 120° electrical.The supply voltage phases are also, optimally, separated 3

Penrose, Howard W Ph.D., “Estimating Motor Life Using Motor Circuit Analysis Predictive Measurements Part 1,” IEEE Electrical Insulation Conference Proceedings, 2003.

4

Penrose, Howard W Ph.D., “Estimating Motor Life Using Motor Circuit Analysis Predictive Measurements Part 2,” IEEE ISEI Proceedings, 2004. 37

Penrose_BMI.indd 37

4/22/08 5:41:50 AM

38

Electrical Motor Diagnostics: 2nd Edition

Figure 25: Single Phase Equivalent Circuit

by 120° electrical. Within each phase, as the voltage increases, current increases due to the impedance of the motor circuit (See Figure 25). As the current increases, the two magnetic poles increase (or sets of poles), then decrease as current decreases.The stator back iron acts to strengthen and direct the magnetic fields within the air gap between the stator and rotor. As the fields pass through the rotor bars (conductors) of the rotor, a second current develops in the rotor which interacts with the rotating fields in the air gap. The rotor follows the rotating fields, although lags behind the synchronous speed of the stator (slip) in order to maintain a rotor current, and resulting rotor magnetic fields. As this is occurring, changes also occur to the insulation system and back iron steel. As the current increases in each phase:  There is a skin effect within the copper conductors that forces more current towards the surface of the conductor.  Insulation dipoles polarize between conductors as the phase voltage and current increases and decreases, causing constantly changing capacitance within the circuit between conductors.  Insulation dipoles polarize between conductors and ground as the phase voltage and current increases and decreases, causing constantly changing capacitance between the winding circuit and ground.  Magnetic dipoles polarize in the area of effect of each pole within the stator core steel. The reluctance to realign, with a change in magnetic field, is termed as hysterisis. Operating voltages force the changes to occur fairly rapidly. Changes to the circuit, or to the dielectric or magnetic properties of the motor effect

Penrose_BMI.indd 38

4/22/08 5:41:50 AM

Electrical Insulation Systems and Theory

39

its operation and the force of the operating voltage causes the defective areas of the insulation or steel to heat. Continued breakdown of the dielectric occurs based upon the severity of the fault.

Insulation and Magnetic Field Effects The electrical insulation circuit is modeled as a series of parallel RC circuits between conductors and conductors and ground. As changes occur to the insulation system, the values of R and C change. The values of the insulation in each phase are the sum of the turn to turn and coil to coil RC values of each phase. Insulation to ground values are the sum of the insulation between conductors and ground for the complete circuit. The capacitance of the electrical insulation is a direct function of the generation of dipoles within the insulation system. As a field is generated across an atom, or molecule, of a dielectric, it will polarize, meaning that the electron orbit of an atom will shift slightly, making one side of the atom more positive and one more negative. As current passes through conductors near electrical insulation, the insulation reacts by polarizing the atoms (dipoles) within the insulation. As the dipoles polarize, there is less leakage (capacitance) between

Figure 26: Insulation Model of Motor Winding System

Penrose_BMI.indd 39

4/22/08 5:41:51 AM

40

Electrical Motor Diagnostics: 2nd Edition

Figure 27: Generation of a Dipole

Figure 28: Dipolar Effect of Insulation

the conductors and ground. This also occurs in the insulation system between conductors when there is a difference in potential. In a good insulation system, the polarization of the insulation system occurs in a larger number of atoms. Once the potential is removed, the atoms return to their original state (dipoles randomize).

Penrose_BMI.indd 40

4/22/08 5:41:51 AM

Electrical Insulation Systems and Theory

41

Figure 29: Magnetic Dipoles

The same effect occurs in a magnetic field. The magnetic dipoles of the backiron and teeth of the stator core line up in the direction of the magnetic field. This helps direct the magnetic flux and adds to the strength of the fields within the airgap. The reluctance of the steel to change polarity shows up as hysterisis losses from the field. Once the field is removed, the magnetic dipoles of the steel quickly randomize. The above descriptions for the polarization of electrical insulation and core steel represent the steady-state application of an applied voltage potential. In an operating three phase system, the effects get far more exciting. As each sinusoidal phase of voltage is impressed across the windings:  As the voltage starts from zero, the beginning of the coil energizes, the insulating dipoles between the insulation to ground and the conductors within the coil are forced to polarize.  As the voltage continues to rise, the potential at the beginning of the coil is higher than the end of the coil, insulating dipoles continue to polarize and the magnetic dipoles begin to polarize in the direction of the magnetic flux generated by the coils.  As the voltage hits its peak at the beginning of the coil, a majority of the magnetic and insulating dipoles associated with the start of

Penrose_BMI.indd 41

4/22/08 5:41:52 AM

42

Electrical Motor Diagnostics: 2nd Edition

the coil have polarized and the ones at the end of the coil continue to polarize. There is a lag in the fields between the beginning and the end of the coil, which causes a potential between conductors to exist.  As the voltage begins to decrease, the insulating and magnetic dipoles begin to randomize (move to neutral) at the beginning of the coil and release energy back into the system as the fields collapse. The fields at the end of the coil hit their peak then start to decrease.  The voltage approaches zero, then passes into the negative sequence of the sine wave.The dipoles and fields continue to react, but align in the opposite direction (as in a piston action). We will define this action as ‘dipolar spin’ of both the electrical insulation and magnetic steel dipoles. The high potential of most electric motors force the changes to the fields and dipoles to happen quickly. As a result, work is performed and heat is generated. The Capacitance of each portion of the circuit is given, at any time, as: Equation 18: Capacitance of Each Portion of Circuit C=

Q ε °S Q −q l

Where an insulator exists between the conductors and conductors and ground.The induced charge, q, increases the capacitance by the ratio Q/(Q-q). The dimensionless ratio q/(Q-q) is a property of the polarizability of the material and is referred to as the electric susceptibility, Xe.5 At the boundary of each insulation system (conductors, slot, phase, etc.), the boundary conditions are such that: Equation 19: Insulation Boundary tan θ 2 = ε r tan θ1 Where εr represents the relative permittivity of the boundary of the insulation surface.

5

P. Hammond and J.K. Sykulski, Engineering Electromagnetism: Physical Processes and Computation, Oxford Science Publications, 1994.

Penrose_BMI.indd 42

4/22/08 5:41:52 AM

Electrical Insulation Systems and Theory

43

By dividing each phase into tubes and slices,6 the total capacitance for m slices and n tubes through the system would be: Equation 20: Phase Capacitance −1

δl C = ∑ (∑ ) εδ S 1 1 n

m

The inductance of the circuit can be figured as the flux linkage per unit of current, and is represented by the unit Henry (H): Equation 21: Circuit Inductance Nφ i For a motor with n coils, the inductance may be defined: L=

Equation 22: Inductance with ‘N’ Coils N p ( K pqφq ) L pq = iq Where Kpq is referred to as the coupling coefficient between two coils (p and q).When p and q are equal, the inductance is termed as selfinductance, when unequal, it is termed mutual inductance.7 The total impedance per phase as viewed from the stator input terminals is given as shown in Equation 23 where X refers to the leakage reactance (capacitive).8 Equation 23: Per Phase Impedance R’ jX M ( 2 + jX ’l 2 ) s Zt = R1 + jX l1 + R ’2 + j( X M + X ’l 2 ) s

6

The purpose of the tubes and slices approach, as introduced by Hammond and Sykulski, is to provide a manageable means to look at variances through a system. This is done by taking the system in small chunks referred to as tubes and slices.

7

Nasar, Electric Machines and Electromechanics, Schaums Outlines, 1981.

8

Mulukutla Sarma, Electric Machines Second Edition, PWS Publishing Company, 1996.

Penrose_BMI.indd 43

4/22/08 5:41:53 AM

44

Electrical Motor Diagnostics: 2nd Edition

In simpler form, impedance can also be viewed as: Equation 24: Inductive Reactance X L = 2π fL = Inductive Reactance Equation 25: Capacitive Reactance Xc =

1 = Capacitive Reactance 2π fC

Equation 26: Simple Impedance (Complex AC Resistance) Z=

R 2 + ( X L − XC )2

When looking at a balanced system, a wye circuit should appear as in Figure 30. The circuit impedance would appear: Equation 27: Simplified Circuit Impedance Example V Z AB = AB I AB Vab Ia

Iab

Iac

45˚ 30˚ 30˚

Ic 45˚

Vac

45˚

Ica

Ibc

Vca

Ib

Figure 30: Balanced Wye System

Penrose_BMI.indd 44

4/22/08 5:41:53 AM

Electrical Insulation Systems and Theory

45

For example: 32.9∠45°Ω =

650.5∠120°V 19.8∠75° A

Armed with this information, we can now review the effects of winding related faults on the operation of the motor.

Winding Faults When a defect occurs in a winding due to a developing short, winding contamination or severely damaged core steel, it effects the electrical properties of the insulation system. In the case of a winding defect, changes to either capacitance or resistance within the insulation system will cause a reactive problem due to changes to the makeup of the insulation system. For instance, in a developing short, the changes to the insulation system cause changes to the capacitance due to changes in how the dipoles are excited (dipole spin). As a result, there are changes to how the insulation reacts in that area, causing a leakage reactance variance and heating due to forcing the insulation to polarize with high applied potential (operating voltage). Winding contamination causes changes to the resistive and capacitive reactance between insulating surfaces, as well. At design voltage, most defects do not become apparent until a distinct change occurs, which may be represented by a severe current unbalance, nuisance tripping or a direct short circuit. In the case of winding contamination, the end result is the same as a winding short: Either a short between conductors or across the insulation system to ground. As a result, as faults occur due to thermal deterioration, contamination, moisture absorption or other reactive faults, the circuit impedance will change, slightly, at first, then more dramatic as the fault progresses.

Traditional Test Methods Most of the traditional test methods require a significant voltage application in order to work. The purpose is to stress the insulation system by forcing a reaction of the insulation dipoles, or ionization of air and insulation medium defects in order to force a potential across a resistive or capacitive fault. In this section, we will review a few of the test methods in brief, including: Insulation to ground testing; Polarization Index; Resistance Testing; and, Surge Comparison Testing.

Penrose_BMI.indd 45

4/22/08 5:41:54 AM

46

Electrical Motor Diagnostics: 2nd Edition

Insulation to Ground Testing (Meg-Ohm Meters) As described in Figure 29, a DC potential is placed across the motor winding conductors and ground. The applied potential is set and a value of current (leakage) crosses the insulation boundary. This value is converted to resistance, usually in meg-ohms. It is, in effect, a method of measuring leakage across the insulation boundary, but only between the surfaces of the conductors and ground. As the insulation dipoles are only excited with DC, some time is required for them to polarize. Standards normally indicate a winding charging time of one minute and, as insulation resistance is directly affected by temperature and moisture, normalization for temperature.

Polarization Index The Polarization Index (PI) test is a measurement of leakage at one minute then at ten minutes.The results are shown as a ratio of the ten minute to one minute readings. It is assumed that a fault will polarize slowly (high ratio) or rapidly (low ratio) due to contamination and changes to the ground wall circuit capacitance.

Resistance Testing Resistance tests use a low voltage DC output and bridge. The primary purpose is to detect high resistance joints, loose connections, broken connections and direct shorts.

Surge Comparison Testing An older method of evaluating windings for shorts. A series of steepfronted higher voltage pulses are sent from the instrument to the stator. The higher voltages occur too fast to properly polarize the insulation system, instead relying upon higher voltage to ionize gasses leaving the ability to detect a reactive fault as creating enough potential to cross the barrier (Paschen’s Law) with the test ending prior to twice the nameplate voltage plus 1000 Volts or once an arc is drawn. This method of testing causes a change to the properties of the insulation at the point of defect either accelerating the fault or completing the fault. In order to force slight defects, a greater potential must be applied in order to stress the complete system. Due to the steep fronted surges, however, the applied voltage is normally applied on the first 2–3 turns in the first coil of each phase.

Penrose_BMI.indd 46

4/22/08 5:41:54 AM

Electrical Insulation Systems and Theory

47

“The situation is quite different for detecting the breakdown of the turn insulation in a winding (parallel or phase) having many coils. The breakdown of the turn insulation in a single coil in a winding of many coils produces a very small relative change in the characteristics (L, C, R) of total load impedance seen by the surge generator. Hence, the change in the VWF [voltage wave form] shape produced by the breakdown of the turn insulation somewhere in a winding of many coils is relatively very small. Hence the surge tests may not reliably verify the presence of one shorted turn in a single phase and may probably lead to wrong conclusions. Perfectly intact windings may appear to have a turn short. More importantly, a turn short induced by the surge test by breaking down the weakened turn insulation may not be detected. In such a case, the stator winding would likely fail after the machine is put back into service. “In view of the above facts, caution is advised in surge testing of the turn insulation in complete windings. These tests carry very significant risks, which should be carefully considered. Such caution is more important for diagnostic tests on machines in service as such tests are carried out quite infrequently in contrast to frequent tests on new, or refurbished, or repaired machines in a manufacturer’s plant.”9 As shown, traditional testing has specific flaws in the ability to detect faults, and the ability to detect these faults in a non-destructive manner.

Modern Low-Voltage Testing: Motor Circuit Analysis Modern MCA devices use a low voltage sine-wave output designed to excite the insulation system dipoles and surrounding magnetic steel dipoles with low current. There are several key benefits to this approach: Size and voltage rating of the machine being tested do not matter; Specific pass/fail criteria can be applied to phase comparison; and, Degradation can be trended over time without any adverse effects to the existing condition. “Based upon the physical and electrical properties of coil windings, insulation, systems, transformer theory and electric motor theory, a set of electronic measurements can provide the necessary information to determine the condition of electrical equipment. The measurements must include circuit DC resistance, circuit inductance, circuit impedance, 9

Bal Gupta, “Risk in Surge Testing of Turn Insulation in Windings of Rotating Machines,” IEEE Electrical Insulation Conference Proceedings, 2003.

Penrose_BMI.indd 47

4/22/08 5:41:54 AM

48

Electrical Motor Diagnostics: 2nd Edition

phase angle, current/frequency response and insulation resistance readings. Resistance readings are used for open or poor connections, inductance and impedance are used to evaluate winding condition in electric motors and phase balance in all other applications, phase angle and current/ frequency response tests evaluate windings for shorts and insulation resistance readings are used to detect winding to ground shorts.”10

Detection of Winding Contamination Due to the fact that one of the last measurements to change due to a turn short fault is inductance (L), a test result of L can be used as a comparative baseline. This is important as the relative position of the rotor in an assembled machine will effect the reading due to mutual inductance. Equation 28: Mutual Inductance m k N a 2 = 1 ( w1 1 )2 m2 kw1N 2 Where 1 represents the stator winding factors and phases and 2 represents the rotor bar factors and bars per phase.The result is a ratio, the same as a transformer ratio.When a rotor is stationary in an electric motor, the ratios are different for each phase. Winding contamination causes small changes to the capacitance of the winding circuit. In most cases, the capacitance increases within the circuit. When referencing the simple impedance formula, earlier in this chapter, it identifies that an increase in capacitance will have a negative impact on impedance. Also, as the applied voltage is very low, capacitive reactance has a more significant impact on the impedance (Ohm’s Law) as the capacitive value is more dominant. The result, using a relatively low frequency and sinusoidal output, is a collapse of impedance towards inductance in the phase which has capacitive effects from the contamination or water absorption. In cases of high humidity, the insulation has to have fissures or defects in order to cause the change.

Overheated Windings Overheated windings have a similar impact as winding contamination. The difference is that the insulation is thermally degrading causing 10

Dave Humphrey, Allison Transmission, “Which Road Will You Take?,” IEEE Electrical Insulation Conference Proceedings, 2003.

Penrose_BMI.indd 48

4/22/08 5:41:54 AM

Electrical Insulation Systems and Theory

49

increased resistance to dipolar action. In this case, the capacitance may decrease, causing a decrease in impedance in one, or more, phases. In both winding contamination and overheated windings, the end result would be a winding short. Winding contamination can be corrected if detected in its early stages. However, once changes occur that allow for the detection of a winding fault, the winding will have to be replaced.

Winding Shorts One of the keys to proper MCA testing is that inductance is not used as a primary method of detection for developing shorts. Instead, two specific measurements are used in combination to determine the type and severity of the defect. These measurements are: The circuit phase angle; and, A Current/Frequency response method. When a defect occurs in the winding, it changes the effective capacitance of the complete circuit. The change in capacitance will directly effect how the low level current lags behind voltage with the usual result being an increase in capacitance and a reduction of the phase angle in the effected phase. Once the fault becomes more severe, it will begin to effect the surrounding phases. This normally occurs when the defect exists in one coil or between coils in the same phase. A very small change to capacitance within the circuit can be detected, allowing the detection of single turn faults and pinhole shorts when using very low frequencies. A second method of fault detection uses a current ratio, similar in method to the frequency response method used for transformer testing. However, the low voltage current is measured, then the frequency is exactly doubled and a percentage reduction in the low-level current is produced. When the frequency is doubled, small changes to capacitance between turns or between phases are amplified, causing a change to the percentage reduction when compared between phases. The combination of phase angle and current/frequency response allow for the detection of winding shorts and the type of short being detected in any size machine. Also, due to the use of low voltage and the result that only a small change to circuit capacitance is required to detect the faults, early winding defects can be detected quickly and trended to failure.

Additional Tests In combination with the above tests, MCA utilizes resistance readings and insulation to ground tests. This allows the technology to detect

Penrose_BMI.indd 49

4/22/08 5:41:55 AM

50

Electrical Motor Diagnostics: 2nd Edition

approximately half of the potential faults in the overall motor system and allows for the comparison of any two sets of insulated coils. Faults and defects can be detected in cables, coils, transformers, motors and rotor defects.

Rotor Testing and Back Iron Effects on MCA The effect of being able to evaluate the condition of the motor rotor is “based upon Faraday’s law of electromagnetic induction, according to which a time-varying flux linking a coil induces an emf (voltage) in it.”11 Equation 29: Primary emf e1 = ωN 1φm cos ωt for the primary emf Equation 30: Secondary emf e 2 = ωN 2φm cos ωt for the induced secondary emf Equation 31: Turns Ratio e1 N 1 = for the turns ratio e2 N 2 Equation 32: Ratio of Primary and Secondary Impedance Z1 N = ( 1 )2 = a 2 Z2 N2 Which is the ratio of the primary and secondary impedances of the circuit. The motor circuit analyzer excites the core steel based upon the amount of current available to the circuit and reacts across the airgap: [15]

Equation 33: Rotor Influence Blg ap Bl nI = iron + uγ uο uο 11

Syed A Nasar, Electric Machines and Electromagnetics, Schaum’s Outline Series, 1981.

Penrose_BMI.indd 50

4/22/08 5:41:55 AM

Electrical Insulation Systems and Theory

51

The direct relationship to the ability to detect the rotor across the airgap depends upon the distance across the airgap, the area of the steel magnetized and the length of the stator core. In longer cores, the effect will carry across the airgap and excite the rotor core and induce the instrument frequency into the rotor circuit. In very short cores, the fringing effect the magnetic field from the stator has a similar effect. In large machines, the amount of energy available from an MCA device allows for the detection of rotor defects only above the area immediately surrounding each coil side. This produces multiple effects: A. The mutual inductance changes as the rotor position changes as a direct result of the change to the transformer ratio between the primary (stator) and secondary (rotor). A good rotor will show as a repeating pattern, a bad rotor will change the transformer ratio and a defect will appear as a non-repeating pattern. B. Fractures will be readily detected as the induced energy is relatively low and the oxides on the surface of the defect will change the transformer ratio. Whereas, in higher voltage rotor tests, the energy may be significant enough to pass through the defect. C. In rare instances, the airgap may be too significant and very little to no variation of the mutual inductance occurs. In this case, larger defects, such as multiple fractures or a broken bar, will show as a variation in a straight line. D. MCA technology has the ability to detect wound rotor, synchronous rotor field and other wound-rotor defects across the airgap. Because of the impedance ratio between the primary and secondary, rotor winding defects will show as a change to phase angle and current/frequency response and will vary based upon rotor position.

Armature and Commutator Contamination Detection One of the unique abilities of MCA is the ability to detect carbon buildup in DC motor armatures. Due to the dielectric (capacitive) properties of carbon, capacitance values of the circuit become unstable. This causes test results of impedance, phase angle, current/frequency response and insulation to ground to become unstable and non-repeatable. As a result, armature circuit contamination is detected by noting non-repeatable test results. This is important in that, if detected early, this type of defect may be corrected by blowing out the armature with low pressure air.

Penrose_BMI.indd 51

4/22/08 5:41:55 AM

52

Electrical Motor Diagnostics: 2nd Edition

Summary of MCA Theory Based upon the engineering principles of motor and transformer design, utilizing low voltage testing technologies allow for the detection of incipient defects in the electric motor circuit including cable insulation, coils, transformers, connections, motor and rotor windings, armatures, air gap issues and squirrel cage rotor defects, covering over 50% of all potential motor system faults of any size or voltage machine through the motor circuit and cabling or directly at the machine. This is achieved by utilizing a low potential sinusoidal output from the instrument which excites the insulation and magnetic dipoles of the circuit. The low potential allows defects to become more readily apparent at early stages as it does not force, but excites, dipolar spin, causing changes to the circuit impedance, phase angle and current at varying frequencies (current/frequency response), depending on the type of fault. These properties of the technology allow for long term trending of developing defects from insulation breakdown and contamination without any harmful testing effects due to insulation breakdown nor winding contamination.

Development of Experiments In the first part of this chapter, a number of fault detection capabilities and the method for how faults are detected, using MCA methodology were outlined. A series of controlled experiments were developed, based upon this chapter and previously developed tolerances, using MCA and surge comparison testing. The experiments were developed to provide the following: 1. Experiment 1: A used 20 horsepower, 3600 RPM, 460 Volt stator only from a BJM Pump, model KZN 150 (submersible, BJM Corp.) was selected from the BJM Corp. repair center discard area. A visible winding fault, caused from what appeared to be mechanical damage, was identified. The original cause of failure for the pump was seal failure, resulting in bearing failure with the bearing causing damage to the winding. The winding was still operational when the pump was sent for repair and was pulled due to detection of the fault by MCA, then visual detection: 1.1 Use MCA to detect fault using standard fault detection rules;

Penrose_BMI.indd 52

4/22/08 5:41:56 AM

Electrical Insulation Systems and Theory

53

1.2 Apply surge comparison voltage value until the visible fault was detected and note voltage value. Determine if the fault was detected in the same phase; 1.3 Note effect of both test methods, during fault detection. 2. Experiment 2: Performed in three parts: 2.1 A 3/4 horsepower, 3-phase, 1750 RPM, 460 Volt, used, complete Dayton motor was selected.The motor was disassembled and inspected, then reassembled.Testing was performed before and after using both an ALL-TEST IV PRO 2000 (AT4) and ALL-TEST PRO 31 (AT31). Surge comparison testing was performed to 1800 Volts for 3–5 seconds. The purpose was to detect and determine field observed changes to impedance, inductance, and other test results following surge comparison testing in the field. 2.2 A Baldor, 1 horsepower, 1725 RPM, 460 Volt, new, assembled motor with a rotor fault generated, involving the removal of two rotor bars from the rotor circuit.The purpose was to detect the condition of the winding and determine the capability of detecting rotor bar faults, and the effect of rotor position on test methods using the AT4, AT31 and surge comparison tester. Confirmation of winding condition was determined by applying 12,000 Volts using the surge comparison tester. 2.3 A 1 horsepower SPV style, 3600 RPM, 460Volt, BJM Submersible Pump stator with grey water contamination was selected from the repair discard pile.A coil-to-coil fault within the same phase, between individual turns, was observed during selection. The stator was selected due to a low (0.98 MegOhms) insulation test value, representing severe winding contamination conditions. The stator was to be evaluated using the AT4 and AT31 for purposes of contamination detection, surge tested to 1800 Volts, then re-tested using MCA. The purpose was to determine the ability of MCA to detect the faults and impact of surge comparison testing on winding contamination. 3. Experiment 3: Insert rotor in the stator from Experiment 1 in order to determine the impact of the rotor on the original test results and to identify severity of the faults. Surge comparison testing was used to degrade the fault condition and the results monitored

Penrose_BMI.indd 53

4/22/08 5:41:56 AM

54

Electrical Motor Diagnostics: 2nd Edition

using MCA. The dielectric strength was trended following each test. The differences between a running equipment fault and progressive degradation from the surge tester used are to be noted. The MCA devices were factory calibrated, as was the surge comparison tester, prior to test evaluation. Testing was performed over three separate days with variations to temperature and humidity (dew point) in which the differences were noted between Experiment 1 and Experiment 3. Through each experiment, the theories presented in the first part of this chapter were supported and winding tolerances confirmed. Testing and results were recorded within instruments and via a video log of critical test results. Variances in test results between the AT4 and AT31 involve the use of different test frequencies ranging between 25 Hz and 800 Hz.

Test Tolerances MCA Tolerances MCA Test tolerances for the technologies presented in this chapter are well established: Table 4: Assembled Motor Tolerances Test Resistance Z and L Fi and I/F Insulation Resistance

Tolerance < 5% Similar Pattern in Both Results (phase comparison) +/−1 Digit from Average Result (phase comparison) > 5MegOhm for 100 Mohm for >600V motor Table 5: Disassembled Motor Tolerances

Test Resistance Z and L Fi and I/F Insulation Resistance

Penrose_BMI.indd 54

Tolerance < 5% < 3% and 5MegOhm for 100 Mohm for >600V motor

4/22/08 5:41:56 AM

Electrical Insulation Systems and Theory

55

Surge Testing Tolerances Requires a visual tolerance between phases. Separation of waveforms indicate a fault in the windings, such as a short.Theoretically, surge testing will provide a ‘shift’ in the right side of the waveform prior to arcing.

Experiment Test Results Experiment 1 The 20 horsepower stator was selected and placed on a wooden crate to ensure no interference from other surrounding materials. Testing was performed using MCA and following the rules for motors tested without rotors, as shown in Table 5. Table 6: Pre-Test Results with MCA

Resistance Impedance Inductance Phase Angle I/F Ins Resistance

1–2

1–3

2–3

0.7209 41 8 83 −49

0.7207 40 8 83 –50 >99

0.7206 40 8 83 −49

Surge testing was performed. All three phases were, initially, limited to testing to 2,000 Volts (Figures 31, 32 and 33).

Figure 31: Surge Results T1-T2

Penrose_BMI.indd 55

4/22/08 5:41:56 AM

56

Electrical Motor Diagnostics: 2nd Edition

Figure 32: Surge Results T1-T3

Figure 33: Surge Results T2-T3

The MCA test identified the phase to phase fault in test leads T1–T3. To ensure detection across T1–T3, the stator was surge tested to failure (Figure 34). The fault was detected on leads T1–T3 at 6,000 Volts. In confirming tests, it was noted that the voltage decreased to 4,500 Volts. No shift, or

Penrose_BMI.indd 56

4/22/08 5:41:57 AM

Electrical Insulation Systems and Theory

57

other pre-arc conditions, were noted prior to distortion and concurrent arcing at the fault point. Smoke was noted at the fault point, which was found to be the arc burning through the remaining insulation during

Figure 34: Fault Detection (Night Vision Video)

Figure 35: Arc Damage (Night Vision Video)

Penrose_BMI.indd 57

4/22/08 5:41:58 AM

58

Electrical Motor Diagnostics: 2nd Edition

surge testing (Figure 35). In the follow-up MCA, it was noted that the impedance measurements across leads T2–T3 increased significantly (doubled) due to a change in the applied frequency as a result of a change to the condition of the circuit. The conclusion was that MCA testing was able to detect the fault sooner, between phases, than the surge comparison test. Due to the location of the fault, this was expected as it was a ‘deep winding’ fault in the center of a coil.

Experiment 2a One occasional observation, in the field, has been changes to MCA results following fast dV/dt high voltage testing, such as surge comparison testing, including inductance, impedance and other results. An assumption has been the dipolar alignment of the magnetic steel and dielectric insulation following the motor being at rest for some length of time. If this were to be the case, then it would have to be assumed that: a) Surge comparison testing would generate dipolar action and alignment; and, b) The core steel would maintain the dipolar action, assuming magnetic domains would be generated, which would cause some level of retained magnetism. Therefore, a used motor with a good winding, that was estimated to show evidence of this effect, was selected and surge tested to 1,800 Volts. Table 7: Before and After AT4 Test Results with Differences

Resistance Impedance Inductance Phase Angle I/F Ins Resistance

Before

After

Diff

85.422 300 458 79 −41

85.402 287 437 78 −41 >99

−0.02 −13 −21 −1 0

Table 8: Before and After AT31 Test Results and Differences at 200 Hz

Impedance Fi I/F

Penrose_BMI.indd 58

Before

After

Diff

525 72 −44

504 72 −45

−21 0 +1

4/22/08 5:41:58 AM

Electrical Insulation Systems and Theory

59

The after tests were performed as quickly as possible following the surge test. Each result was performed twice to ensure repeatability (all results were repeatable within 1%).The motor was quickly disassembled, a metal piece placed against the stator, which did not attract, then placed in front of a fan and reassembled and re-tested. All tests returned to the original test findings. The original concept of dipolar action discounts the following considerations: a) Paschen’s Law requires ionization of air and materials; b) As noted in the first part of this chapter, the fast rise time does not allow for significant polarization of materials; and, c) The assumption that there would be remaining magnetic flux would require steel with high magnetic retention properties, which would result in extremely high hysterisis losses in the machine. The observations provide the preliminary conclusion that the cause of change has to do with ionization of the air within the stator, in these instances. Ionized air and ozone (detected as an ozone ‘odor’ when the motor was disassembled) results in conductivity of the atmosphere between the conductors of the motor. Changes to the readings, during follow-up testing with MCA, indicate that this condition lasts a short period of time. This may also identify the result of partial discharge in VFD applications. The exchange of air supported the conclusion as well as the lack of retained magnetism following the test. Finally, none of the tests noted in this chapter, and other tests performed for support of the findings of this chapter, performed on stator-only conditions showed any of the same results, regardless of the amount of time not used. This particular area has been noted for additional research for its value in MCA.

Experiment 2b The motor was set up and isolated.The AT4 and AT31 were both used to analyze the condition of the windings, the AT31 was used to analyze the condition of the rotor.The surge comparison tester was used to evaluate the condition of the windings and rotor. Following all tests, the surge comparison tester was increased to 12,000 Volts to ensure condition of each phase. As outlined in the first part of this chapter, the impedance and inductance of MCA followed the same pattern, showing the insulation system was clean and dry. MCA identified a good winding in both instances.The AT31 identified bad rotor bars using a visual inductance rotor test. Each of the surge waveforms followed the same pattern, but required rotor

Penrose_BMI.indd 59

4/22/08 5:41:59 AM

60

Electrical Motor Diagnostics: 2nd Edition

position movement in order to determine condition of the windings, but did not identify two missing rotor bars.The surge test to 12,000 Volts identified the new winding was in excellent condition. Table 9: AT4 Test Results on Motor

Resistance Impedance Inductance Phase Angle I/F Ins Resistance

1–2

1–3

2–3

20.368 186 147 80 −43

20.336 183 144 79 −44 >99

20.334 154 122 79 −45

Table 10: AT31 Test Results on Motor (60Hz) Impedance Fi I/F

1–2 79.6 57 −38

1–3 78.3 57 −38

2–3 71.8 56 −38

Experiment 2c A stator was selected with winding contamination from grey water in the windings. The test results identified as 0.89 MegOhms to ground. The motor was tested with the AT4 and AT31, before and after the surge comparison test. Surge comparison was limited to 1800 Volts, max. A visible coil to coil fault was identified and detected using the MCA instruments. Table 11: AT4 Test Results on Stator

Resistance Impedance Inductance Phase Angle I/F Ins Resistance

Penrose_BMI.indd 60

1–2

1–3

2–3

0.6400 27 5 79 −49

0.5577 26 6 80 −49 0.89

0.6189 27 5 79 −49

4/22/08 5:41:59 AM

Electrical Insulation Systems and Theory

61

Table 12: AT31 Test Results on Stator (60Hz) Impedance Fi I/F

1–2 2.12 25 −20

1–3 2.08 29 −20

2–3 2.12 25 −20

Table 13: AT4 Test Results on Stator Post Test

Resistance Impedance Inductance Phase Angle I/F Ins Resistance

1–2

1–3

2–3

0.6400 27 5 79 −49

0.5577 26 6 80 −49 0.0

0.6189 27 5 79 −49

Table 14: AT31 Test Results on Stator Post Test Impedance Fi I/F

1–2 2.12 25 –20

1–3 2.08 29 –20

2–3 2.12 25 –20

Figure 36: Effect of Surge Comparison Testing on Winding Contamination

Penrose_BMI.indd 61

4/22/08 5:42:00 AM

62

Electrical Motor Diagnostics: 2nd Edition

The MCA tests identified the winding contamination and coil to coil short. There were no changes to the surge comparison waveform until the stator arced to ground at 1,600 Volts.The stator was in a condition to clean dip and bake prior to high voltage testing, but had a visible hole to ground following 3–5 seconds of applied test voltage.

Experiment 3 Part of the purpose for the exposed ‘virtual’ motor is to reduce the influence of ionization on the test results. The stator was tested with MCA, then re-tested until the applied voltage was less than 1,000 Volts before fault detection. Table 15: AT4 Test Results on Virtual Motor

Resistance Impedance Inductance Phase Angle I/F Ins Resistance

1–2

1–3

2–3

0.6821 21 8 77 −47

0.6798 21 8 76 −45 > 99

0.6865 39 7 76 −47

Figure 37: Virtual Motor with Spacers to Hold Airgap

Penrose_BMI.indd 62

4/22/08 5:42:00 AM

Electrical Insulation Systems and Theory

63

Figure 38: Surge Values on Each Test

Table 16: AT4 Test Results After Test 6 and Before Test 7 on Virtual Motor

Resistance Impedance Inductance Phase Angle I/F Ins Resistance

1–2

1–3

2–3

0.6604 21 8 77 −47

0.6592 21 8 77 −45 > 99

0.6569 40 7 77 −47

Table 17: AT4 Test Results on Virtual Motor After High Voltage Tests

Resistance Impedance Inductance Phase Angle I/F Ins Resistance

1–2

1–3

2–3

0.6786 21 8 77 −47

0.6797 21 8 77 −43 > 99

0.6819 39 7 76 −47

While this experiment proved that high voltage testing will degrade defects in windings, MCA will detect the fault and identify changes to

Penrose_BMI.indd 63

4/22/08 5:42:01 AM

64

Electrical Motor Diagnostics: 2nd Edition

Figure 39: Arc Damage to Winding and Insulation

the condition of the winding, and leads to an observation of the ability to detect and trend degradation, it represents a fault different from an operational fault. In an operational fault, the defect will overheat and a greater area of insulation will be effected, causing a more significant change to the phase angle and/or current/frequency test result as the insulation degrades. However, the test does provide evidence that the prediction of the first part of the chapter, concerning changes to results, due to a progression of winding degradation, does occur. It is observed that the final change to the current/frequency reading came from a point where additional conductors, other than the original two, were effected and the area of carbonization of the insulation increased. It is noted, in Figure 39, that the separation of the phase chapter around the fault point is a direct result of the surge comparison arc, evidence of Paschen’s Law in action. Low voltage test methods had no impact on the existing condition of the insulation system.

Experiment Conclusions The purpose of this chapter, and the experiments contained within, has been to provide additional supportive evidence for chapters, research and field conclusions using MCA.

Penrose_BMI.indd 64

4/22/08 5:42:01 AM

Electrical Insulation Systems and Theory

65

Experiment 1 Conclusions It has been determined that MCA is capable of detecting a phase to phase deep-winding defect before traditional test methods as it does not require a steep dV/dt. An arc is often formed when high voltage surge comparison testing detects a defect. The high energy involved degrades the remaining insulation life of the insulation defect.

Experiment 2a Conclusions Variations in MCA test results following high voltage testing appear to be due to ionization and not polarization. Additional work is being conducted to determine how best this effect can be used for fault detection, in particular in applications involving partial discharge.

Experiment 2b Conclusions A number of important conclusions concerning condition based monitoring resulted from this experiment. It was determined that MCA is capable of evaluating the condition of the stator insulation system without having to move the motor shaft. This makes MCA an excellent method for Predictive and Condition Based Maintenance of motors in-place. MCA was also found to be able to detect defective rotor bars using inductive-based rotor tests and movement of the shaft. High voltage surge comparison testing was incapable, during this experiment, of detecting rotor condition and required movement of the shaft in order to detect any winding defects. This makes it less capable as a method for PdM and CBM programs. It was noted that new insulation systems should be capable of withstanding high voltage surge comparison testing, including accidental over-potential. However, it is noted that in all aged systems that showed any defect, including minor, correctable defects, the high voltage surge comparison test degraded the insulation system significantly further than before testing. This presents the conclusion that any defect that could provide an estimated time to failure using MCA would represent a near-immediate fault in high voltage surge comparison testing. This further supports the conclusion that it is less capable as a method for PdM and CBM programs.

Experiment 2c Conclusions MCA was determined to have the capability to detect winding contamination without a negative impact on insulation condition. This provides

Penrose_BMI.indd 65

4/22/08 5:42:02 AM

66

Electrical Motor Diagnostics: 2nd Edition

for an excellent ability to evaluate system condition during predictive maintenance or condition based monitoring. High voltage surge comparison testing was determined to destroy correctable insulation system conditions in contaminated winding conditions.This supports the EASA (Electrical Apparatus Service Association) ANSI/EASA AR-100 Motor Repair Standard recommendation to clean insulation systems prior to applying high voltage tests. Also, it provides evidence of risk in using high voltage test methods in predictive maintenance applications.

Experiment 3 Conclusions It was determined that MCA has the ability to detect changes to the insulation system over time, providing the ability to estimate time to failure following fault detection. High voltage surge comparison testing was found to significantly degrade the insulation system after each successive test following fault detection, rendering it less capable of estimating remaining insulation life following fault detection.

Summary Each of the experiments supported the theory and conclusions of the first part of this chapter. However, it has been determined that additional experiments are to be planned, with an expansion to include Electrical Signature Analysis. The purpose of the additional work will be to prove further the original intent of Experiment 3 by generating an insulation defect, with the surge comparison tester, in a used submersible pump motor.The pump will then be run in normal operating conditions, with periodic MCA tests and periodic ESA tests in order to provide evidence of trending using MCA and to determine the point at which ESA can detect a winding short. Some of the conclusions of this study were surprising to the investigators including the repeated condition of the fault not showing until an arc was drawn, although the arcs were sometimes only visible with the lights out. In each case, there were no ‘waveform shifts,’ nor other warnings, prior to waveform distortion and arcing on the stators and assembled motors selected with real-life, versus laboratory induced, faults.

Penrose_BMI.indd 66

4/22/08 5:42:02 AM

Electrical Insulation Systems and Theory

67

Chapter 4 Questions 1. What kind of circuit are insulation systems modeled as? 2. What happens as a field is placed across an insulation system or magnetic steel? 3. How does a MegOhm meter work? 4. What is Polarization Index and what does it detect? 5. How does MCA technology work? 6. Compare MCA and Surge Test technologies.

Penrose_BMI.indd 67

4/22/08 5:42:02 AM

Penrose_BMI.indd 68

4/22/08 5:42:02 AM

Chapter

5 Common Electric Motor Testing

The requirement for an inclusive AC induction machinery standard has been needed that outlines known condition-based testing technologies and their capabilities. Over the past nine years, an IEEE Power Engineering Society Standards Committee has developed the “Guide for Induction Machinery Maintenance Testing and Failure Analysis,” designated IEEE Standard 1415–2006, published in May, 2007 through the Institute of Electrical and Electronics Engineers, Inc. What makes this particular standard unusual is that it provides an overview of both electrical and mechanical test methods and provides test limits where possible. In many cases, it splits a particular technology into sub-tests. For instance, an MCA test that involves resistance, impedance, inductance, phase angle, current/frequency and insulation to ground is broken into individual tests such as: Winding Resistance, Insulation Resistance, Phase Angle, Phase Balance (Inductance and Impedance) and Variable Frequency. The purpose is to both cover existing technologies and provide room for future technologies that may use different combinations.

Electrical Motor Diagnostics Defined One of the most troublesome areas that has come along with our modern times is keeping track of definitions. For instance, on-line can mean using the internet or while equipment is running. Lately, the concept of 69

Penrose_BMI.indd 69

4/22/08 5:42:03 AM

70

Electrical Motor Diagnostics: 2nd Edition

Electrical Motor Diagnostics has been considered as only the technologies of Motor Circuit Analysis (MCA) and Electrical or Current Signature Analysis. However, in 2004, the Institute of Electrical Motor Diagnostics (IEMD) defined Electrical Motor Diagnostics as all technologies used to test or evaluate the condition of the electric motor system or capable of being used in motor system maintenance and management programs. Motor system maintenance and management was then defined: Motor system maintenance and management is the philosophy of continuous improvement of all aspects of the motor system from incoming power to driven load. It involves all components of energy, maintenance and reliability from system cradle to grave. The result of these definitions is that a broader scope of technology is encompassed, providing the concept of a broader range of tools available for electric motor health diagnosis. This definition, combined with the IEEE Std 1415 gives us the slate of technologies that are to be discussed in this chapter. However, the focus will be on the motor only, including: • • • • • • •

Stator winding and core Rotor winding and core Vibration and noise Bearings and shafts Structure and frame Ventilation Accessories.

Condition-Based Technologies The following technologies are covered by the standard: • AC High Potential: Is a pass/fail test applied at twice the rated voltage plus 1,000 volts for new insulation systems and 125–135% of motor nameplate voltage for existing insulation systems. • Acceleration Time: Increased or decreasing starting times may indicate problems with power supply, motor or load. • Bearing Insulation: Evaluation of the insulation integrity of the bearings for purposes of reduced shaft currents and resulting bearing damage. Performed following IEEE Std 43-2000. • Bearing Temperature: Measured by RTD, thermocouple or bulbtype thermometer. Temperature limits vary but generally fall in the range of 90–100C for alarm and 105–120C for shutdown.

Penrose_BMI.indd 70

4/22/08 5:42:03 AM

Common Electric Motor Testing

71

• Capacitance: Measurement is trended and values to ground increasing over time indicate surface contamination, high humidity, high temperature or insulation breakdown. • Core Loss (Loop Test): Test performed during motor repair to evaluate the interlaminar insulation of the stator core. No spot should be greater than 10C than the ambient core temperature. • Coupling Insulation: Performed to ensure that no shaft currents flow into driven equipment. Performed following IEEE Std 43-2000. • Current Demodulation: Used in motor current signature analysis as a method of removing the fundamental frequency from current FFT spectra. • Current Running: Can be used as an indication of load. Pulsating current, measured with an analog current probe, is an indicator of rotor bar problems. • Current Signature Analysis: Used to provide analysis of electromechanical condition and driven equipment condition. Requires analysis of current FFT spectra. • Current Starting: Inrush and starting current is evaluated for anomalies. • DC High Potential: DC High Potential is a trendable test when leakage is recorded. Uses twice the voltage plus 1,000 volts time 1.7 as the maximum applied. If, while increasing voltage, the leakage value increases very quickly, then the test has failed. • Dielectric Absorption: Is a ratio of the DC insulation resistance readings of the 60 second value to the 30 second value.A ratio of 1.4 or greater, in pre-1970 insulation systems, is considered acceptable. Otherwise, trending is required. Reference IEEE Std 43-2000. • Dissipation Factor and Power Factor: Are both tests that use an alternating current voltage at the rated voltage of the motor being tested. The trended value should not exceed a change of 2% over the period of test. • Grease Analysis: Used to trend and evaluate deterioration of lubrication properties of grease. • Growler: Used to evaluate the condition of rotor bars when the rotor is removed from the electric motor. • Insulation Resistance: Measures the insulation value between conductors and ground after 1 minute. The applied voltage is less than the motor rated voltage with a temperature corrected result of 5 MegOhms for random-wound machines and 100 MegOhms for form-wound machines. Reference IEEE Std 43-2000.

Penrose_BMI.indd 71

4/22/08 5:42:03 AM

72

Electrical Motor Diagnostics: 2nd Edition

• Oil Analysis: Used to evaluate the degradation of the lubricating properties of oil. Can also be used to detect excessive mechanical wear in equipment. • Partial Discharge: Is a measurement of capacitive discharges within the electrical insulation itself. This value is trended generally on machines over 6,000 Vac. • Phase Angle: The timed measurement between the peak voltage and current at about 7Vac applied to a coil. When two coils are compared, the value should be within one digit of both results. • Phase Balance (Inductance and Impedance): Used to detect severe winding unbalances or to compare in order to detect winding contamination. Test results are compared phase to phase to determine if the pattern is the same, or not. • Polarization Index:The ratio of the 10 minute insulation to ground test and the one minute insulation to ground test. A ratio of 2 or more is required on pre-1970 insulation systems. Reference IEEE Std 43-2000. • Shaft Grounding Current: A measurement of the shaft current. Can identify that shaft currents are not flowing through the shaft grounding system. • Shaft Testing: Magnetic particle, liquid penetrant and ultrasonic examination are used to evaluate the condition of the motor shaft material. • Shaft Voltage: Voltage measurements taken from the shaft of the motor. Variations in the voltage value indicate problems with the motor. • Single Phase Rotor Test: 10 percent of the motor nameplate voltage is applied across one phase of the motor. The rotor is turned and current values taken.Variations of 3%, or more, of the current value through 360 degrees of rotation identify probable broken rotor bars. • Speed: Uses measurements of motor RPM in order to determine if potential motor or load problems exist. • Surge Test: High frequency, high voltage impedance-based test used to check the turn-to-turn dielectric strength of the insulation system. Waveforms compared with deviations indicating faults. • Surge PD: Variation of the surge test, evaluates partial discharges that result from the high voltage, fast rise-time test. • Thermography: Utilizes an infrared camera to compare the background (ambient) to the test component. Defects can cause a high temperature rise at the point of fault.

Penrose_BMI.indd 72

4/22/08 5:42:03 AM

Common Electric Motor Testing

73

• Torque Analysis: Uses three phases of voltage and current in order to calculate torque. The value is then displayed and analyzed as torque FFT spectra. • Ultrasound/Ultrasonic: Used to detect bearing and other electromechanical defects on motors. Also used to detect other motor system opportunities. • Variable Frequency: Using about 7 Vac, the motor current is measured then the applied frequency doubled and the resulting current compared to the initial current result.The value should be no more than one to two digits different between phases. • Vibration: FFT spectra of vibration information is used to trend and detect mechanical and some electrical faults. • Voltage Balance: Voltage measurements used to detect voltage unbalance defects in the supply. • Voltage Distortion: Harmonic content of voltage. If this value is too high, rotor and stator heating will occur. • Voltage Drop: Is a trended measurement of the voltage drop when starting a large electric motor. Changes may indicate electric motor defects. • Voltage Level: Voltage measurements are used to ensure that the supply voltage remains within +/− 10% of nameplate voltage. • Voltage Spikes: Monitoring voltage spikes allows the ability to evaluate supply and control conditions. • Winding Resistance: Used to detect broken wires and loose connections. • Winding Temperature: Winding temperature can be trended over time in order to determine if overload conditions or insulation failure is going to occur. Each of the tests is described, effectiveness determined (ie: Effective for trending?), online or offline, typical test frequencies, any precautions or considerations and related standards cited.

Voltmeter Troubleshooting of an AC Motor For the purpose of these chapter, we will assume that we have a True RMS multi-meter, a True RMS current clamp, an Analog current clamp and a 500/1,000 Volt insulation tester with an analog display with a range from 0.01 to 1,000 MegOhms. The systems that we will be testing will

Penrose_BMI.indd 73

4/22/08 5:42:03 AM

74

Electrical Motor Diagnostics: 2nd Edition

be across the line, AC induction electric motors with the intention of understanding basic troubleshooting with basic tools. It is assumed, in throughout this book, that you are following all appropriate safety and PPE requirements during testing.

The Power of a Volt Meter For most applications, in modern electrical environments, a True RMS digital volt meter is important to give you accurate values.This is because older averaging volt meters would display inaccurate, non-repeatable, values on digital displays or the needle would bounce in voltage harmonic situations.The True RMS meter compensates for these variations and harmonics. You will also want a meter with a range 10 times more accurate than the value you are looking for. For instance, if you are looking for values to the nearest 1 Volt, you want a meter accurate to at least 0.1 Volt. The most obvious methods of voltage testing are to check the voltage value and phase balance. In both cases, the proper method of taking voltage readings is to go phase to phase, which provides more accurate test values than phase to ground. The pattern is also important, if you are going to compare test results or analyze a system. A common pattern is Phase A to Phase B, Phase A to Phase C, then, Phase B to Phase C. Table 18: Phase to Phase Voltage Test Results Phase

Voltage

A–B

460

A–C

458

B–C

466

These three values can provide important clues as to the condition of the system and help identify some problems. For instance, over/under voltage conditions can change the operating conditions for the electric motor. NEMA identifies the maximum deviation from the nameplate of an electric motor as +/-10% for design purposes. In order to determine the maximum deviation, you determine the nameplate voltage and take the measured value that is the furthest from the nameplate.

Penrose_BMI.indd 74

4/22/08 5:42:04 AM

Common Electric Motor Testing

75

Equation 34: Voltage Deviation (460 Volt Motor) 466V − 460V × 100 = 1.3% 460V The impact as the voltage deviates from the nameplate can be significant, see Figure 40, with the maximum recommended deviation for energy purposes as 5%.

Figure 40: Impact of Voltage Deviation

The next concern is Voltage Unbalance, which results in unbalanced currents and magnetic fields in the motor. As the unbalance becomes greater, the temperature rise of the motor increases, generating the need to de-rate the motor. Equation 35:Voltage Unbalance (1) Determine the Average Voltage (460V + 458V + 466V )

3

= 461.3V

Equation 36: Voltage Unbalance (2) Determine the Unbalance 466V − 461.3V × 100 = 1% 461.3V The unbalance is then compared against Figure 41, which provides a multiplier against the motor horsepower. The electric motor is designed to work within 5% voltage unbalance, with an energy application recommendation of not more than 2% voltage unbalance.

Penrose_BMI.indd 75

4/22/08 5:42:04 AM

76

Electrical Motor Diagnostics: 2nd Edition

Figure 41: Voltage Unbalance Multiplier

NOTE: It is important to note that the motor may not be operated into its service factor in either condition.The motor service factor is only to be used at the motor nameplate voltage and frequency values.

Checking Contacts with a Volt Meter One method of testing contacts in a motor starter is to use a volt meter. When an infrared camera is used, the operator is looking for the I2R losses which show as Watts or Heat. Damaged connections show as a loose connection and related resistance, as do conditions of glazing that can occur in some operating environments. One way of detecting contact and connection problems without the use of an infrared thermometer or imager is to perform a voltage drop survey. When performing a survey, it is important to start with the volt meter set at a voltage equal to or larger than the circuit voltage. Check to ensure that voltage is being supplied to the starter or contactor by performing a phase to phase test on the supply side of the starter or contactor, as described above. Next, place one lead from the volt meter on the Phase A input side of the starter and one lead on the Phase A output side of the starter. Adjust the value of the volt meter down, if no value shows, until you get to a value or less than one volt.

Penrose_BMI.indd 76

4/22/08 5:42:05 AM

Common Electric Motor Testing

77

A good contact will have a value less than one volt.A poor, or failed, contact will have a value of one volt or greater. Perform the same steps across each phase of the starter or contactor. Once completed, check to ensure that you still have phase to phase voltage by re-testing the supply side phase to phase.

Checking Fuses with a Volt Meter Testing fuses while equipment is running is a straight forward process. There are two steps that are performed including the phase to phase and line testing. The phase to phase test for fuses involves ‘cross checking’ the fuses. Start by measuring phase to phase on the supply side of each fuse, as noted in the phase balance test. You place one lead of the voltmeter on the supply side of the Phase A fuse and the load side of Phase B fuse.You should see a full phase to phase voltage value. If you do not, then there is a problem with the Phase B fuse. Repeat the steps by placing one lead on the load side of the Phase A fuse and the supply side of Phase B fuse. Repeat the steps between Phase B and Phase C. Once finished, re-check phase to phase on the supply side. A less accurate method of checking fuses, but important for control circuit fuses or single phase applications, is to measure across the fuse.The steps and resulting values should be performed and evaluated in the same way as starter or contactor testing, as described above.

Ammeter Testing of an AC Motor Both analog and digital ammeters have their application, special features and capabilities of some ammeters will provide different capabilities. However, we will assume generic instruments.

Analog Ammeters The analog ammeter uses an electro-mechanical deflection of a needle that is directly proportional to the value of current that is being measured. Hand held meters will normally have a clamp and a display while panel mounted analog meters will have a Current Transformer (CT) providing a percentage of the actual circuit current to the display.The display is then incremented so that it shows the actual full value of current. The purpose of these meters is to provide a relative idea of existing load. However, they also provide some additional troubleshooting capability.

Penrose_BMI.indd 77

4/22/08 5:42:05 AM

78

Electrical Motor Diagnostics: 2nd Edition

For instance, because an ammeter measures the ‘real time’ value of current, with some delay that is related to the CT winding and the coil in the display, it can provide the ability to detect such things as broken rotor bars and torsional load problems. In the case of broken rotor bars, the needle will deflect at a rate that is similar to the pole pass frequency of the operating motor.As this frequency is often around a fraction of a Hz to several Hz, the movement of the needle will be easily visible to the technician.The result would be a ticking motion that will have a magnitude related to the operating current based upon the severity of the number of broken rotor bars. Unfortunately, this can also occur in applications that have a regular varying torque.

Digital Ammeters Digital ammeters come in two types: Averaging and True RMS. The averaging ammeter has a built-in algorithm that compensates for applications that include harmonic distortion, such as variable frequency drives or systems that have a lot of electronic systems connected on the circuit. True RMS (Root-Mean-Square) meters have a built-in circuit that provides readings based upon the real power of the circuit. In a harmonic environment, an averaging meter and an analog meter will not read correctly. However, a True RMS meter will provide an accurate value. The power of True RMS digital ammeters, hand held or panel, is the ability to compensate for these harmonic conditions and to provide an easily measurable value. Other features can include a min/max capture capability and datalogging capabilities.

Load Measurements Knowing the percentage of load can help determine if a motor is being overloaded. The value will change depending upon voltage unbalance and the supply voltage versus nameplate voltage. The effective load can be calculated, when the motor is operating above 50% of load, by considering both the voltage and current.The average phase to phase voltage (Equation 37), the average phase current (Equation 38), the nameplate voltage (Vn) and the nameplate current (An) are required. Equation 37: Average Voltage (Va) V + V2 + V3 Va = 1 3

Penrose_BMI.indd 78

4/22/08 5:42:05 AM

Common Electric Motor Testing

79

Equation 38: Average Current (Aa) Aa =

A1 + A2 + A3 3

Equation 39: Percentage of Motor Load ⎛V %Load = ⎜ a ⎜V ⎝ n

⎞ ⎛ Aa ⎟⎟ × ⎜⎜ ⎠ ⎝ An

⎞ ⎟⎟ × 100 ⎠

In an application where the motor nameplate is 460 Volts and the nameplate current is 38 Amps. The measured phase to phase voltage is V1 = 475;V2 = 482; and,V3 = 485.The measured phase to phase current is A1 = 33;A2 = 36; and A3 = 38. If you were to consider the current values only, it would appear that the motor was operating less than 100% of full load. By using current only, the average current would be 35.7Amps, and the load would be considered 93.9%. If using Equation 39, the average voltage would be 480 Volts and the resulting percentage load would be 98% load, over 4 points of efficiency different. If the motor is rated at 1.15 Service Factor, the motor would be required to have nameplate voltage and frequency supplied. In the example, above, the motor is operating at 4.3% over voltage and 1% Voltage unbalance. The actual load at 98% would be satisfactory. If, however, the voltage unbalance was greater than 2%, then the motor would be experiencing overload.

Current Unbalance Conditions The NEMA MG-1 calls for a maximum current unbalance of 7% when a motor is being factory tested. However, when a motor is being tested in a plant motor system, there are a number of reasons that can cause current unbalance. Some of the issues can include changes to the circuit impedance, phase unbalances, etc. A high current unbalance, with relatively low voltage unbalance, can be caused by conditions of a failing motor winding, loose connections or bad power factor correction capacitors. For instance, in one application, a 100 horsepower motor had current draw of 109 Amps, 109 Amps and 72 Amps.The motor user kept removing the motor and sending it in for rewind (which the repair shop was happy to do). After the third time, it was noticed that the system had power factor correction capacitors. They were tested and it was determined that one of three power factor

Penrose_BMI.indd 79

4/22/08 5:42:05 AM

80

Electrical Motor Diagnostics: 2nd Edition

correction capacitors had a blown fuse. The three capacitor leads were removed and the motor was retested with currents of 102 Amps, 101 Amps and 99 Amps. Due to other electrical conditions in the distribution circuit, such as single phase loads that cause power factor conditions, and unbalanced loads, not only can there be significant voltage unbalances, but current unbalance can be excessive. Where an unbalance of up to 7% unbalance is acceptable in factory conditions, unbalanced distribution power factor can cause unbalances far beyond that value. Additionally, this can be exaggerated when the motor is lightly loaded (less than 50% or idle). An idle motor will often have a current draw 20 to 40% of full load current.

Other Current Conditions Some ammeters have a min/max capability. This allows the operator to determine the actual peak current on each phase, when determining the maximum load. You can also determine the peak inrush for an electric motor by setting the ammeter to capture the peak current and starting the motor. This can be compared to the kVA Code of the motor. Table 19: kVA Code AC Induction Motors Code

Multiplier

Code

Multiplier

A

0–3.15

Code G

Multiplier 5.6–6.3

N

11.2–12.5

B

3.15–3.55

H

6.3–7.1

P

12.5–14.0

C

3.55–4.0

J

7.1–8.0

R

14.0–16.0

D

4.0–4.5

K

8.0–9.0

S

16.0–18.0

E

4.5–5.0

L

9.0–10.0

T

18.0–20.0

F

5.0–5.6

M

10.0–11.2

U

20.0–22.4

Equation 40: Calculating the kVA/HP (3 Phase Motors) kVA

HP

=

3 × Amps ×Volts HP × 1, 000

The locked rotor current can be calculated from knowing the nameplate information and the kVA/HP code (often listed on the nameplate as ‘Code.’).

Penrose_BMI.indd 80

4/22/08 5:42:06 AM

Common Electric Motor Testing

81

Equation 41: Calculating the Locked Rotor Amps (LRA) LRA =

1, 000 × HP × kVA / HP 3 ×Volts

For instance, if there is a 10 horsepower, Code F motor at 460 Volts and a nameplate current of 11.2 Amps, the LRA would be ~70 Amps. When testing the motor on startup, the max capture of the ammeter should fall close to 70 Amps. If it does not, then it can indicate a rotor defect or winding problems, especially when commissioning a new or repaired electric motor. With ammeters that can datalog or operate and log the min/max over a long period of time, then the loading can be determined over a full operation. For instance, if a motor starts tripping unexpectedly, the loading can be monitored to see if something has changed in either the loading or driven equipment. This type of reading can be used as a trendable value and as a proof test by maintenance to indicate if changes have been made to the system that need to be investigated.

DC Resistance Testing of AC Machines The use of the Ohm meter has been around long before AC induction motors. In these modern times, we have instruments that can measure extremely high resistances and those that can measure extremely low resistances. We will concentrate on Ohm meters, milli-Ohm meters and micro-Ohm meters.The proper selection of which one to use for analyzing your machines, within the capabilities of simple DC resistance testing, is extremely important.

Selecting the Right Measurement Tool There are both analog and digital Ohm meters, just as with other basic measurement instruments. In the past, analog Ohm meters provided excellent means of measuring the condition of machines with modern digital instruments providing additional accuracy. One of the issues, when taking resistance measurements on electric motors has to do with the resistance of the electric motor circuit. A rule of thumb, on machines under 600 Volts AC, is that the resistance will become less as the horsepower increases. For instance, an average

Penrose_BMI.indd 81

4/22/08 5:42:06 AM

82

Electrical Motor Diagnostics: 2nd Edition

Figure 42: Wheatstone bridge

one horsepower motor may have a resistance of 40 Ohms while an average 100 horsepower may have a resistance of 0.010 Ohms. Electric motors over 1,000 Volts AC will tend to have resistances over a few Ohms. The type of measurement method used by the instrument will be of importance as to the accuracy of the test.We will cover the basic circuits, the Wheatstone bridge, the Double Kelvin Bridge and the Four Wire Kelvin Bridge. The Wheatstone bridge (Figure 42) works by comparing resistances and a null balance meter to compare DC voltage in the circuit being tested. Basically, when the voltage between point 1 and the negative side of the battery is equal to the voltage between point 2 and the negative side of the battery, the null detector will show a zero voltage, and the bridge can be considered balanced. With the circuit, the state of balance is solely dependant upon the ratios of Ra/Rb and R1/R2. When used to measure an unknown resistance, the unknown

Penrose_BMI.indd 82

4/22/08 5:42:06 AM

Common Electric Motor Testing

83

is connected in place of Ra or Rb. The remaining three resistors are precision and any of these resistors can be replaced or adjusted until the bridge is balanced. Once the balance is obtained then the unknown value can be determined by the ratios of the known resistances. The accuracy of the meter depends upon the stray voltages that exist in the conductors that affect the null detector which result from wire and connection resistances within the bridge, and the accuracy of the precision resistors.

Figure 43: Double Kelvin Bridge

The Double Kelvin Bridge (Figure 43) is a modified version of the Wheatstone bridge. It compensates for the stray voltages within the circuit, and the circuit being tested, due to the wire and connection resistances. With the ratio Rm/Rn set equal to the ratio of RM/RN, Ra is adjusted until the null detector indicates balance. The actual balance equation for the Double Kelvin Bridge can be found in Equation 42.

Penrose_BMI.indd 83

4/22/08 5:42:07 AM

84

Electrical Motor Diagnostics: 2nd Edition

The accuracy of this type of bridge is at least 0.05% and usually used for circuits measuring in the milli-Ohm range. Equation 42: Balance Equation Double Kelvin Bridge Rx Ra

=

RN RM

+

Rwire ⎛ Rm ⎜⎜ Ra ⎝ Rm + Rn + Rwire

⎞ ⎛ RN Rn ⎞ − ⎟⎟ ⎜⎜ ⎟⎟ R ⎠ ⎝ M Rm ⎠

The next type is the 4-Wire Kelvin Bridge (Figure 44) which uses Ohms Law in order to determine the resistance. Its particular application is in circuits where there are long test leads and significant stray voltages. It also allows for very low (micro-Ohm) measurements with a high degree of accuracy. Figure 45 shows a more common Kelvin circuit in which the losses and inaccuracies due to an Ammeter are compensated for by using a calibrated shunt resistor, which also sets the accuracy of the instrumentation.

Figure 44: Basic 4-Wire Kelvin Bridge

In digital instrumentation, the circuit board is used in part to provide a portion of the bridge with the logic providing the adjustment for the null circuit.

Penrose_BMI.indd 84

4/22/08 5:42:07 AM

Common Electric Motor Testing

85

Figure 45: 4-Wire Kelvin Bridge

Some instruments will use the 4-Wire Kelvin Bridge for very accurate micro-Ohm measurements and others will compensate for circuit resistance, using a Double Kelvin bridge by measuring through the test wires first, before measuring the circuit under test.

Considerations When Using an Ohm Meter There are a number of things to consider when using an Ohm Meter to test an electric motor circuit. These considerations include: 1. Voltage present on the circuit when testing from a Motor Control Center (MCC), or disconnect. This voltage may include ElectroMagnetic Induction (EMI); 2. The temperature of the motor windings; 3. Connections through the circuit, as well as differing lengths of conductors, when testing from an MCC may cause an unbalance; and/or, 4. Whether or not there are series or parallel resistances in the circuit.

Penrose_BMI.indd 85

4/22/08 5:42:07 AM

86

Electrical Motor Diagnostics: 2nd Edition

In precision instruments, EMI or other voltages present in the circuit will cause readings to be non-repeatable.This can be observed if the measured value does not ‘settle’ during the measurement. The temperature of the motor windings must be compensated for when trending measurements. If you are troubleshooting an electric motor, comparing phase to phase does not require compensation.The reverse can be performed, as well, in order to determine the winding temperature of a motor, if an original temperature and resistance are known. In both of the formulae below, the K (constant) is based upon the material being tested, with Aluminum = 225 and Copper = 234.5, and RC being the resistance of the winding cooler than during the test and vice-versa. Equation 43: Resistance of Cold Winding ⎛ K + TC RC = RH × ⎜ ⎜ K +T H ⎝

⎞ ⎟⎟ ⎠

Equation 44: Resistance of Hot Winding ⎛ K + TH RH = RC × ⎜ ⎜ K +T C ⎝

⎞ ⎟⎟ ⎠

For example: If a copper motor winding was measured as 2.5 Ohms at 25oC, the technician later tests the motor at 3.6 Ohms. What is the winding temperature? Answer 1: RH ⎡ 3.6Ohms ⎤ ⎢ 2.5Ohms ( 234.5K + 25C ) ⎥ − 234.5K = 139.2C ⎣ ⎦ The next consideration is whether or not the windings are in series or parallel.This will affect the total resistance such that, for instance, if the induction motor is connected in series for the high voltage connection or parallel for lower voltage connections. It is also important to understand if there are multiple wires in parallel, even of different resistances, should one or more be broken. Equation 45: Series Resistances RS RS = R1 + R2 + ... + Rn

Penrose_BMI.indd 86

4/22/08 5:42:08 AM

Common Electric Motor Testing

87

Equation 46: Parallel Resistances RP 1 1 1 1 = + + ... + R p R1 R2 Rn If a circuit has a combination of series and parallel circuits, solve for the parallel circuits individually, then solve for the system as a series circuit.

Troubleshooting Motors with Ohm Meters When testing from phase to phase on an electric motor circuit, you must connect the motor for operating voltage. This way you are testing three equivalent circuits. Alternately, you may be testing from the MCC or a disconnect from phase to phase. The most accurate way to test an electric motor is to ensure that you are using an instrument capable of testing out to 0.001 Ohms, or to 0.0001 Ohms for higher horsepower 460 Volt motors. It should be noted that the low voltage and high voltage resistances of the motor will be different. If you have a resistive unbalance, the losses measured in Watts will relate to the resistance of each circuit and the amount of current. Additionally, spot resistance, such as with a loose connection, will cause local heating in relation to Equation 47. Equation 47: Resistive Losses Watts = I2R If a circuit is carrying 100 Amps and a connection has a resistance of 0.1 Ohms, then the loose connection will generate 1002 x 0.1 Ohms = 1000 Watts, or 1kW, at that point. Depending upon the standard used, the allowable Resistive unbalance, from phase to phase, should be no more than 2% at the motor. Unbalances greater than this amount, or unbalances that have increased, will often relate to loose connections, direct shorts or broken conductors within the motor or conductors.

Classical Insulation Resistance Testing The insulation to ground tester, or Meg-Ohm meter, is also one of the earliest instruments used by technicians in evaluating and troubleshooting

Penrose_BMI.indd 87

4/22/08 5:42:08 AM

88

Electrical Motor Diagnostics: 2nd Edition

insulation systems, including electric motor insulation systems. In this chapter, we will concentrate on the testing method as outlined in IEEE Standard 43-2000 (R2006), “The Recommended Practice for Testing Insulation Resistance of Rotating Machinery,” (IEEE 43) and a few additional methods for evaluating findings, we will also refer to the testing method as Insulation Resistance (IR) testing. The standard that we generally reference within industry is the IEEE 43, which went through a major revision in May, 2000. It was updated because post 1970 insulation systems went through a series of changes in their chemical makeup. The chemical makeup of newer insulation systems are very different from the older systems, including how they react through testing methodologies. The revised standard drastically changed a number of traditional testing programs for insulation resistance that had been in place for over 50 years, including the Polarization Index (PI), insulation to ground tests and AC vs DC testing of insulation systems. The purpose of the IR reading is to evaluate the condition of the insulation between the conductors in the stator slots and ground. This is done by applying a direct voltage between the conductors (windings) and the casing of the electric motor (machine) and measuring current leakage across the insulation system. The measurement of current and voltage, applied, provide a finding measured as resistance (Ohm’s Law: R = V/I). In the case of an insulation system, the leakage current may be measured in milli- or micro-Amps, with the lower the current reading, the higher the insulation resistance value. These IR readings change over time because of ‘insulation polarization.’ In effect, the insulation system consists of polarized atoms that ‘line up,’ or polarize, with the applied DC voltage. As they polarize, the insulation resistance will increase.

The Basic Insulation Resistance Test Straight insulation resistance testing has been used to troubleshoot and evaluate the condition of machines for over a century, often with disastrous results, in the hands of an inexperienced user. There are very clear limitations on the ability of insulation resistance testing, alone, to evaluate the condition of an electric motor for operation. For one thing, there has to be a clear path between the insulation system and the casing of the machine. Air, mica, or any other non-conducting material between the winding and ground will provide a high insulation resistance. Faults on

Penrose_BMI.indd 88

4/22/08 5:42:09 AM

Common Electric Motor Testing

89

the end-turns of motor windings will also not provide a clear path to ground, with most winding faults starting as internal winding shorts that might graduate to insulation faults. So, care must be taken when using IR as a troubleshooting tool. In performing IR, the proper method is to connect all leads together, test with the IR meter for a period of one minute, ensuring that the red test lead (negative) is on the leads and the black lead is on the housing. Once the IR measurement is obtained, it is then adjusted for temperature while the leads are grounded for four, or more, minutes. The values for IR applied voltage and minimum test values can be found in Tables 20 and 21. Table 20: Insulation Resistance Test Voltage Winding Voltage

IR DC Voltage Applied

1000

500

1001–2500

500–1000

2501–5000

1000–2500

5001–12000

2500–5000

12001

5000–10000

Table 21: Insulation Resistance Minimum Values Min

Winding Being Tested

Insulation Resistance at 1 Min kV + 1MegOhms

Most windings made before 1970

100 MegOhms

Form wound stators after 1970

5 MegOhms

Random wound stators under 1,000 Volts after 1970

There are a few things that have to be considered when performing insulation resistance from a Motor Control Center (MCC) or disconnect a distance from the motor under test. For one thing, if you tie all of the cable leads together and test, because of the surface area under test, it is possible that the readings may be only a few MegOhms. This does not necessarily mean that the system is bad, and a few tricks can be used to evaluate the condition of the cable. Additionally, any capacitors or lightning arrestors should be disconnected from the circuit and

Penrose_BMI.indd 89

4/22/08 5:42:09 AM

90

Electrical Motor Diagnostics: 2nd Edition

variable frequency drives or amplifiers must be disconnected from the motor. First, take each conductor and test between the conductor and ground. If the reading is greater by an order of magnitude then chances are that no problem exists. Next, disconnect the other end of the cable and separate the conductors from each other and ground. At the other end, perform the insulation resistance test between conductors. If the readings are above the minimum, then the insulation resistance of the cable is OK (however, it does not definitively clear the cable of any potential faults). The same process can be used on some motors, with the exception of a phase to phase test, unless the internal connections of the motor can be broken, such as in a wye-delta motor or all twelve leads are brought out of the machine. If the phases can be separated, then an insulation resistance measurement can be taken between phases. The results should be above the minimum values shown above.

Figure 46: Insulation Resistance Temperature Correction R40 = Kt x Rt

Penrose_BMI.indd 90

4/22/08 5:42:09 AM

Common Electric Motor Testing

91

During these tests, if you are using an analog IR meter, if the needle is not steady, or the digits ‘dance’ around in a digital IR meter, then there is the strong possibility that moisture or contaminants have gotten into the windings. The bouncing is the result of ‘capacitive discharge,’ or the build up of the DC energy within the winding that suddenly discharges and then starts to re-charge. Figure 46 represents the insulation resistance temperature correction chart for correcting to 40ºC. Using this chart, if the winding temperature is 60ºC and the insulation resistance was 200 MegOhms, the correction factor (Kt) would be ‘4,’ and the result would be 4 times 200 MegOhms which would be a corrected insulation resistance of 800 MegOhms. Dielectric Absorption

The dielectric absorption test, or ‘DA,’ is a ratio of the sixty second IR reading to the 30 second IR reading. As shown in Figure 47, the value

Figure 47: Dielectric Absorption

Penrose_BMI.indd 91

4/22/08 5:42:10 AM

92

Electrical Motor Diagnostics: 2nd Edition

at position A is divided by the value at position B. In a good insulation system, the IR will increase as a curve which will start fairly steep then will plateau, depending upon how fast the insulation system polarizes. Pass/Fail criteria can be found in Table 22. However, in insulation systems manufactured after 1970, it is not uncommon for insulation

Table 22: Dielectric Absorption Chart Insulation Condition

Dielectric Absorption Ratio

Dangerous

1

Questionable

1.0–1.4

Good

1.4–1.6

Excellent

1.6

systems to polarize rapidly and insulation systems with a temperature corrected one minute reading greater than 5,000 MegOhms may show a low value. In these instances, the test results should be used for trending only, and in the new IEEE 43, the test results must be corrected for temperature. Polarization Index

The Polarization Index, or PI, is the ratio of the 10 minute to 1 minute insulation resistance test. As shown in Figure 48, the result is the value at position A divided by position B. In a good insulation system, the IR will increase as a curve which will start fairly steep then will plateau, depending upon how fast the insulation system polarizes. Pass/Fail criteria can be found in Table 23. However, in insulation systems manufactured after 1970, it is not uncommon for insulation systems to polarize rapidly and insulation systems with a temperature corrected one minute reading greater than 5,000 MegOhms may show a low value. In these instances, the test results should be used for trending only, and in the new IEEE 43, the test results must be corrected for temperature.

Penrose_BMI.indd 92

4/22/08 5:42:10 AM

Common Electric Motor Testing

93

Figure 48: Polarization Index

Table 23: Polarization Index Insulation Condition

Dielectric Absorption Ratio

Dangerous

1

Questionable

1.0–2.0

Good

1.4–4.0

Excellent

4

Using the PI, the user should watch the needle if the meter is analog. If the needle bounces as it increases, then it represents capacitive discharge and an impending insulation problem such as contamination. If the meter graphs the PI as a chart, the user should review the data to see if there are any downward spikes or the graph shows a decreasing value across the ten minutes. This would also indicate insulation resistance defects.

Penrose_BMI.indd 93

4/22/08 5:42:10 AM

94

Electrical Motor Diagnostics: 2nd Edition

Utilizing Hi-Pot Testing for Insulation to Ground Stress Testing The type of testing performed by Hi-Pot testing only evaluates that plane between the conductors and the slot wall of the stator core, or the slot cell wall.They do not detect inter-turn winding faults or developing winding shorts. An additional important requirement in all high voltage testing is ensuring that the winding is clean and dry prior to testing.These conditions, of course, limits the abilities of this type of testing. However, there are a few tricks that can expand their test capability. All Hi-Pot testing is set up in a similar fashion: If possible, each phase is separated with each phase not being tested, RTD’s and other coils in the system are shorted to ground. This allows the insulation system between the coils being tested and the other insulation systems to be tested while also ensuring that there are no circulating currents and the leads are away from operators. While AC Hi-Pot testing is the most dangerous form of testing, the AC applied voltage and current generate some excitation of the insulation dipoles. This gives a more complete pass/fail analysis of the condition of the system. The operator must also ensure that the lead is held rigidly against the conductor prior to applying voltage otherwise the arc that is generated will cause spikes that may cause latent damage to the insulation system.When testing an electric motor in place, the danger to the equipment is even more severe because of the additional surface area of the cabling. Any additional components such as capacitors, variable frequency drives, etc. including current and potential transformers, must be disconnected and grounded to reduce the chance for damage. With the DC Hi-Pot, the safest approach is the step voltage test. If evaluating an electric motor rated under 600 Volts, step increments at 500 Volts, if above 600 Volts, step the voltage at 1,000 Volts.This reduces the charging current stresses on the insulation system. With the leads of other windings connected to ground (only if you can break the connections between phases) and components, you are also evaluating the condition of the insulation system between those phases as well as the phase being tested to ground. Just as with the AC Hi-Pot, everything should be disconnected if you are testing through the cable system to the motor. In both cases, the leakage current should be trended, this is the current that the meter stabilizes on after the voltage is increased. The trend should be a steady increase and any sharp increase in leakage current before the test reaches the calculated voltage indicates an insulation defect that should be corrected.

Penrose_BMI.indd 94

4/22/08 5:42:11 AM

Common Electric Motor Testing

95

The VLF DC Hi-Pot provides a slightly more inclusive test which is handled in the same way as the DC Hi-Pot. The primary difference between the two is that the VLF provides some level of excitation of the insulation system dipoles. This will more closely identify insulation to ground defects.

Evaluating Motor Condition with Low Voltage Advanced Diagnostics Up to this point, we have discussed the traditional and customary methods of testing AC induction motors while they are running or de-energized. We will now discuss three ‘modern’ methods of evaluating the condition of the insulation system including the two methods of Motor Circuit Analysis (MCA) and surge comparison testing. Each of these methodologies operates differently in how they evaluate the condition of the insulation to ground and the inter-turn insulation system of an electric machine. As we progressed through the previous sections, you will have noticed that the primary part of the insulation analysis focused on the groundwall insulation system. In Chapter 4 we discuss the operation of the insulation system, itself, and how high voltage ground-wall insulation test instruments work. However, the greatest number of electrical failures in a motor winding actually occur between conductors or coils with only 17% of insulation failures ending with an insulation to ground fault. Because of the reactive nature of turn and coil insulation breakdown, DC resistance readings from phase to phase will not detect them at any stage up until conductors break or somehow come into direct contact with each other.This means that you must either detect insulation weakness or insulation degradation. The technologies we are going to discuss in this section are insulation weakness (surge comparison testing) and insulation degradation (MCA) by looking at the reactive components of the system. We will also explore the three major players in these technologies: ALLTEST Pro, Baker Instruments and PdMA and how they use these technologies, in combination with some of the other testing we have discussed in this series, to provide motor diagnostics and condition-based solutions. We will discuss advanced diagnostic techniques later in this book.

Motor Circuit Analysis Modern MCA devices use low voltage sinusoidal outputs designed to excite the insulation system dipoles and surrounding magnetic steel

Penrose_BMI.indd 95

4/22/08 5:42:11 AM

96

Electrical Motor Diagnostics: 2nd Edition

dipoles as well as other DC insulation tests in order to obtain a more complete view of the condition of the insulation system. There are several key benefits to this approach: Size and voltage rating of the machine being tested do not matter; in many instances, specific pass/fail criteria can be applied for phase to phase comparisons; and, Degradation can be trended over time without any adverse effects on the existing condition. Detection of Winding Contamination There are several ways to detect winding contamination using MCA. These methods include: The traditional insulation resistance test; Polarization Index and Dielectric Absorption Tests; Capacitance; and, Impedance and inductance comparison. Due to the fact that one of the last measurements to change due to a winding fault is inductance (L), a test result of L can be used as a comparative baseline. This is important as the relative position of the rotor in an assembled machine will effect the reading due to mutual inductance.

Figure 49: Good Impedance and Inductance Pattern

Penrose_BMI.indd 96

4/22/08 5:42:11 AM

Common Electric Motor Testing

97

When considering the stator as the primary in a transformer and the rotor as the secondary, the number of turns, in relation to each other, will give slightly different readings. In each rotor position of an induction motor, the ratio of turns will be slightly different in each phase, causing a slightly different value of inductance. Winding contamination causes changes to the capacitance of the winding circuit due to the polarization of the contaminants (ie: moisture). When referencing the simple impedance formula, it identifies that a change in capacitance will have a negative impact on impedance. Also, with instruments that use very low voltages, the capacitive reactance has a more significant impact on the impedance as the capacitive value is more dominant.The result, using a relatively low frequency and sinusoidal output, is a collapse of impedance towards inductance in the phase which has the greatest impact from contamination or moisture. In cases of high humidity, the insulation has to have fissures or defects in order to

Figure 50: Bad Impedance and Inductance Pattern

cause the change. However, it is able to detect inter-turn, inter-coil and coil end contamination.

Penrose_BMI.indd 97

4/22/08 5:42:12 AM

98

Electrical Motor Diagnostics: 2nd Edition

Another impact will be on trended circuit capacitance.When compensating for the ambient humidity, on machines that have cooled to ambient temperature, and testing temperature, which also has a direct impact on capacitance-only measurements. The level of sensitivity of this type of measurement is a double-edged sword. While it has the sensitivity to ambient temperature and humidity changes from test to test, it can also detect changes very early, making it an excellent trendable test. Users with limited experience should consider additional testing prior to making a call based on this measurement. This measurement can also detect inter-turn, inter-coil and coil end contamination.

Figure 51: PI Curve and Capacitive Discharges

The Polarization Index (PI) and Dielectric Absorption (DA) will detect when contaminants build up between the conductors and ground.As previously mentioned, the ratio of the 10 minute to 1 minute insulation resistance reading for PI and the 1 minute to 30 second ratio for DA will detect most contamination issues when trended. However, for single PI or DA testing, a different approach can be taken. In this approach, the curve is monitored for the whole time. If the winding is clean and dry, the curve will be smooth. However, if there are sudden dips in the curve, it identifies capacitive discharges that indicate winding contamination. Overloaded Windings Overheated windings have a similar impact as winding contamination. The difference is that the insulation is thermally degrading causing

Penrose_BMI.indd 98

4/22/08 5:42:12 AM

Common Electric Motor Testing

99

increased resistance to dipolar action. In this case, the capacitance value will change causing a decrease in impedance in one or more phases. In both winding contamination and overloaded winding conditions, the end result will be a winding short. Winding contamination can be corrected through a clean, dip and bake, if detected and corrected in its early stages. However, once changes occur that allow for the detection of a winding fault, the winding will have to be replaced. If the issue is an overloaded winding, the corrective action is a rewind. Winding Shorts One of the keys to proper MCA testing is that inductance is not used as a primary method of detecting developing shorts. Instead, two specific measurements are used in combination to determine the type and severity of the defect.These measurements are the circuit phase angle and the current/frequency response method. When a defect occurs in the winding, it changes the effective capacitance of the complete circuit. The change in capacitance will directly effect how the low level current lags behind voltage with the usual result being an increase in capacitance and a reduction of the phase angle in the effected phase. Once the fault becomes more severe, it will begin to effect the surrounding phases. This normally occurs when the defect exists in one coil or between coils in the same phase. A very small change in capacitance within the circuit can be detected, allowing the detection of single turn faults and pinhole shorts when using very low frequencies. A second method of fault detection uses a current ratio, similar in method to the frequency response test used for transformer testing. However, the low voltage current is measured, then the frequency is exactly doubled and a percentage reduction in the low-level current is produced. When the frequency is doubled, small changes to capacitance between conductors or phases are amplified, causing a change to the percentage reduction when compared between phases. The combination of phase angle and current/frequency response allow for the detection of winding shorts and the type of short being detected in any size machine. Also, due to the use of low voltage and the result that only a small change to circuit capacitance is required to detect the faults, early winding defects can be detected quickly and trended to failure.

Penrose_BMI.indd 99

4/22/08 5:42:13 AM

100

Electrical Motor Diagnostics: 2nd Edition

Another method of detecting late-stage winding faults is to perform a rotor inductance test. At the completion of the test, if one phase is significantly shifted from the other phases, it either indicates a significant phase unbalance or a winding short. If the motor had been previously tested, a shift would indicate a winding short. Rotor Testing with MCA MCA inductance readings can be used to determine if rotor bar faults exist in the electric motor. The MCA device excites the core steel based upon the amount of current available to the circuit and reacts across the airgap. The direct relationship to the ability to detect the rotor circuit across the airgap depends upon the distance across the airgap, the area of the steel magnetized and the length of the stator core. In longer cores, the effect will carry across the airgap and excite the rotor core and induce the instrument frequency into the rotor circuit. In large machines, the amount of energy available from an MCA device allows for the detection of rotor defects only above the area immediately surrounding each coil side. This produces multiple effects: 1. The mutual inductance changes as the rotor position changes as a direct result of the change to the transformer ratio between the primary (stator) and secondary (rotor). A good rotor will show a repeating pattern, a bad rotor bar will change the transformer ratio and a defect will appear as a non-repeating pattern. 2. Fractures will be readily detected as the induced energy is relatively low and the oxides on the surface of the defect will block the low-level current with the same result as a bad rotor bar, as above. Whereas, in higher voltage rotor tests, the energy may be significant enough to pass through the defect. 3. In rare instances, the airgap may be too significant and very little to no variation of the mutual inductance occurs. In this case, larger defects, such as multiple fractures or a broken bar, will show as a variation in a straight line. 4. MCA technology has the ability to detect wound rotor, synchronous rotor field and other wound-rotor defects across the airgap. Because of the impedance ratio between the primary and secondary, rotor winding defects will show as a change to phase

Penrose_BMI.indd 100

4/22/08 5:42:13 AM

Common Electric Motor Testing

101

angle and current/frequency response and will vary based upon rotor position. The ALL-TEST IV PRO 2000™ MCA Instrument The MCA device manufacturer ALL-TEST PRO, LLC, provides lower cost, hand-held, MCA devices.This device uses two test leads, much like a multi-meter, and a 500 data-set memory that can be uploaded to proprietary software. The software system uses simple rules to evaluate the condition of the winding for the user and to trend data. The types of data collected by the unit include:  A DC milli-Ohm test that is compared for balance phase to phase. This allows for a percentage resistive unbalance that does not require temperature correction. It is basically used to detect gross loose connections, broken conductors or other resistive faults.  Circuit impedance at frequency selections of 100, 200, 400 and 800 Hz that the instrument picks based upon proprietary logic.  Circuit inductance.  Phase angle readings at the same frequency as impedance.  Current/Frequency readings with the base frequency the same as impedance.  An insulation to ground test to 100 MegOhms at a selectable 500 or 1,000 Volts. The primary purpose of the ALL-TEST IV PRO 2000 is to detect insulation defects indicating developing winding shorts, existing winding shorts, winding contamination and rotor faults.The resistance and insulation to ground tests are used to detect significant resistive and groundwall insulation faults. The PdMA MCE Instrument The MCA device manufacturer PdMA provides a laptop and case unit device.This device is able to connect to all three phases and ground during the testing cycle and the data is entered directly into the included laptop. The software provides alarms and trending capabilities to detect faults.

Penrose_BMI.indd 101

4/22/08 5:42:13 AM

102

Electrical Motor Diagnostics: 2nd Edition

The types of data collected by the unit include:  A 4-wire Kelvin bridge DC resistance test.This value is temperature corrected for trending and phase to phase comparison.  AC testing performed at 300 or 1200 Hz.  Inductance testing for phase to phase inductance balance comparison.  Circuit capacitance testing for trending.  Insulation resistance testing with temperature correction and the ability to perform a PI and/or DA test. The PI and DA tests produce the curves in order to detect capacitive discharges.  A higher voltage insulation test system is available as an option to allow for up to 5,000 Volt test values and higher PI and DA tests. The primary purpose of the MCE is to detect insulation to ground defects, rotor faults and later stage winding shorts.

Surge Comparison Testing The concept of high voltage testing for turn insulation faults has been around for over 80 years, but only put into practical application over the past fifty plus years. In modern times, if a motor repair shop or motor manufacturer is not using surge comparison technology to confirm the quality of their turn insulation system, they cannot meet modern motor repair or manufacturing standards or specifications. These include the ANSI/EASA AR-100 motor repair standard by EASA, the NEMA MG-1 standard and numerous IEEE standards. Surge comparison test instruments operate by sending out a high voltage, fast rise-time impulse. When the impulse is introduced to a coil, the peak of the impulse resonates based upon the impedance of the coils and a ‘ringing’ effect will appear at the peak of the impulse. If impulses are applied to two or more identical coils, and the peaks are viewed on an oscilloscope, the ringing should be exactly the same. If any of them deviate, it represents a difference between the coils. This technology operates via Paschen’s Law. The basic definition is that this guides the voltage required to ionize the space (gas or material) and then draws and electrical arc between two conductive surfaces. Jabob’s ladder is an example of Paschen’s Law at work and was often used in old horror and science fiction movies to show an arc forming between two antennae then climb to the top and repeat.When breakdown occurs

Penrose_BMI.indd 102

4/22/08 5:42:13 AM

Common Electric Motor Testing

103

between two conductors, how the energy reacts will depend on the type of breakdown or failure. When dealing with two points in atmosphere (air), a point of partial discharge will occur at a voltage below the point where an arc would be drawn. If there is weakness between two points in insulation, the fault may react with some level of partial discharge, or may not and go directly to an arc. The amount of voltage required depends on the distance across the fault point, the material across the fault and the depth into the coil. In smaller machines, there are a larger number of turns per coil and per phase. As the high energy impulse is applied to the winding, it dampens quickly across the first two to four turns up to hundreds of turns per phase. In larger machines, the depth of penetration of the high energy impulse will vary by the design and size of the coil.This means that when the new winding or used winding voltage is applied, the surge may not detect weakness further than a few turns into the winding. Both found conditions or missed faults are frequent enough that those that support the technology and those who object to the technology in field applications can identify numerous examples. The user should understand that both situations are very possible. When a breakdown does occur, it will happen at a specific voltage that relates to the severity of the fault.Typically, the minimum voltage to pass can be found in the following Equations. Equation 48: New Winding Surge Voltage Maximum Eapp = 2E + 1000 Volts Where Eapp is the Surge Voltage and E is the motor nameplate voltage rating Equation 49: Used Winding Surge Voltage Eapp = 0.75( 2E + 1000 Volts ) When an electric machine has a rotor, the position of the rotor has a direct impact on the shape of the waveform and will be different for each phase. With a standard surge comparison tester, the position of the rotor must be adjusted whenever comparing two coils in an assembled machine. Newer technology surge comparison testers, used in field applications, electronically compensate for rotor position. The size and voltage rating of the surge comparison tester will vary based upon the size of the machine to be tested. Commonly if there are a number of medium voltage motors that will be tested, up to 6000 Vac,

Penrose_BMI.indd 103

4/22/08 5:42:14 AM

104

Electrical Motor Diagnostics: 2nd Edition

a surge tester with a voltage limit of 12,000 Volts is used. If all of the motors to be tested are rated under 600 Vac, new testers are available with an upper limit of 2,000 Volts. When a winding is evaluated, and is in new and good condition, the waveforms associated with the condition of the machine will be identical between phases when electronically compensated. In fact, a new winding can often have voltages applied many times the maximum values calculated in Equations 48 and 49 without degrading the insulation system. It is not until energy crosses between conductors that any latent damage, or accelerated degradation, may occur. As a result of the potential damage that may occur when weakness is found, the newer digital surge testers monitor for the point where partial discharge occurs, which, by eyesight, would be an unstable surge waveform. This is very important as previous testers relied upon user operation and interpretation.The newer digital technology testers are designed to avoid potential destructive fault detection by identifying the variation in the waveform prior to the arc being drawn. This reduces the instances where the detection of insulation weakness causes immediate faults. Two of the primary digital surge comparison instruments are the Electrom TIG instruments and the Baker Instruments Advanced Winding Analyzer (AWA).With the AWA providing a trendable phase to phase value and software, we will discuss that system. There are two primary models, the AWA IV, which has a 12,000 Volt limit and includes a built-in surge comparison tester, polarization index, DC Hi-Pot, insulation resistance and winding resistance testing. This unit weights 42 pounds and requires a power outlet. The AWA II is a solution for users of smaller electric motors and includes the same tests with a 2150 volt limit. This one weighs 18 pounds and also requires a power outlet. Both contain a software for trending, record keeping and troubleshooting.

Comparison of the ‘Big Three’ There has been a long battle of chapters and chapters by the manufacturers of both low and high voltage winding analyzers. To be quite blunt, the verbiage is presented in such a way that claims one or the other is not effective or unable to live up to the manufacturer’s claims. These discussions are, quite simply, market speak – as the users of each of the technologies will attest.

Penrose_BMI.indd 104

4/22/08 5:42:14 AM

Common Electric Motor Testing

105

Following is a short summary of the strengths of each of the primary technologies’ capabilities at the time of the writing of this book, in alphabetical order:

• ALL-TEST Pro, LLC: Manufacture the ALL-TEST PRO 2000 and ALL-TEST PRO 31 analyzers. Both are hand-held instruments that are focused on evaluating insulation degradation, contamination, insulation troubleshooting and rotor testing. While they each include an insulation to ground tester, this is not their primary focus. The ALL-TEST IV PRO 2000 instrument saves data in memory that can be uploaded and downloaded to software that provides data analysis and trending. Requires a minimal level of training for successful use. • Baker Instruments: Manufacturer of surge comparison test instruments including the AWA II and AWA IV. The built-in software provides trending capabilities with a focus on insulation weakness due to insulation defects and wear. The insulation resistance test capabilities also test for insulation to ground weakness and built in limits provide an increased level of automatic protection which are designed to reduce the chance of damage and will stop testing if a fault is detected early enough. These instruments require a medium level of training and experience for successful use. • PdMA Corp.: Manufacturer of the MCE instrument which is portable and includes a laptop. The focus of this instrument includes insulation to ground testing, winding contamination and rotor testing with the ability to detect later stage turn insulation faults. This instrument requires a medium level of training and experience for successful use. All have the capability of testing AC induction motors from the motor control center, disconnect or right at the motor. Each will detect cable faults as well as motor winding faults and each will detect insulation fault sin advance of failure.

Introduction to Electrical Signature Analysis In the early 1980’s, several different approaches were taken to look at the electrical signatures of rotating machines. One approach was to look at the electrical current, which became known as Motor Current Signature Analysis (MCSA) and one was developed by Oak Ridge National Labs

Penrose_BMI.indd 105

4/22/08 5:42:14 AM

106

Electrical Motor Diagnostics: 2nd Edition

for the detection of broken rotor bars in Motor Operated Valves (MOV’s) in the nuclear power industry. This second method looked at both the voltage and current signatures and became known as Electrical Signature Analysis (ESA). MCSA is primarily used by the vibration industry using special current probes which allow the vibration data collectors to take current input. This current is then converted from analog to digital, filtered and produced as an FFT (Fast Fourier Transform) spectra of amplitude versus frequency. ESA has been primarily used by the dedicated ESA instrument manufacturers and includes the voltage waveform as an input. The primary difference is that current tells the user what is from the point of test towards the load and voltage provides information from the point of test towards the supply. This allows the user to quickly determine where a particular signature exists. We will discuss Electrical Signature Analysis and its application in AC induction motor circuits. ESA provides the capability of detecting power supply issues, severe connection problems, airgap faults, rotor faults, electrical and mechanical faults in the motor and driven load, including some bearing faults. It is important to note that the technology should not be considered a replacement for vibration analysis in mechanical analysis, but provides excellent data on motor condition from incoming power through to the rotor. From the bearings to the mechanical load still remains in the realm of vibration, in most cases.

Fault Detection Using ESA One of the original concepts behind the development of ESA was to eliminate the loss of instrumentation to test MOV’s in the dangerous areas within nuclear power plants.The primary failure of these machines is the rotor which would overload and melt when limit switches failed. It was discovered that the rotor bar failure signature was unique enough that not only could the signature be quickly identified, but that condition values could be applied easily. When the Pole Pass Frequency sidebands (P1 and P2) of Figure 1 are compared to the values in Table 1 the condition of the rotor bars can be determined. However, in this case, the motor is 4,160 Vac and the data was taken from the Motor Control Center (MCC) Current Transformers (CT). The result can be a dampening effect on those peaks resulting in the analyst needing to estimate the severity of the fault.

Penrose_BMI.indd 106

4/22/08 5:42:14 AM

Common Electric Motor Testing

107

Figure 52: Broken Rotor Bar Signature Table 24: Rotor Bar Failure Levels dB

Rotor Condition Assessment

Recommended Action

>60

Excellent

None

55–60

Good

None

49–54

Moderate

Trend Condition

43–48

High Resistant Connection or Cracked Bars

Increase Test Frequency and Trend

37–42

Broken Rotor Bars Will Show in Vibration

Confirm with Vibration, Plan Repair/Replace

31–36

Multiple Cracked/Broken Bars, Poss Slip Ring Problems

Repair/Replace ASAP

100 MegOhm16

Indicates poor insulation to ground (ie: ground fault)

When a motor does not have a rotor in place, such as in a motor repair shop with a stator only, the tolerances change: Table 36: MCA Tolerances (Disassembled Motor – All sizes) Test Result

Tolerance

Resistance (R)

99M

0.168 44 8 77 –45

4/22/08 5:42:33 AM

160

Electrical Motor Diagnostics: 2nd Edition

Figure 61: Inductive Unbalance/Rotor Test

The unbalance was found to be striking, and related to the unbalanced current, vibration and overheating of the motor. Possibilities were explored ranging from power quality to test equipment calibration. All were satisfactory.When interviewed, the manufacturer noted that process changes were made at a particular location for larger concentric wound machines. In a motor of this size and speed, the first set of concentric coils (one phase) curls under the following phases, reducing the equipment’s winding appearance and mechanical strength. In order to combat that issue, the manufacturer made a decision to significantly increase the size of the first set of coils in their automated process (first phase) that also happens to be the furthest from the rotor. This allows the coil ends to appear without having to make post-winding modifications to the coils. No dynamometer testing, full load testing, or otherwise was performed on the motor design other than an applied voltage impedance test that ‘met design requirements.’ During the interview, a statement was made that any equivalent manufactured motor would exhibit the same results. Several other motors of the same frame size, other manufacturers, etc. were evaluated. None resulted with the same findings.

Penrose_BMI.indd 160

4/22/08 5:42:33 AM

Motor Circuit Analysis Testing AC Machines

161

Figure 62: Impedance Balance Test Using Alternate Impedance Tool at 1200 Hz Table 43: Alternate Motor

Resistance Impedance Inductance Phase Angle I/F Insulation

T1-T2

T1-T3

T2-T3

0.223 36 7 70 −39

0.225 35 7 69 −39 #.#

0.225 36 7 70 −39

It is interesting to note that on the ‘alternate’ motor example, while the general phase balance is common to other motors in this frame size, with the same concentric winding type (with the exception that all coils are equivalent in size) a defect in the waveform is identified. As it turns out, this particular defect is common to this particular manufacturer, and is identified as a casting void defect, which alters the smooth curvature of the sine-wave. Through experience, it has been found that

Penrose_BMI.indd 161

4/22/08 5:42:34 AM

162

Electrical Motor Diagnostics: 2nd Edition

Figure 63: Inductive Unbalance/Rotor Test, Alternate Motor

Figure 64: Impedance Unbalance, Alternate Motor

Penrose_BMI.indd 162

4/22/08 5:42:35 AM

Motor Circuit Analysis Testing AC Machines

163

casting voids that impact the sides of an inductive waveform have a minimal impact on the reliability of a motor, however, if the deformity is at the crest, it will significantly impact the motor’s ability to produce torque. The result of this study concluded that the particular motor design was not acceptable to the user for the critical application. As a result of these findings, the power plant now performs MCA testing on all new critical motors to check for winding and rotor defects that may impact equipment reliability.

Case 2: Casting Void 200 HP Motor A motor repair shop reported that a brand-new 200 horsepower, 1800 RPM motor was unable to provide enough torque to obtain no-load speed at full voltage with the coupling detached from the load.A replacement motor was sought and compared to the original using inductive measurements with changes to the rotor position. The primary change was to the inductance at the peak of the sinewave. The fault was found to be a large casting void whose position was identified by a mass of balancing weight on the rotor. The casting void was confirmed using vibration analysis, then found by drilling a small hole in the rotor in the approximate position then a small wire placed through the hole. The casting void was determined to be close to the rotor core, interrupting the current path.

Figure 65: Rotor Defect

Penrose_BMI.indd 163

4/22/08 5:42:36 AM

164

Electrical Motor Diagnostics: 2nd Edition

Figure 66: Good Rotor

Case 3: ID Fan Motor Rebuild Defect This case is slightly different in that MCA was not applied. A 5000/9000 horsepower, 712/886 RPM, 6600 Vac PAM motor was sent to a motor repair shop for a rebuild. The rotor was remanufactured and the motor returned to service. The initial startup changed from the original 14 seconds to ~40 seconds with a voltage drop from 7200 Vac to 5800 Vac. When the motor changed from low to high speed, the original speed change was 7 seconds to a new time of 39 seconds. When questioned, the repair facility claimed that it used new rotor laminations that would improve the motor efficiency and that the fault must be in the application. Their calculated torque curve appeared to exceed the original motor’s capabilities, however, observed operation proved otherwise. A consultant was brought in to review the application and, upon obtaining the original then new information, was able to determine, over several weeks of gathering information, the cause of the changes in operation. When the rotor was determined to have loose rotor bars and damaged laminations, it was decided to rebuild the rotor. The repair facility assumed that the rotor bar material was the same naval brass used in similar sized motors by the manufacturer. In addition, newer silicone steel laminations were manufactured and the core slots moved in by 0.125 inches per side and the rotor bar slot dimensions were increased by over 10%. As it turns out, the original rotor bar material was actually a Low Silicon Bronze.

Penrose_BMI.indd 164

4/22/08 5:42:36 AM

Motor Circuit Analysis Testing AC Machines

165

The resulting modifications to the rotor resulted in a change of : Table 44: Modification to Rotor

Circ Mils DC Resistance Endring Circ Mils

Before

After

1.4324 241.27/bar 7.1618

1.4773 108.31/bar 8.9124

The rotor lamination modification also changed the back iron density of the rotor. In effect, the torque curve actually changed so that the starting torque was significantly lower than the original, as well as the starting current being significantly higher.The breakdown torque did increase, as did the full load operating speed. If MCA had been applied to before and after tests, a defect would have been immediately apparent in all AC readings (Z, L, Phase Angle and I/F), before re-installation, resulting in a cost avoidance of weeks of downtime.

Case 4: Cable Faults During an onsite visit it was noted that six of twelve motors on a baghouse at a steel manufacturer had been routinely failing since the baghouse’s initial installation 18 months before. The initial observation was that the 150 horsepower motors would trip immediately upon startup, in particular on warm days or if restarted immediately after operating. Over time each of the six motors ended up with winding faults termed as ‘lightning strikes’ by the repair facilities.The other six motors exhibited none of the same results. The fan motors were set up as two groups of six immediately next to each other. All components were of the same manufacturer installed at roughly the same time. Each of the six motors were fed from two transformers fed from the same bus. Engineering firms, repair centers and consultants were presented with the opportunity to resolve the issue. The motors were evaluated and the cables tested conductor to conductor with Meg-Ohm meters and DC high potential test sets. In addition, power quality measurements were made and operation chart recorded. No faults could be found. Out of curiosity, the off-line motor circuit analysis was applied. The first application was on a motor that was being removed because it would

Penrose_BMI.indd 165

4/22/08 5:42:37 AM

166

Electrical Motor Diagnostics: 2nd Edition

not start from the motor control center. It was assumed that the winding had faulted. Table 45: Good 150 HP Motor

Resistance Impedance Inductance Phase Angle I/F Insulation

T1–T2

T1–T3

T2–T3

0.026 18 3 66 −24

0.025 18 3 66 −24 #.#

0.026 18 3 66 −24

As shown in the above table, the motor appeared completely balanced. Testing was then performed from the MCC to the motor, in which there was a disconnect at the motor. In a few instances, the simple resistance was not repeatable, and it was determined that the cables ran through a common cable tray to the motors. This resulted in mutual inductance between the energized and de-energized equipment, so a series of manual AC tests were performed. On one motor, a phase-to-phase impedance of ~1500 Ohms at 800 Hz was detected. This was exceedingly high (they were averaging 18 Ohms at 800 Hz). It was immediately discovered that the disconnect had been open between the starter and the motor.The cable was isolated from the starter and the disconnect, and the value remained roughly the same. Each of the cables was tested, on the side that the motors were failing, and the values fell between 1200 Ohms Z at 800 Hz to 3500 Ohms Z at 800 Hz. The lower readings were found on the cables that had a load just before they were disconnected. These results could not be duplicated on the other group of six motors. When queried further, it was discovered that there was a very interesting specification for how the cable was to be installed. A very precise engineering spec stated that the cable had to be unrolled prior to being run. It appeared that the cables that were installed on the side that the motors were failing on had been strung across a road and had been run over prior to installation by a variety of heavy vehicles. The result was damage between conductors that showed up during startup, and that gradually became worse as the defects broke down.The condition of the cable from wire to wire was not seen in any DC measurement. In the

Penrose_BMI.indd 166

4/22/08 5:42:37 AM

Motor Circuit Analysis Testing AC Machines

167

case of the motor that was about to be replaced because of a suspected winding fault, it was found that the cable had finally shorted. The complete MCA field study took approximately 1 hour.

Case 5: Evaluation of an 8000 HP Synchronous Motor A large 8000 horsepower, 13.2kV, 200 RPM synchronous motor that operates a reciprocating compressor at a chemical plant tripped on ‘shortcircuit’ warning during startup, using a General Electric Multilin®, in 65 milli-seconds. The motor was tested with a number of technologies, including surge testing, which did not detect a fault. The ALL-TEST IV PRO™ (older model) was employed to investigate the winding and isolate the fault. The initial test provided results as follow: Table 46: Initial Synchronous Motor Tests (8,000 hp)

Resistance Impedance Inductance Phase Angle I/F Insulation

T1–T2

T1–T3

T2–T3

0.322 189 37 81 −42

0.319 181 38 85 −49 >99

0.319 190 37 83 −46

These readings clearly indicated a fault. In order to investigate further, and to isolate the fault, additional readings were taken with the rotor in different positions and several with the motor neutral opened (one phase tested at a time). Table 47: Synchronous Motor, Second Reading

Resistance Impedance Inductance Phase Angle I/F Insulation

Penrose_BMI.indd 167

T1–T2

T1–T3

T2–T3

0.318 190 37 83 −45

0.316 192 38 86 −49

0.321 190 37 81 −44

>99

4/22/08 5:42:37 AM

168

Electrical Motor Diagnostics: 2nd Edition

As observed, with additional tests to confirm the finding, there will most likely be rotor faults as the fault readings follow the re-positioning of the rotor. In addition, impedance and inductance readings do not follow each other, indicating that the windings are in poor condition (confirmed based upon previous partial discharge testing). The initial conclusion was, based upon how quickly the motor protection tripped, that there was either a catastrophic stator winding failure or that there was a problem with one of the secondary circuits. In any case, it was concluded that there were definite rotor faults.This particular synchronous motor would start with the rotor field circuits shorted, using them to assist the amortizer windings to produce enough torque during startup. Upon further investigation, once the motor was partially disassembled, additional testing found a catastrophic failure in the rotor fields and the rotor field circuit open. This would cause the rotor bar resistance to increase dramatically, magnetically saturating the amortizer winding, which would result in an immediate and rapid increase in current. The control instrument detected this sharp current increase as a short. The stator windings were tested as being poor, but operable.

Case 6: AC Traction Motor – Good The following results are from a good rebuilt subway traction motor: Table 48: Traction Motor Test Data

Resistance Impedance Inductance Phase Angle I/F Insulation

T1–T2

T1–T3

T2–T3

0.037 4 0 51 −45

0.037 4 0 52 −45 >99M

0.037 4 0 51 −45

Several key points can be determined from the above data: All readings are very balanced – representing a good motor. Inductance readings are ‘0’ – This is a result of a very low inductance reading (less than 1mH). In the automatic mode, readings beyond the integral number is removed. Resistance readings of traction motors are very low, often in the low milli-Ohm range.

Penrose_BMI.indd 168

4/22/08 5:42:38 AM

Motor Circuit Analysis Testing AC Machines

169

Case 7: Spindle Motor – Good Following is a Yaskawa Spindle motor that tested good on a test bench: Table 49: Good Spindle Motor

Resistance Impedance Inductance Phase Angle I/F Insulation

T1–T2

T1–T3

T2–T3

0.237 6 1 55 −43

0.235 7 1 55 −43 #.#

0.234 6 1 55 −43

Case 8: Brushless DC Servo – Coil-to-Coil Short The following Brushless DC Servo motor tested coil-to-coil short on a test bench: Table 50: Shorted DC Brushless Servo

Resistance Impedance Inductance Phase Angle I/F Insulation

T1–T2

T1–T3

T2–T3

2.239 101 20 81 −47

1.554 40 8 76 −47 >99M

1.367 25 5 70 −46

Figure 67: Visual Representation of Faulted Brushless DC Servo

Penrose_BMI.indd 169

4/22/08 5:42:38 AM

170

Electrical Motor Diagnostics: 2nd Edition

Note the slight variation in resistance, significant changes to impedance, inductance, over 5 points of phase angle, with a balanced I/F. The phase angle compared to I/F indicates a coil-to-coil fault. The relationship between phases, can be visually represented as follows:

Case 9: 100 Horsepower, 1800 RPM with Dirty Windings The following 100 Horsepower, 1800 RPM motor was found to have seized bearings and had been submerged during operation. One concern was whether the motor was salvageable. Table 51: Winding Condition – Dirty Windings Resistance Impedance Inductance Phase Angle I/F Insulation

T1–T2 0.883 173 34 78 −43

T1–T3 0.883 150 30 79 −44 >99M

T2–T3 0.883 79 31 79 −44

Figure 68: 100 HP Visual Data

Penrose_BMI.indd 170

4/22/08 5:42:38 AM

Motor Circuit Analysis Testing AC Machines

171

The following information can be gathered from the above data and the following visual data: All but the impedance and inductance are fairly balanced. The steep angle found as the difference between impedance and inductance is due to leakage to ground from contamination that has effected the insulation. The small ‘shift’ in phase angle and I/F indicates that some minimal damage may have occurred due to contamination. However, the windings may be salvageable.

Case 10: Servo Motor – Turn-to-Turn Fault The following servo stator was found to have a turn-to-turn short during bench testing: Table 52: Shorted Servo Data

Resistance Impedance Inductance Phase Angle I/F Insulation

T1–T2

T1–T3

T2–T3

0.068 1 0 21 −33

0.068 1 0 22 −34 #.#

0.082 3 0 42 −45

Of particular interest is that the resistance is balanced outside the range of a standard Ohm-meter. The phase angle and I/F readings are significantly different. When both of these readings are out of range, the most common cause is a turn-to-turn short.

Case 11: Rotor Position Related Unbalance In the following data, the shaft of the assembled motor could not be turned. By using the visual representation provided by HAL Junior™, the unbalance could be determined as rotor position by comparing the inductance and impedance readings on the graph. The other data shows that the stator winding is in excellent condition.

Penrose_BMI.indd 171

4/22/08 5:42:39 AM

172

Electrical Motor Diagnostics: 2nd Edition Table 53: Rotor Position Readings T1–T2

T1–T3

T2–T3

Resistance

12.08

11.69

11.97

Impedance Inductance Phase Angle I/F Insulation

159 126 82 −51

151 120 82 −51 #.#

170 135 82 −51

Figure 69: Rotor Position Visual Representation

Case 12: Motor Tested from MCC – Shorted Windings The following data was collected from a Motor Control Center. Follow-up analysis determined the fault was due to shorted windings in the motor. The motor was still operating. I would occasionally trip overloads at 20% full load of operation. All unbalances point towards a severe winding fault. Phase angle and I/F deviations indicate turn faults.

Penrose_BMI.indd 172

4/22/08 5:42:39 AM

Motor Circuit Analysis Testing AC Machines

173

Impedance to Inductance relationship indicates winding contamination. Conclusion: The winding is failing due to winding contamination and will require a repair versus replace decision immediately. Table 54: Shorted Windings Measured from MCC

Resistance Impedance Inductance Phase Angle I/F Insulation

T1–T2

T1–T3

T2–T3

0.842 52 20 79 −44

1.084 96 19 83 −39 #.#

0.164 93 18 76 −39

Figure 70: Visual Representation of Shorted Windings

Case 13: Brand New 300 Horsepower Motor with Shorted Phase Just because an electric motor is brand new, it cannot be taken for granted that faults do not exist. The following example was taken from a brand new motor at a show, in the manufacturer’s booth.

Penrose_BMI.indd 173

4/22/08 5:42:40 AM

174

Electrical Motor Diagnostics: 2nd Edition

The unbalanced I/F readings indicate a phase fault within the electric motor. New electric motors will have occasional phase unbalances with winding faults rare and rotor casting void faults being more common. Table 55: New 300 HP Motor with Winding Fault

Resistance Impedance Inductance Phase Angle I/F Insulation

T1–T2

T1–T3

T2–T3

0.011 5 1 47 −39

0.011 6 1 46 −34 #.#

0.01 6 1 46 −33

Case 14: 3000 Horsepower with Fractured Rotor Bar The following data was taken with the motor assembled. Inductance readings were taken every 15 degrees.

Figure 71: Broken Rotor Bar Data

Penrose_BMI.indd 174

4/22/08 5:42:41 AM

Motor Circuit Analysis Testing AC Machines

175

Particular issues found with this rotor data are: In smaller electric motors there will be one ‘sine-wave’ for each pole of the motor. For instance, a 3600 RPM motor would have two sinewaves. On a larger motor this will double (ie: medium voltage motors), so that a two pole motor will have four sine-waves. This is because the instrument puts out enough power to only read the influence of the rotor bars above each coil side, whereas it will influence the complete coil on smaller motors. In the above case, the motor was a 3000 Horsepower, 4160 VAC motor, 1800 RPM. As the rotor position is changed, the broken rotor bars will cause a dramatic change to the inductance between the stator and rotor.

Case 15: New 50 Horsepower, 2-pole Motor Combined Faults The following information was gathered on a new 50 Horsepower, 3600 RPM electric motor prior to shipment to a customer for a critical application. The combined problems found within this motor caused its rejection prior to shipment and installation. The faults were confirmed using vibration analysis.

Figure 72:

Penrose_BMI.indd 175

Combined Rotor and Stator Faults, Inductance

4/22/08 5:42:42 AM

176

Electrical Motor Diagnostics: 2nd Edition

The shift of inductance and impedance indicate that there will be an inherent current unbalance. The slight indentations in the sides of the sine-waves indicate casting voids within the endturns and possibly the rotor bars. The casting voids are serious enough that they impact the impedance of each phase.

Figure 73:

Combined Rotor and Stator Faults, Impedance

Case 16: 300 HP, 2300 Volt, 3600 RPM Broken Rotor Bars Motor repaired prior to shipping, a rotor bar fault was detected in copper alloy rotor bars. Detected due to deviations in rotor inductance tests. Notes: Impacts to the inductance waveforms at the peaks were used to detect rotor bar faults. The change to the rotor inductance test over each position where the readings gradually increase then decrease, also indicate that the rotor was off-center.

Penrose_BMI.indd 176

4/22/08 5:42:42 AM

Motor Circuit Analysis Testing AC Machines

177

Figure 74: Broken Rotor Bars, 350 HP, 3600 RPM, 2300 Volt

Case 17: 1000 HP, 3600 RPM – Special Signature on Good Motor Special signature found on a 1000 HP, 3600 RPM motor. Unusual signature has to do with winding and rotor interaction – special rotor and winding design. Referred to as the ‘Batman signature.’

Figure 75: 1000 HP, 3600 RPM, Batman Signature

Penrose_BMI.indd 177

4/22/08 5:42:43 AM

178

Electrical Motor Diagnostics: 2nd Edition

Case 18: 700 HP, 3600 RPM, Broken Rotor Bars and Cracked Shaft Instead of a smooth waveform, because of a combination of multiple broken rotor bars and the cracked shaft, significant variations can be seen in the waveforms.

Figure 76: 700 HP, 3600 RPM, Broken Rotor Bars and Cracked Shaft

Penrose_BMI.indd 178

4/22/08 5:42:44 AM

Motor Circuit Analysis Testing AC Machines

179

Chapter 7 Questions 1. Describe how MCA can be used to identify existing winding shorts. 2. What conditions can cause changes to phase angle and current/ frequency response in an AC induction motor? 3. Describe how to perform a rotor-compensated test on a Generator. What would the tolerances be?

Penrose_BMI.indd 179

4/22/08 5:42:44 AM

Penrose_BMI.indd 180

4/22/08 5:42:44 AM

Chapter

8 Motor Circuit Analysis Testing DC Motors

The key to proper operation of a DC machine is its ability to have an alternating current available in the armature such that the magnetic fields are about 90 degrees from each other. This means that there must be a contact system that acts as a ‘switch’ to allow current to flow in a specific direction such that the fields remain at 90 degrees. The switching system used to convert the incoming DC to AC, in the DC motor, are the brushes and the commutator. In this chapter, we are going to discuss some of the conditions that must be considered for both components of the system.

The Commutator The commutator is made up of many copper/copper alloy bars with mica insulation between each bar. These bars are normally linked together with a core, or hub, that can be tightened to ensure a tight fit for the bars. When installed, the commutator should have a shiny, protective gloss that is made up of a surface film of copper oxide and graphite. The film, after the machine has been running less than 24 hours, will vary based upon the environment, but should have an appearance color of light straw to jet black. The color normally falls somewhere in between. Water vapor absorbed by this layer actually provides lubrication that reduces both the brush and commutator wear.

181

Penrose_BMI.indd 181

4/22/08 5:42:44 AM

182

Electrical Motor Diagnostics: 2nd Edition

A commutator that is out of round, grooved, burned, has flat spots or is in otherwise bad condition, should be turned and undercut. Many years ago, the commutator surface could be ground by hand, but safety requirements will normally not allow for it, so we will not cover that option in this chapter. Instead, we will assume that the commutator will be cut when the machine is out of service and the armature is on a lathe, or a portable lathe is being used. When turning a commutator, it is first important to check for loose bars. This can be accomplished using some modern instruments that provide information on commutator roundness. An older method was to use a light-weight hammer and tap gently on each bar. If there is a bell-like tone, then the bar is OK, if there is a dull thud, then the bar is loose. In these cases, the commutator can be tightened to manufacturer’s specifications using a torque wrench as too much pressure may cause the bars to distort. The cut, itself, should be just enough to remove any imperfections from the bars and to ensure that it is round. Once the commutator is ‘turned,’ it should be polished to an exceptionally smooth finish with a mirror-like appearance. Some repair shops will ‘thread’ the commutator slightly with the explanation that this will seat the brushes in operation. Instead, this will end up causing excessive brush and commutator wear while reducing the contact surface area. Once the commutator has been turned, it must be undercut to ensure that mica is below the copper bars. The optimal cut is when the depth of the mica is equal to the thickness of the mica between the bars.This is important as it maintains the mechanical integrity of the commutator while preventing the mica, which is harder than copper, to create ridges as the copper wears. These ridges will cause sparking and chipping of the brushes. Following undercutting, the edges of the cuts should be chamfered to ensure a smooth transition of the brushes across the bars. After the commutator has been prepared through turning and undercutting, it must be handled in a way that contaminants, including finger prints, do not come into contact with the surface of the commutator.

Brush and Commutator Operation The surface film, or glazing, on the commutator is extremely important for proper lubrication and requires some level of humidity for proper lubrication between the brushes and bars. Raw copper surfaces can

Penrose_BMI.indd 182

4/22/08 5:42:44 AM

Motor Circuit Analysis Testing DC Motors

183

be caused for a number of reasons such as brush grade, low current density, or gritty contaminants. When this occurs, the brushes will wear rapidly.

Figure 77: Coefficient of Friction

The optimal current density of the brush and commutator contact is 40 Amps/inch2 (average) to maintain a good surface film. Most DC machines are manufactured assuming that they will operate within a specific range and the current density of the brushes averaging in the range above. For light loads, a number of strategies can be considered, including reducing the number of brushes and using a ‘softer’ brush grade that will film easier.The brush company or distributor can often assist in the proper selection of a proper brush grade specific to an application. In addition to water vapor providing lubrication, the proper current density will cause the brush temperature to increase, reducing the coefficient of friction between the surfaces (Figure 77). Another concern for the life of the brush and commutator is atmospheric contamination such as oil, saline (salty air), corrosive gasses, or silicone vapors. Of these, silicone vapor is insidious in its attack, as some maintenance staffs will use silicone sealant on brush rigging covers, connection boxes, etc. in an effort to seal the motor

Penrose_BMI.indd 183

4/22/08 5:42:45 AM

184

Electrical Motor Diagnostics: 2nd Edition

from contamination. However, it only requires a very little amount of silicone material such as silicone tape, silicone-rubber lead wire, sealant, or any other source of silicone to cause extremely rapid brush wear. The fast brush wear will generate a significant amount of carbon within the motor, causing low insulation to ground results, which will also accelerate brush wear, and will lead to both ground and inter-turn winding shorts. Oil contamination is often indicated when a thick, black film builds up on the commutator surface. This film causes very poor brush contact and excessive heating at the few conductivity spots which can cause the copper to become excessively hot.When the temperature becomes high enough, very small globules of copper will come away from the commutator of which some will embed in the brushes causing threading of the commutator bars. In worst-case conditions, copper globules will be found within the motor that will often be described as ‘thrown copper.’ One of the more effective solutions for this condition, in addition to reducing the oily contamination, is to install harder brushes that will scour the contaminants from the commutator. Another area of concern within the brush/commutator system is the connections between the brush shunts (conductors) and the brush rigging. If a shunt is loose, less current will flow in the affected brush, reducing its operating temperature while the brush with a good connection will have an increased operating temperature. As brushes are dielectric, as the temperature increases, the conductance will improve causing more current to flow in that brush. In effect, when inspecting the brushes with infrared, the hotter brushes are actually the good brushes.

Seating, Tensioning and Cleaning When installing brushes, it is important to ensure that they are in proper contact with the commutator. This process is referred to as ‘seating the brushes.’There are a number of ways of approaching this, for this chapter we will discuss a common method used in repair shops: 1. Wrap sandpaper around the commutator such that the surface is facing outwards. If a significant amount of seating is required, start with a rough chapter.The sandpaper MUST be sandpaper and not an emery cloth or paper that contains metals. 2. Place the brushes in the brush holders and use the brush springs to put pressure on them.

Penrose_BMI.indd 184

4/22/08 5:42:45 AM

Motor Circuit Analysis Testing DC Motors

185

3. Rotate the shaft in the direction that the machine will normally operate. 4. Lift the brushes and ensure that over 95% of the brush surface is in contact. 5. Use a finer grade of sandpaper to clean up the surfaces of the brushes. 6. Remove all of the carbon using a vacuum cleaner or low pressure instrument air. Be careful that carbon is kept away from the motor windings. 7. Re-install the brushes and turn the shaft again. Over 95% of the brush surface should be in contact with the commutator surface. If the motor is in operation, or to clean up the brush contact, a seating stone may be used. This is placed directly against the commutator while the machine is operated at speed. A seating stone is very soft and gritty. The grit is caught by the commutator and will cause the brushes to wear very quickly while at the same time being soft enough not to damage the commutator surface. All company and regulatory safety considerations should be met if performing this particular task. Once the brushes are properly seated, they can be tensioned. This is normally done with a brush tension gauge which is used to lift the brush directly out of the brush holder and observe the value on the scale. The value will depend on the brush type and application, however average brush tension values are in Table 56. Adjustments are normally made at the spring. Table 56: Recommended Brush Tension Brush Grade

Tension, psi

Carbon

1.75–2.5

Carbon-Graphite

1.75–2.5

Graphite-Carbon

1.75–2.5

Electrographitic

2–3

Graphite

1.25–2

Metal Graphite

2.5–3.5

Fractional HP Motors

4–5

Finally, cleaning of commutators and brushes should be performed periodically during operation. The commutator itself, can be cleaned using a canvas wiper, which is a system of canvas layers and an insulated

Penrose_BMI.indd 185

4/22/08 5:42:45 AM

186

Electrical Motor Diagnostics: 2nd Edition

pole.This is held against the commutator to remove grit, carbon and other contaminants from the film while not removing the film, itself. Carbon dust can also be removed using low pressure instrument air (60 55–60 49–54 43–48

5

37–42

6

31–36

None None Trend Increase Test Intervals and Trend Confirm with motor circuit analysis Overhaul

7