Service Handbook for Transformers

DISCLAIMER OF WARRANTIES AND LIMITATION OF LIABILITY

THE INFORMATION, RECOMMENDATIONS, DESCRIPTIONS AND

SAFETY NOTIONS IN THIS DOCUMENT ARE BASED ON OUR
EXPERIENCE, JUDGEMENT, AND DOCUMENTS IN THE PUBLIC
DOMAIN WITH RESPECT TO TRANSFORMERS. THIS INFORMATION
SHOULD NOT BE CONSIDERED TO BE ALL INCLUSIVE OR COVERING
ALL CONTINGENCIES. IF FURTHER INFORMATION IS REQUIRED,
THE TRANSFORMER DIVISION OF ABB INC. SHOULD BE
CONSULTED.

NO WARRANTIES, EXPRESSED OR IMPLIED, INCLUDING

WARRANTIES OF FITNESS FOR A PARTICULAR PURPOSE OR
MERCHANTABILITY, OR WARRANTIES ARISING FROM COURSE OF
DEALING OR USAGE OF TRADE, ARE MADE REGARDING THE
INFORMATION, RECOMMENDATIONS, DESCRIPTIONS, AND SAFETY
NOTATIONS CONTAINED HEREIN. IN NO EVENT WILL ABB LTD. BE
RESPONSIBLE TO THE USER IN CONTRACT, IN TORT (INCLUDING
NEGLIGENCE), STRICT LIABILITY, OR OTHERWISE FOR ANY
SPECIAL, INDIRECT, INCIDENTAL, OR CONSEQUENTIAL DAMAGE OR
LOSS WHATSOEVER. THIS INCLUDES, BUT IS NOT LIMITED TO,
DAMAGE TO OR LOSS OF USE OF EQUIPMENT, PLANT OR POWER
SYSTEM, COST OF CAPITAL, LOSS OF PROFITS OR REVENUES,
COST OF REPLACEMENT POWER, ADDITIONAL EXPENSES IN THE
USE OF EXISTING POWER FACILITIES, OR CLAIMS AGAINST THE
USER BY ITS CUSTOMERS RESULTING FROM THE USE OF THE
INFORMATION, RECOMMENDATIONS, DESCRIPTIONS, AND SAFETY
NOTATIONS CONTAINED HEREIN.

i
 ACKNOWLEDGEMENTS

This Transformer Service Handbook is meant to provide a general understanding of

service as it relates to transformers. Service is a technical product that a transformer
needs until the end of its lifetime.

These pages provide an introduction to transformer service and maintenance, and are a
guide to help increase the value of the product, by protecting and prolonging the asset
life for customers and/or owners.

The material was compiled and written by ABB experts from our Transformer Business
Unit, based on their vast knowledge of transformers and many years of global
experience in the field of transformer manufacturing and service.

You are holding in your hands the end result of this challenging work – the Service
Handbook for Transformers.

Leif Carlzon, Group Vice President and Product Group Manager for Transformer
Service, Asim Fazlagic, Vice President for Transformer Service North America, Dr.
George Frimpong, Transformer Service expert in USA, Pierre Boss, Senior Transformer
expert in Switzerland and Pierre Lorin, Technology Manager for Product Group
Transformer Service have led the project by compiling, writing and editing the material
in this handbook.

We also thank the ABB employees and industry partners who supplied valuable input
and information, as well as a number of organizations which generously permitted us to
use their materials and documentation in the creation of this handbook.

Their support and contributions made this project possible.

We are convinced that readers will find our Transformer Service Handbook a very
useful and comprehensive source of answers to the many questions relating to
transformers and a trouble-free product life.

At ABB, we don’t just build high quality transformers - we take care of them so they stay
that way.

Tarak Mehta
Group Senior Vice President
Head of Business Unit Transformers
Power Product Division
Zurich, Switzerland

ii
 FOREWORD

ABB possesses the technology rights of more than 30 brands including ABB, ACEC,
ASEA, Ansaldo, Bonar Long, Breda, BBC, CGE, Challenger, Elektrisk Bureau, Elta, GE
(> 40 MVA), GTE, Gould, IEL, ITC, ITE, Indelve, Industrial Design, Italtrafo, Lepper,
MFO, Marelli, Moloney Electric, National Industri, Nitran, No-Tra-Mo, Ocren, OEL, OTE,
Richard Pfeiffer, Sécheron, Strömberg, TIBB, Thrige, Westinghouse, Zinsco.

At some utilities these transformers can account for up to 70-80 % of the utility’s total
transformer asset base. With this in mind, we undertook the task of providing for the
industry (users of ANSI/IEEE as well as IEC standards) a reference guide with detailed,
yet easy to understand, information for the proper care and maintenance of
transformers. This information should in no way supersede the maintenance guidelines
provided by the transformer manufacturer.

The engineering staffs at ABB keep abreast of new information and techniques
available for analyzing problems in transformers. In many cases, we are the pioneers of
such new ideas. In keeping up with new ideas, we have realized there is a wealth of
information on transformers available in the open literature. However, this information is
at times found in little known journals, brochures, and books. What we have attempted
to do with this handbook is to compile the most useful information into a single
document. The goal is that this will serve as the preferred reference manual for all who
are involved in the operation and maintenance of transformers. We have melded this
information with our many years of experience in designing transformers and providing
maintenance and diagnostic guidance to customers. This book can also be used as
training material in many universities and schools, to help students gain specific
knowledge about transformer service and maintenance.

The material presented in this handbook is not meant to provide theoretical insights into
the methods used for maintaining transformers. Instead, it is written to help the user
gain a better understanding of why certain measurements are recommended, and in
some cases, how to interpret the results of these measurements. There are three ABB
publications that provide theoretical coverage and discussions on transformers, short
circuit strengths as well as the testing of power transformers and shunt reactors
(Transformer Handbook, Short circuit duty of Power Transformers and Testing of Power
Transformers and Shunt Reactors available from the ABB website:
www.abb.com/transformers).

The layout of the handbook is as follows. We open with a general description of

transformer design to help the user understand the nature of the various components
that require maintenance in a transformer. Knowing the condition of a fleet of
transformers is important for making informed decisions about any maintenance, repair
or replacement activities. Therefore we address the topic of risk
assessment/management of transformers. We present ABB’s methodology of risk
assessment as applied to populations of transformers with the view of identifying the
few that need the attention of asset managers. This provides them the ability to focus on

iii
condition based rather than time based maintenance activities. This method has been
successfully applied to transformer fleets of many utilities and industrial customers
worldwide. The result has been to improve the availability of the fleet as a whole and at
the same time optimize the maintenance spending where it has the best impact. This is
followed by a general discussion of the various methodologies available for diagnosing
potential problems in transformers. The subsequent sections, which constitute the bulk
of the material in the handbook, provide detailed descriptions and discussions on the
test methods and interpretation of results used to maintain and repair transformers,
either in workshops or at site. Finally, we cover the environmental aspects related to
transformers and the important topic of economics of transformer asset management.

We would like to thank all the authors for their valuable contribution to making such a
comprehensive book about using the transformer as a valuable asset for improving
Power and Productivity for a Better World™.

Leif Carlzon Asim Fazlagi Pierre Lorin

Group Vice President Vice President & General Manager Product Group Service
Head of Product Group Service ABB TRES North America Head of Technology
Zurich - Switzerland Saint Louis, Missouri - USA Geneva - Switzerland

iv
 AUTHORS

The first international version of this handbook was written in collaboration with ABB employees from
several countries. We want to thank them all for this impressive team work.

In Brazil
Lars Eklund and Dr. Jose Carlos Mendes

In China
Henry-HongGuang Huang and Fred Samuelsson

In Germany
Sonia Berhane and Dr. Peter Werle

In India
Jivraj Sutaria

In Ireland
Mark Turner

In Italy
Paolo Capuano

In Norway
Knut Herdlevar and Arnt-Sigmar Todenes

In Spain
Miguel-Angel del-Rey, Rafael Santacruz and Nicolas Toribio

In Sweden
Dr. Dierk Bormann, Dr. Kjell Carrander, Dr. Mats Dahlund, Dr. Uno Gäfvert, Bjorn Holmgren, Lars
Jonsson, Peter Labecker, Lena Melzer, Peter Olsson, Dr. Lars Pettersson and Bengt-Olof Stenestam

In Switzerland
Dr. Jose-Luis Bermudez, Pierre Boss, Cedric Buholzer, Thomas Horst, Paul Koestinger, Pierre Lorin,
Jean-François Ravot, Ralf Schneider, Serge Therry, Olivier Uhlmann and Thomas Westman

In Thailand
Manoch Sangsuvan and Ekkehard Zeitz

In Turkey
Taner Danisment, Sener Ertuna and Burhan Gundem

In United Kingdom
Liam Warren

In United States of America

Wayne Ball, Gary Burden, Dr. Clair Claiborne, Eric Doak, Asim Fazlagi , Dr. George Frimpong, Ed Fry,
Dr. Ramsis Girgis, Axel Kalt, Greg Leslie, Dr. T.V. Oommen, Mark Perkins, Eric Pisila, Rich Ronnau,
Craig Stiegemeier and Brian Twibell.

v
A special recognition goes to our colleagues who wrote the first ANSI/IEEE version of the handbook used
as a base for the international version.

Also we would also like to thank Doble Engineering, IEEE, CIGRE, GE Energy, FLIR
Thermograpgy, Megger, Physical Acoustics, Electrical World Magazine, and the various other
organizations that allowed the use of their materials in this handbook.
Special thanks go to the three general reviewers
Pierre Boss, Dr. George Frimpong and Mark Turner

vi
 CONTENTS
DISCLAIMER OF WARRANTIES AND LIMITATION OF LIABILITY ........................................................ I
ACKNOWLEDGEMENTS .......................................................................................................................II
FOREWORD..........................................................................................................................................III
AUTHORS ..............................................................................................................................................V
1 TRANSFORMER DESIGN CONSIDERATIONS ........................................................................... 17
1.1 CONFIGURATION ..................................................................................................................... 17
1.2 MECHANICAL CONSIDERATION .................................................................................................. 17
1.3 THERMAL CONSIDERATIONS ..................................................................................................... 18
1.4 DIELECTRIC CONSIDERATIONS.................................................................................................. 19
1.5 CONSTRUCTION TYPES ............................................................................................................ 19
1.5.1 Shell Form........................................................................................................................ 19
1.5.1.1 Design Features ................................................................................................................... 19
1.5.1.2 Mechanical Strength ............................................................................................................. 20
1.5.1.3 Thermal Capability ................................................................................................................ 22
1.5.1.4 Dielectric Characteristics........................................................................................................ 24
1.5.2 Core Form ........................................................................................................................ 26
1.5.2.1 Design Features ................................................................................................................... 26
1.5.2.2 Mechanical Strength ............................................................................................................. 27
1.5.2.3 Thermal Capability ................................................................................................................ 29
1.5.2.4 Dielectric Characteristics......................................................................................................... 30
1.6 BUSHINGS ............................................................................................................................. 32
1.6.1 Design and Construction of Capacitances in Condenser Bushings Complying
with the IEEE Standards .................................................................................................. 32
1.6.2 Bushings Voltage Tap....................................................................................................... 36
1.6.3 Connections ..................................................................................................................... 38
1.6.3.1 Internal Electrical Connections ................................................................................................... 38
1.6.3.2 Draw Lead Connected Bushings................................................................................................ 38
1.6.3.3 Bottom Connected Bushings ..................................................................................................... 38
1.6.4 Liquid Level Indication ...................................................................................................... 38
1.6.5 Painting ............................................................................................................................ 39
1.7 ON-LOAD TAP CHANGERS ....................................................................................................... 40
1.7.1 Introductions..................................................................................................................... 40
1.7.2 North-American Practices ................................................................................................ 41
1.7.2.1 General Description of LTCs ................................................................................................. 41
1.7.2.2 Reactance Type LTCs........................................................................................................... 41
1.7.2.3 Arcing Control Methods......................................................................................................... 42
1.7.2.3.1 Arcing Tap Switch ............................................................................................................ 42
1.7.2.3.2 Arcing Switch and Tap Selector ........................................................................................ 42
1.7.2.3.3 Drive Mechanism for Reactance Type LTCs...................................................................... 43
1.7.2.4 Vacuum Interrupter Type LTCs.............................................................................................. 43
1.7.2.5 Resistance Type LTCs .......................................................................................................... 44
1.7.2.6 Drive Mechanisms for Resistance Type LTCs ........................................................................ 45
1.7.2.7 Failure Mechanisms for LTCs................................................................................................ 45
1.7.2.7.1 Electrical Connections ...................................................................................................... 45
1.7.2.7.2 Insulation System ............................................................................................................. 46
1.7.2.7.3 Control System................................................................................................................. 47
1.7.2.7.4 Mechanism ...................................................................................................................... 47
1.7.3 European Practices .......................................................................................................... 47
1.7.3.1 Resistance Type OLTCs ....................................................................................................... 47
1.7.3.2 Diverter Switch OLTC ........................................................................................................... 48
1.7.3.3 Selector Switch OLTC........................................................................................................... 49

vii
 1.7.3.4 Tie-In Resistors..................................................................................................................... 51
1.7.3.5 Failure Mechanisms for OLTCs ............................................................................................. 52
1.7.3.5.1 Electrical Connections ...................................................................................................... 52
1.7.3.5.2 Insulation System ............................................................................................................. 53
1.7.3.5.3 Motor Drive Mechanism.................................................................................................... 53
1.7.3.5.4 Mechanism ...................................................................................................................... 53
1.8 STREAMING ELECTRIFICATION .................................................................................................. 54
1.8.1 Charging Tendency and its Effect of Streaming Electrification ........................................... 55
1.8.2 Mitigation Strategies for Streaming Electrification............................................................. 56
2 A PRACTICAL APPROACH TO ASSESSING THE RISK OF FAILURE OF POWER
TRANSFORMERS ................................................................................................................................ 59
2.1 BACKGROUND......................................................................................................................... 59
2.2 LIFE MANAGEMENT PROCESS ................................................................................................... 59
2.2.1 Risk Assessment .............................................................................................................. 60
2.2.2 Layout of the Evaluation Procedure .................................................................................. 63
2.2.3 Evaluation Procedure........................................................................................................ 64
2.2.4 Probability of Failure – Individual Failure Rate................................................................... 66
2.3 ASSESSMENT OF THE TECHNICAL RISK OF FAILURE BY CATEGORY (MTMPTM PROGRAM) ............... 67
2.3.1 Mechanical Aspects.......................................................................................................... 67
2.3.2 Thermal Aspects............................................................................................................... 67
2.3.3 Electric Aspects - Risk of Dielectric Failure........................................................................ 67
2.3.4 Aspects Related to Accessory Failure ............................................................................... 67
2.3.5 Total Technical Risk of Failure .......................................................................................... 68
2.4 RISK MITIGATION .................................................................................................................... 70
2.5 SUMMARY .............................................................................................................................. 70
3 DIAGNOSIS OF TRANSFORMERS.............................................................................................. 71
3.1 DIAGNOSTICS METHODS FOR POWER TRANSFORMERS AND ACCESSORIES .................................. 71
3.1.1 Diagnostic Methods for Power Transformers..................................................................... 71
3.1.1.1 Stresses Acting on Power Transformers ................................................................................ 72
3.1.1.2 Deterioration Factors and Failure Mechanisms....................................................................... 73
3.1.1.3 Diagnostic Methods............................................................................................................... 73
3.1.2 Diagnostic Methods for Bushings ...................................................................................... 74
3.1.2.1 Stresses Acting on Bushings ................................................................................................. 75
3.1.2.2 Deterioration Factors and Failure Mechanisms....................................................................... 75
3.1.2.3 Diagnostic Methods............................................................................................................... 76
3.1.3 Diagnostic Methods for Surge Arresters............................................................................ 76
3.1.3.1 Stresses Acting on Surge Arresters ....................................................................................... 77
3.1.3.2 Deterioration Factors and Failure Mechanisms....................................................................... 77
3.1.3.3 Diagnostic Methods............................................................................................................... 78
3.2 GENERAL DIAGNOSIS TOOLS ..................................................................................................... 79
3.2.1 Oil Quality Assessment..................................................................................................... 79
3.2.1.1 Factors Affecting the Health and Life of Power Transformers ................................................. 79
3.2.1.2 Methods for Assessing the Quality of Transformer Oils........................................................... 80
3.2.1.2.1 Dielectric Breakdown Strength (BDV)................................................................................ 80
3.2.1.2.2 Interfacial Tension (IFT).................................................................................................... 80
3.2.1.2.3 Acid Neutralization Number .............................................................................................. 81
3.2.1.2.4 Power Factor.................................................................................................................... 82
3.2.1.2.5 Test for Oxygen Inhibitor................................................................................................... 82
3.2.1.2.6 Furan Analysis ................................................................................................................. 82
3.2.1.2.7 PCB Content .................................................................................................................... 83
3.2.1.2.8 Corrosive Sulphur............................................................................................................. 83
3.2.1.3 Moisture in Transformer Insulation Systems .......................................................................... 83
3.2.1.3.1 Transformer Oil ................................................................................................................ 84
3.2.1.3.2 Relative Humidity ............................................................................................................. 84
3.2.1.3.3 Paper (Cellulose).............................................................................................................. 85
3.2.1.3.4 Where Does the Water Come From .................................................................................. 86
3.2.1.3.5 Moisture Equilibrium between Oil and Paper in Transformers............................................. 86

viii
 3.2.1.3.6 Cautions in Estimation of Moisture Using Moisture Equilibrium Curves ............................... 88
3.2.1.4 Limits for Measurement Oil Quality Parameters ..................................................................... 89
3.2.1.5 Moisture and Bubble Evolution in Transformers ..................................................................... 92
3.2.2 Dissolved Gas in Oil Analysis (DGA) ................................................................................ 96
3.2.2.1 Introduction........................................................................................................................... 96
3.2.2.2 Procedure............................................................................................................................. 97
3.2.2.3 Sampling .............................................................................................................................. 97
3.2.2.4 Extraction ............................................................................................................................. 97
3.2.2.5 Analysis................................................................................................................................ 97
3.2.2.6 Interpretation ........................................................................................................................ 99
3.2.2.7 Air ........................................................................................................................................ 99
3.2.2.8 Gas Spectrum – Types of Faults............................................................................................ 99
3.2.2.8.1 Hot Metal Surface............................................................................................................. 99
3.2.2.8.2 Examples of Hot Metal Surfaces ....................................................................................... 99
3.2.2.9 Overheated cellulose .......................................................................................................... 100
3.2.2.9.1 Examples of Overheated Cellulose ................................................................................. 100
3.2.2.10 Electrical Faults .................................................................................................................. 100
3.2.2.10.1 Examples of Electrical Faults......................................................................................... 100
3.2.2.11 Factors affecting gas concentration in transformers.............................................................. 101
3.2.2.11.1 Type and Brand of Oil ................................................................................................... 101
3.2.2.11.2 Oxygen......................................................................................................................... 101
3.2.2.11.3 Load............................................................................................................................. 101
3.2.2.11.4 Oil Preservation Systems .............................................................................................. 101
3.2.2.11.5 Gas Mixing ................................................................................................................... 102
3.2.2.11.6 Temperature................................................................................................................. 102
3.2.2.11.7 Gas Solubility in Oil....................................................................................................... 103
3.2.2.11.8 Other Factors................................................................................................................ 104
3.2.2.12 DGA Interpretation Methods................................................................................................ 106
3.2.2.12.1 Key Gas Method of Interpreting DGA............................................................................. 106
3.2.2.12.2 Individual and Total Dissolved Key-Gas Concentration Method ...................................... 107
3.2.2.12.3 Rogers Ratio Method .................................................................................................... 110
3.2.2.12.4 IEC Method .................................................................................................................. 112
3.2.2.12.4.1 Carbon Dioxide/Carbon Monoxide (CO2/CO) Ratio .........................................................112
3.2.2.12.4.2 IEC C2H2/H2 Ratio ..............................................................................................................113
3.2.2.12.4.3 IEC Recommended Method of Interpretation ...................................................................113
3.2.2.12.5 Duval Triangle Method for Diagnosing a Transformer Problem Using
Dissolved Gas Analysis ................................................................................................ 114
3.2.2.12.6 ABB's Advanced Dissolved Gas Analysis Software (ADGA) ........................................... 117
3.2.3 Analysis of Particles in Transformer Oils ........................................................................ 118
3.2.3.1 Oil Sampling for Particle Analysis ........................................................................................ 118
3.2.3.2 Particle Counting ................................................................................................................ 118
3.2.3.2.1 Normal and Abnormal Particle Count Levels.................................................................... 119
3.2.3.3 Trace Metal Content of Particles.......................................................................................... 120
3.2.3.3.1 Method of Measurement ................................................................................................. 120
3.2.3.3.2 Normal and Abnormal Metallic Content of Particles in Oil................................................. 120
3.2.3.4 Diagnostic Examples of Particle Analysis............................................................................. 121
3.2.3.5 Effect of particles on dielectric strength of insulating oil ....................................................... 122
3.2.3.5.1 Current filtering practices on new transformers ................................................................ 122
3.2.3.5.2 Classification of contamination level ................................................................................ 123
3.2.3.5.2.1 Bare electrodes ....................................................................................................................123
3.2.3.5.2.2 Covered electrodes ..............................................................................................................123
3.2.3.5.3 Contamination deposited on insulating surface................................................................ 124
3.2.3.5.4 Recommended corrective action..................................................................................... 125
3.2.4 Winding Resistance Test ................................................................................................ 126
3.2.4.1 Measurement Method for Winding Resistance Measurement................................................ 126
3.2.5 Transformer Turns Ratio Test (TTR) ............................................................................... 128
3.2.6 Insulation resistance ....................................................................................................... 131
3.2.6.1 Measurement...................................................................................................................... 131
3.2.6.2 Interpretation ...................................................................................................................... 132
3.2.6.3 Polarization Index ............................................................................................................... 133
3.2.7 Insulation Power Factor Tests......................................................................................... 134

ix
 3.2.7.1 Two-Winding Transformer ................................................................................................... 135
3.2.7.1.1 Testing of Two-Winding Transformers............................................................................. 136
3.2.7.2 Three-Winding Transformer................................................................................................. 139
3.2.7.3 Typical Insulation Power Factor Values................................................................................ 140
3.2.7.4 General Guidelines for Assessing Power Factor Values ....................................................... 141
3.2.7.5 Power Factor Tip-up Tests .................................................................................................. 141
3.2.8 Core Insulation Resistance Measurement ....................................................................... 142
3.2.8.1 Measurement and Diagnosis of Inadvertent Core Grounds................................................... 142
3.2.9 Excitation Current Tests.................................................................................................. 144
3.2.9.1 Measurement Setup............................................................................................................ 145
3.2.9.2 Analysis of Excitation Current Results.................................................................................. 148
3.2.10 Infrared Thermography Analysis of Transformers and Accessories ............................. 149
3.2.10.1 The Thermography Process ................................................................................................ 149
3.2.10.2 Criteria for Evaluating Infrared Measurements ..................................................................... 150
3.2.10.3 Example Uses of Infrared Thermography diagnostics on Power Transformers ..................... 150
3.2.10.3.1 Loose connection at bushing outlet terminal................................................................... 150
3.2.10.3.2 Blocked oil flow in radiators or radiator shut off .............................................................. 151
3.2.10.3.3 LTC overheating ........................................................................................................... 151
3.2.10.3.4 Low oil level in transformer or bushing ........................................................................... 152
3.2.10.3.5 Moisture contamination of surge arrester ....................................................................... 152
3.2.11 Bushings .................................................................................................................... 153
3.2.11.1 ANSI & IEC – Common Diagnostic Tools............................................................................. 153
3.2.11.1.1 Oil leakage inspection................................................................................................... 153
3.2.11.1.2 Insulator inspection and cleaning................................................................................... 153
3.2.11.1.2.1 Porcelain insulators ............................................................................................................153
3.2.11.1.2.2 Silicon rubber insulators.....................................................................................................153
3.2.11.1.3 Thermovision................................................................................................................ 153
3.2.11.1.4 Oil sampling from bushing ............................................................................................. 154
3.2.11.1.5 Dissolved Gas Analysis (DGA) ...................................................................................... 156
3.2.11.1.6 Moisture analysis .......................................................................................................... 156
3.2.11.1.7 Dielectric Frequency Response Analysis (DFRA)........................................................... 157
3.2.11.1.8 Partial Discharge measurements................................................................................... 157
3.2.11.1.9 De-polymerization analysis............................................................................................ 157
3.2.11.2 Diagnostics techniques for bushings complying with the ANSI/IEEE Standards..................... 158
3.2.11.2.1 Condenser Bushing Power Factor Tests........................................................................ 158
3.2.11.2.2 Factors Affecting C1 and C2 Capacitance and Power Factor Measurements .................. 159
3.2.11.2.3 Bushing Hot Collar Test ................................................................................................ 162
3.2.11.2.4 What to do when Bushing Power Factor Tests are Doubtful............................................ 164
3.2.11.2.5 Special Case – Type “U” Bushings ............................................................................... 164
3.2.11.2.5.1 History.................................................................................................................................164
3.2.11.2.5.2 Recommendation ...............................................................................................................170
3.2.11.2.6 Type “T” Bushings......................................................................................................... 173
3.2.11.3 Diagnostics and Conditioning on ABB Bushings Complying with the IEC Standard................ 174
3.2.11.3.1 Capacitance and tan measurement.............................................................................. 175
3.2.11.3.2 Temperature correction................................................................................................. 175
3.2.12 Measurements for Assessing the Condition of OLTCs/LTCs ....................................... 178
3.2.12.1 Number of Operations......................................................................................................... 178
3.2.12.2 Resistance of the Electrical Connections ............................................................................. 178
3.2.12.3 Temperature....................................................................................................................... 178
3.2.12.4 Motor Current ..................................................................................................................... 178
3.2.12.5 Acoustic Signal ................................................................................................................... 178
3.2.12.6 Relay Timing....................................................................................................................... 179
3.2.12.7 Gas-in-Oil Analysis ............................................................................................................. 179
3.2.12.7.1 Items Specific to the European Practice........................................................................ 179
3.2.12.7.1.1 Scope ..................................................................................................................................179
3.2.12.7.1.2 History.................................................................................................................................179
3.2.12.7.1.3 Faults in OLTCs possible to indicate by DGA...................................................................179
3.2.12.7.1.4 The Stenestam ratio...........................................................................................................180
3.2.12.7.1.5 Important principals for interpretation of DGAs in OLTC .................................................180
3.2.12.7.1.6 Interpreting the Stenestam ratio ........................................................................................180
3.2.12.7.1.7 Typical gas concentrations ................................................................................................181

x
 3.2.12.7.2 Important to bear in mind .............................................................................................. 182
3.2.12.7.3 North-American Practice ............................................................................................... 182
3.2.12.8 Moisture ............................................................................................................................. 183
3.3 ADVANCED DIAGNOSTIC TOOLS .............................................................................................. 184
3.3.1 Assessment of Mechanical Properties - Frequency Response Analysis (FRA) ................ 184
3.3.1.1 Introduction......................................................................................................................... 184
3.3.1.1.1 Purpose of FRA measurements ...................................................................................... 184
3.3.1.1.2 When should FRA measurements be performed?............................................................ 184
3.3.1.2 Standards........................................................................................................................... 185
3.3.1.3 General description of the FRA method ............................................................................... 185
3.3.1.3.1 Principle of the measurement.......................................................................................... 185
3.3.1.3.2 Practical set-up .............................................................................................................. 186
3.3.1.4 Commercial equipment ....................................................................................................... 187
3.3.1.5 Detailed measurement procedure........................................................................................ 187
3.3.1.5.1 Test preparation ............................................................................................................. 188
3.3.1.5.2 Tap changer position ...................................................................................................... 188
3.3.1.5.3 Treatment of un-tested terminals..................................................................................... 189
3.3.1.5.4 Test leads: ..................................................................................................................... 189
3.3.1.5.5 Test Set-up .................................................................................................................... 189
3.3.1.6 Reporting of FRA measurements......................................................................................... 192
3.3.1.6.1 General information: ....................................................................................................... 192
3.3.1.6.2 Transformer information:................................................................................................. 192
3.3.1.6.3 Description of each measurement: .................................................................................. 192
3.3.1.6.4 Instrumentation: ............................................................................................................. 193
3.3.1.6.5 Cabling: ......................................................................................................................... 193
3.3.1.7 Basic interpretation and on-site quality check....................................................................... 193
3.3.1.7.1 Some “normal” FRA spectra ........................................................................................... 194
3.3.1.7.2 Meaning of different frequency ranges in an FRA spectrum ............................................. 197
(A) When only the current FRA measurement data are available:....................................................... 197
3.3.1.7.3 Comparison between open- and short-circuit measurements ........................................... 197
3.3.1.7.4 Comparison between high- and low-voltage windings ...................................................... 197
3.3.1.7.5 Comparison between phases.......................................................................................... 197
(B) When further data are available................................................................................................... 198
3.3.1.7.6 Comparison with historical data....................................................................................... 198
3.3.1.7.7 Comparison with twin or sister units ................................................................................ 198
3.3.1.7.8 History of the unit ........................................................................................................... 198
3.3.1.7.9 Other diagnostic data...................................................................................................... 199
3.3.1.8 Examples of problems diagnosed using FRA ....................................................................... 199
3.3.1.8.1 Axial Winding Collapse................................................................................................... 199
3.3.1.8.2 Hoop Buckling................................................................................................................ 200
3.3.1.8.3 Shorted Turns ................................................................................................................ 202
3.3.2 Assessment of Thermal Properties ................................................................................. 204
3.3.2.1 Degree of Polymerization (DP) ........................................................................................... 204
3.3.2.1.1 DP versus Life Plots ....................................................................................................... 204
3.3.2.1.2 Latest Research Findings on DP Analysis ....................................................................... 207
3.3.2.2 Furanic Compound Analysis................................................................................................ 207
3.3.2.2.1 Origin of Furanic Compounds ......................................................................................... 207
3.3.2.2.2 Detection of Furanic Compounds .................................................................................... 208
3.3.2.2.3 Correlation Curves of Furanic Content with DP................................................................ 208
3.3.2.2.4 Issues to Consider in Using Furan Analysis..................................................................... 209
3.3.3 Dielectric Frequency Response as a Tool for Troubleshooting Insulation
Power Factor Problems .................................................................................................. 211
3.3.3.1 Introduction......................................................................................................................... 211
3.3.3.2 Dielectric frequency response and X-Y model ...................................................................... 211
3.3.3.3 Causes of High Power Factor in Transformer Insulation ....................................................... 214
3.3.3.3.1 Comparison of DFR to Power Factor Measurement ......................................................... 214
3.3.3.3.2 Influence of Oil Conductivity and Moisture on PF and DFR .............................................. 215
3.3.3.4 Dielectric Frequency Response Signature and Identification Techniques .............................. 216
3.3.3.4.1 Identification of high Core-Grounding Resistance Problems ............................................. 217
3.3.3.4.2 Identification of Paper Contamination Problems............................................................... 220
3.3.3.4.3 Low Temperature Effect on Insulation Power Factor........................................................ 220
3.3.3.5 Summary............................................................................................................................ 222

xi
 3.3.4 Assessment of Electrical Properties - Partial Discharge Measurements .......................... 223
3.3.4.1 Purpose of measurement .................................................................................................... 223
3.3.4.2 Electrical PD Measurement on Transformers ....................................................................... 224
3.3.4.2.1 Calibration...................................................................................................................... 225
3.3.4.2.2 PD measuring procedure................................................................................................ 226
3.3.4.2.3 An Advanced PD system ................................................................................................ 226
3.3.4.3 Procedure for Investigation of PD Sources........................................................................... 228
3.3.4.4 Acoustical Partial Discharge Measurement on Transformers ................................................ 233
3.3.4.4.1 Acoustic PD Wave Characterization................................................................................ 233
3.3.4.4.2 Acoustic Partial Discharge Localization ........................................................................... 235
4 FAULT ANALYSIS ..................................................................................................................... 237
4.1 GUIDANCE FOR PERFORMING FAILURE ANALYSIS ....................................................................... 237
4.1.1 Introduction..................................................................................................................... 237
4.1.2 Failure definition ............................................................................................................. 239
4.1.3 Classification of failures .................................................................................................. 239
4.1.4 General information on malfunctions and failures ............................................................ 240
4.1.5 Systematic failure analysis.............................................................................................. 241
4.1.5.1 Collecting information on the unit concerned ........................................................................ 242
4.1.5.2 Data and information at the time of fault inception ................................................................ 243
4.1.5.3 Deciding on continued operation or additional investigations ................................................ 246
4.1.5.4 Assessment of the extent of damage on site ........................................................................ 247
4.1.5.5 Assessment of external damage on site............................................................................... 248
4.1.6 Diagnostic measurements and their interpretation ........................................................ 249
4.1.6.1 Routine measurements on site ............................................................................................ 250
4.1.6.1.1 Oil analysis .................................................................................................................... 250
4.1.6.1.2 Insulation resistance and tan ........................................................................................ 250
4.1.6.1.3 Measurement of transformer ratio ................................................................................... 250
4.1.6.1.4 Measurement of winding resistances............................................................................... 251
4.1.6.1.5 Measurement of short-circuit impedance ......................................................................... 251
4.1.6.1.6 Excitation at low voltage ................................................................................................. 252
4.1.6.2 Special diagnostic measurements ....................................................................................... 252
4.1.6.2.1 Gas-in-oil analysis .......................................................................................................... 252
4.1.6.2.2 Measurement of partial discharges.................................................................................. 254
4.1.6.2.3 FRA method................................................................................................................... 255
4.1.6.2.4 Measurement of polarization effects for assessing the moisture ....................................... 256
4.1.6.3 Inspection of core-and-coil assembly on site ........................................................................ 256
4.1.6.3.1 General preconditions..................................................................................................... 256
4.1.6.3.2 Safety precautions.......................................................................................................... 257
4.1.6.3.3 Checks to be conducted ................................................................................................. 257
4.1.6.4 Dismantling the defective transformer .................................................................................. 258
4.1.6.4.1 Preconditions ................................................................................................................. 258
4.1.6.4.2 Inspection ...................................................................................................................... 259
4.1.6.4.3 Inspection of the core-and-coil assembly after lifting out of the tank.................................. 259
4.1.6.4.4 Inspection of the windings............................................................................................... 260
4.1.6.5 Typical fault patterns of windings ......................................................................................... 260
4.1.6.5.1 Short-circuit faults........................................................................................................... 260
4.1.6.5.2 Electrical flashover ......................................................................................................... 261
4.1.6.5.3 Thermal faults ................................................................................................................ 263
4.1.6.6 Inspection of the core and the tank ...................................................................................... 263
4.1.7 Final assessment of the failure and the fault.................................................................... 264
4.1.8 Case Studies .................................................................................................................. 265
4.1.8.1 Case 1: Examination of a transformer affected by partial discharges.................................... 265
4.1.8.2 Case 2: Analysis of a failure caused by overvoltages at no-load switching operation ........... 275
4.1.8.3 Case 3: Fault analysis on a generator step-up Transformer following an internal flashover . 279
5 ONLINE DIAGNOSTIC MONITORS FOR TRANSFORMERS AND KEY ACCESSORIES .......... 284
5.1 POWER TRANSFORMER (TANK & CORE) .................................................................................. 284
5.2 LOAD TAP CHANGER ............................................................................................................. 285
5.3 BUSHING & CT ..................................................................................................................... 285

xii
 5.4 EXAMPLE MONITORING SYSTEMS ........................................................................................... 286
6 PREVENTIVE MAINTENANCE OF TRANSFORMERS............................................................... 294
6.1 BASIC AGEING PROCESSES .................................................................................................... 294
6.1.1 Introduction..................................................................................................................... 294
6.1.2 Paper Degradation.......................................................................................................... 295
6.1.3 On-site Drying Methods .................................................................................................. 298
6.1.3.1 Traditional methods............................................................................................................. 298
6.1.3.2 On-site drying with low frequency heating (LFH) in combination with hot-oil spray ................ 299
6.1.4 Oil reclaiming.................................................................................................................. 300
6.1.4.1 Online oil reclaiming technology ......................................................................................... 300
6.1.4.2 Comparison with oil change................................................................................................. 300
6.1.4.3 Long- term stability.............................................................................................................. 300
6.2 GENERAL MAINTENANCE OF TRANSFORMERS .......................................................................... 302
6.2.1 Recommended schedule of Maintenance activities ......................................................... 302
6.2.1.1 Monthly Maintenance Schedule........................................................................................... 302
6.2.1.2 Quarterly Maintenance Schedule......................................................................................... 303
6.2.1.3 Annual Maintenance Schedule with The Transformer De-energized ..................................... 304
6.2.2 Maintenance of Components .......................................................................................... 305
6.2.2.1 Transformer liquid and insulation......................................................................................... 305
6.2.2.2 Bushings and joints............................................................................................................. 306
6.2.2.3 Off-load tap changer (DETC)............................................................................................... 306
6.2.2.4 On-load tap changer ........................................................................................................... 307
6.2.2.5 Motor drive unit ................................................................................................................... 307
6.2.2.6 Oil filtering unit.................................................................................................................... 307
6.2.2.7 Coolers............................................................................................................................... 307
6.2.2.8 Liquid conservator with rubber diaphragm (COPS)............................................................... 307
6.2.2.9 Gaskets.............................................................................................................................. 307
6.2.2.10 Surface protection............................................................................................................... 308
6.2.2.10.1 Painted surfaces ........................................................................................................... 308
6.2.2.10.2 Zinc coated surfaces ..................................................................................................... 308
6.2.3 Investigation of Transformer Disturbances ...................................................................... 308
6.2.3.1 Recording of disturbances................................................................................................... 308
6.2.3.2 Fault localizations advice for oil-immersed transformers ....................................................... 309
6.2.4 Internal Inspection .......................................................................................................... 312
6.2.4.1 Opening the Transformer .................................................................................................... 312
6.2.4.2 The Inspection .................................................................................................................... 313
6.2.4.3 Electrical Tests ................................................................................................................... 314
6.2.5 Maintenance of Bushings................................................................................................ 315
6.2.6 Maintenance and Service of OLTCs/LTCs ..................................................................... 317
6.2.7 General Quality Information for Various Types of LTCs................................................... 318
6.2.7.1 North-American Practices.................................................................................................... 318
6.2.7.2 European Practice .............................................................................................................. 322
7 REPAIR, REFURBISHMENT AND RETROFIT ........................................................................... 324
7.1 PREPARATION PHASE ............................................................................................................ 325
7.2 UNTANKING AND DISASSEMBLY OF ACTIVE PART ....................................................................... 326
7.3 REPAIR OF THE TRANSFORMER ............................................................................................... 327
7.4 ASSEMBLY AND TANKING OF THE ACTIVE PART .......................................................................... 328
7.5 DRYING ................................................................................................................................ 328
7.6 FINAL ASSEMBLY ................................................................................................................... 329
7.7 HIGH VOLTAGE TESTING ........................................................................................................ 329
7.8 QUALITY PLAN ..................................................................................................................... 330
7.9 FACILITIES FOR SITE REPAIR .................................................................................................. 330
7.9.1 Temporary Workshops.................................................................................................... 331
7.9.1.1 Steel Buildings.................................................................................................................... 331
7.9.1.2 Large Tents ........................................................................................................................ 331
7.9.1.3 Foundation for a Temporary Workshop................................................................................ 332
7.9.2 Facilities for Heavy Lifting ............................................................................................... 332

xiii
 7.9.3 Moisture control .............................................................................................................. 332
7.9.4 Oil processing................................................................................................................. 333
7.9.5 Drying equipment ........................................................................................................... 333
7.9.6 High voltage test equipment............................................................................................ 333
8 ENVIRONMENTAL ASPECTS ................................................................................................... 334
8.1 CONTAMINATION OF OILS WITH PCB (POLYCHLORINATED BIPHENYLS) ....................................... 334
8.1.1 General .......................................................................................................................... 334
8.1.2 Dehalogenation Processes Using Sodium and Lithium Derivatives.................................. 335
8.1.3 Dehalogenation Processes Using Polyethyleneglycol and Potassium Hydroxide ............. 335
8.1.4 Dehalogenation in Continuous Mode by Closed Circuit Process...................................... 335
8.2 ELECTROMAGNETIC COMPATIBILITY (EMC) ............................................................................. 335
8.2.1 Introduction..................................................................................................................... 335
8.2.2 Methods to Reduce EMF Levels in Existing Substations ................................................. 336
8.3 AUDIBLE NOISE ..................................................................................................................... 336
8.3.1 Introduction..................................................................................................................... 336
8.3.2 Background .................................................................................................................... 337
8.3.2.1 Characteristics of Transformer Noise................................................................................... 337
8.3.2.2 Propagation of Sound ......................................................................................................... 337
8.3.3 Criteria for Community Noise Levels ............................................................................... 337
8.3.4 Requirements ................................................................................................................. 338
8.3.5 Methods of Substation Noise Control .............................................................................. 338
8.4 RELEASE OF INSULATING OIL.................................................................................................. 340
8.4.1 Introduction..................................................................................................................... 340
8.4.2 Use of Synthetic Ester .................................................................................................... 340
8.4.3 Use of Natural Ester ...................................................................................................... 341
9 ECONOMICS OF TRANSFORMER ASSET MANAGEMENT ..................................................... 342
9.1 FAILURE STATISTICS FOR POWER TRANSFORMERS................................................................... 342
9.1.1 CIGRE Survey of Failures in Large Power Transformers ................................................ 342
9.1.2 Canadian Electricity Association Forced Outage Report ................................................. 344
9.2 ECONOMICS OF TRANSFORMER MANAGEMENT FOR FLEETS AND SPECIFIC UNITS ........................... 347
9.2.1 Introduction..................................................................................................................... 347
9.2.2 General Concept for Economics of Transformer Management......................................... 348
9.2.3 Description of the Simulation Model ................................................................................ 349
9.2.4 Case Study by a Utility.................................................................................................... 350
9.2.5 Conclusions.................................................................................................................... 352
10 HEALTH AND SAFETY ASPECTS / RECOMMENDATIONS ..................................................... 353
10.1 PREAMBLE ........................................................................................................................... 353
10.2 INTRODUCTION ..................................................................................................................... 353
10.3 SCOPE ................................................................................................................................. 353
10.4 DEFINITIONS ......................................................................................................................... 354
10.5 SAFETY MANAGEMENT .......................................................................................................... 355
10.6 DOCUMENTATION .................................................................................................................. 355
10.7 ELECTRICAL SAFETY RULES ................................................................................................... 356
10.7.1 General Rules ............................................................................................................ 356
10.7.2 Communication and Control Rules.............................................................................. 356
10.7.3 Rules for working on dead Electrical Equipment.......................................................... 357
10.7.4 Rules for working on or very near live Electrical Equipment......................................... 361
10.7.5 Switching.................................................................................................................... 361
10.7.6 Work on or very near live conductors .......................................................................... 361
10.7.7 Testing and Commissioning........................................................................................ 362
10.8 W ORK AT HEIGHT: ADDITIONAL SAFETY EQUIPMENT FOR POWER TRANSFORMERS. ...................... 362
10.8.1 “NO-RISK SYSTEM”................................................................................................... 362
10.8.2 “Fall Arrest Towers and Base Plates”.......................................................................... 365

xiv
 10.9 APPENDICES......................................................................................................................... 365
10.9.1 Appendix 1 - Minimum working clearance ................................................................... 366
10.9.2 Appendix 2 - Minimum design clearances where power lines cross or are in close
proximity 368
10.9.3 Appendix 3 - Minimum separation across point of disconnection in air......................... 369
10.9.4 Appendix 4 - Principles of Risk Assessment................................................................ 370
10.9.5 Appendix 5 - Example of Sample Risk Assessment Sheet. ......................................... 371
10.9.6 Appendix 6 - Electrical Job Hazard Analysis Sheet. .................................................... 372
10.9.7 Appendix 7 - Sample Safety Check Sheet................................................................... 374
10.9.8 Appendix 8 - Sample Safety Permit to work. ............................................................... 376
10.9.9 Appendix 9 - Sample Energized Electrical Work Permit............................................... 378
REFERENCES.................................................................................................................................... 379
INDEX................................................................................................................................................. 390
ABB TRANSFORMERS SERVICE GENERAL BROCHURES ........................................................... 395
ABB TRANSFORMERS SERVICE PRODUCT LEAFLETS................................................................ 405
ABB TRES NORTH AMERICA SERVICE BROCHURES.................................................................... 423
CONTACT LIST FOR MAIN ABB SERVICE CENTERS...................................................................... 454

xv
1 TRANSFORMER DESIGN CONSIDERATIONS [1]
1.1 CONFIGURATION

There are two basic configurations for power transformers: core form and shell form. The
principal physical difference between the two constructions is related to the geometry of
the magnetic circuit and the position, alignment, and types of the windings employed for
each design.

Fundamentally, for the shell form designs, the magnetic circuit forms a shell around a
major portion of the windings. Three phase shell form designs use 4 and 7 limb cores with
the usual horizontal orientation of the core limbs. Shell form 7 limb cores are used on
newer shell form designs due to lower weight, manufacturing simplicity, and lower core
loss. Single phase shell form transformers use 3 limb cores. In the shell form design, the
windings are interleaved; that is, the high-voltage and low-voltage windings are
subdivided into groups with the groups adjacent to each other in the axial (horizontal)
direction. Each group is assembled using interconnected rectangular pancake coils.

In core form designs the magnetic circuit forms a core through the windings. Three phase
core form transformers are usually constructed with a three limb core that has the center
limbs vertically oriented with the top and bottom yokes for main flux return paths oriented
horizontally. When shipping height becomes a limiting design factor, a five limb core may
be used to keep the shipping height within the shipping limitation. This configuration
enables the yoke depth to be reduced by providing a return flux path external to the
wound limbs. The only other occasion in which a three-phase, five limb core might be
necessary is when it is required to provide a value of zero sequence impedance of similar
magnitude to the positive sequence impedance. The core form single phase geometry
uses 2, 3, or 4 limb cores. Generally, the core form design uses several types of circular
coils (layer, helical, disc) that are concentric with each other and the vertical core limb.

For power transformers, there will be design requirements where one form of construction
will have an advantage over the other. The major parametric elements of the comparison
are MVA size, voltage class, impedance requirements, and loss performance
characteristics. ABB has the flexibility in design knowledge and manufacturing capability
to produce either construction.

1.2 MECHANICAL CONSIDERATION

The mechanical design of a transformer involves the analysis and determination of the
expected operational forces, the structural stress analysis of the insulation system and
support elements, and the proper choice of materials. A transformer must be strong
enough to withstand the mechanical stresses imposed by system-related events such as
short circuits. The mechanical stresses developed during normal operation are low, but
the stresses generated by a system short circuit event can be quite large. Also, the

17
magnitude of these stresses increase with the size and complexity of the transformer. The
majority of the mechanical stresses must be taken by the insulation system, which is
primarily composed of cellulose-based materials. These materials are weakest in bending
and tension. It is therefore best to apply these materials in compression. Also, to keep the
total forces as low as possible, the design of the windings should be made using the best
arrangement and overall geometry of the individual windings.

1.3 THERMAL CONSIDERATIONS

Temperature is one of the most important factors affecting transformer life. As the
temperature of the insulation increases, the insulation life decreases. The transformer
must be designed to operate within the guaranteed temperature parameters and the
prescribed standard allowances to ensure long transformer life. In an oil-filled
transformer, the insulating oil is used to conduct the heat away from the windings and
the magnetic core. To perform this function, the oil must circulate through the winding
assembly and usually through externally applied cooling apparatus.

For thermosiphon oil flow (natural oil flow), oil circulation is created when the weight of the
column of oil in the cooling equipment is greater than the weight of the column of oil in the
core and coil assembly. Also, the center of cooling must be above the center of heating.
This distance has a direct affect on the top-to-bottom temperature difference – the larger
the distance between the center of cooling and the center of heating, the larger the oil
flow and the lower the top-to-bottom temperature difference. This configuration is defined
in the standards as ONAN (oil natural, air natural) – the old nomenclature was OA).

Additional transformer capacity can be created by adding auxiliary cooling equipment,

such as fans. Fans increase the airflow over the external cooling equipment without
changing the mode of internal oil flow. Fans can be added in one or two stages. Using
the ONAN rating as the base or 100 % rating, a rating of 133 % can be attained by
adding one stage of fans. Additional fans (2nd stage), usually equal in number to the
first stage of fans, can be added to obtain a rating of 167 %. The energizing of the fan
stages is normally controlled by temperature-actuated contacts provided in the winding
temperature device. The current industry designations for fans-only auxiliary cooling
with natural oil flow are defined in the standards as:
ONAN/ONAF (oil natural, air natural/oil natural, air forced – 100 %/133 %) – the
old designation was OA/FA
ONAN/ONAF/ONAF – (oil natural, air natural/oil forced, air forced/ oil forced, air
forced – 100 %/133 %/167 %) – the old designation was OA/FA/FA

For larger transformer ratings, some design configurations may require the addition of oil
circulating pumps to meet the required temperature rise guarantees. With the addition of
oil circulating pumps, the top-to-bottom oil temperature difference attained by the forced
oil flow is usually in the order of single digits. The increased oil flow is usually
accompanied by internal means to direct the oil flow through the windings; this is
generally known as directed flow. When two stages of auxiliary cooling are employed, the
equipment is generally divided equally among the two stages. The designation for cooling

18
with auxiliary fans and pumps is defined in the standards as (past nomenclature shown in
parentheses):

ONAN/ODAF (oil natural, air natural/oil directed, air forced – 100 %/133 %) – the
old designation was OA/FOA.
ONAN/ODAF/ODAF (oil natural, air natural/oil directed, air forced/ oil directed, air
forced – 100 %/133 %/167 %) – the old designation was OA/FOA/FOA.

Other configurations for the use of auxiliary fans and pumps are sometimes applied, such
as using fans only for the first rating increase and energizing all of the pumps for the
second stage of cooling.

Additionally, transformers can be designed with a single rating that uses auxiliary cooling
equipment consisting of oil circulating pumps with an associated oil-to-air heat exchanger
or forced oil with a water-cooled heat exchanger.

1.4 DIELECTRIC CONSIDERATIONS

The transformer insulation system must be designed to withstand the normal operating
voltages as well as over-voltages during lightning events, system short circuits, and
system switching surges. In addition, consideration must be given to produce
transformers that withstand these voltages with all elements operating below the corona
onset voltage.

A transformer is a simple inductance when considering low frequency operating voltages

and over- voltages. However, to an impulse voltage, the transformer presents a complex
combination of inductances and capacitances. Initially, when an impulse voltage impinges
upon a transformer winding, the initial distribution is determined by the winding coil-to-coil
and coil-to-ground capacitances. The final voltage distribution is ultimately distributed in
line with the winding coil inductances. For many transformers, the initial distribution of an
impulse voltage is less than perfect. This results in increased stress at the line end of the
winding. There are several solutions for these increased stresses. For the lower voltage
ratings, the usual method is to accept the higher stress and insulate accordingly. For
higher voltage ratings, there are a number of winding arrangements, conductor
interleaving schemes, and electrostatic shielding methodologies that are employed to
reduce the voltage stresses produced at the line end of the winding.

1.5 CONSTRUCTION TYPES

1.5.1 SHELL FORM

1.5.1.1 DESIGN FEATURES

The ABB Shell Form-Form Fit design features a rectangular shaped coil system made up
of a series of inter-connected pancake coils. The coil and insulation assembly is mounted
vertically in the tank bottom section. The core is positioned horizontally around the
outside of the winding and acts as a protective shell around the coil. The upper section

19
of the tank fits snugly over the core and coils to form a unit assembly with the
mechanical support completely outside the winding (see Figure 1-1).
The heat generated by the core and coils is dissipated by the circulation of the oil. The
oil flow from the bottom to the top of the tank is supported by the temperature
differential or thermal head during self-cooled operation. The addition of pumps and fans
for forced cooling will increase the flow of oil through the core and coils and the flow of
air through the heat exchanger. With either mode of cooling, the oil passes through a
heat exchanger where it cools prior to reentering the tank at the bottom.
The shell form insulation system consists of high dielectric strength pressboard sheets
and precisely located oil spaces designed to control voltage stress concentration.

Figure 1-1: Partial Cutaway of a Shell Form Transformer

1.5.1.2 MECHANICAL STRENGTH

The coils in a shell form design are large surface area pancake coils, and they are
assembled into winding groups with their faces adjacent to flat pressboard washers
which contain a planned pattern of spacer blocks cemented to the surface. The spacer
blocks provide a uniform support system to the turns and strands of the individual coils.
The complete phase is installed vertically in the tank bottom, and the core is stacked
around it. The upper section of the tank is fitted snugly over the core and shimmed with
vertical wooden slits spaced around the periphery of the core.

The total force between transformer winding groups varies as the square of the ampere
turns per group. If the current during fault conditions is ten times the normal load
current, the short circuit force will be one hundred times the normal load winding forces.
As transformers get larger, the ampere turns per winding group are reduced in a shell
form design by increasing the number of winding groups, or high-low spaces; thus
controlling the magnitude of the total force. Increasing the number of high-low spaces

20
does not increase the length of the average mean turn in a shell form winding; therefore,
it can be done economically.
The forces within successive winding groups in a shell form transformer are in opposite
directions. As they traverse the winding, the forces tend to cancel each other out. As a
result, the net total restraining force that must be applied external to the windings is only
the force corresponding to a single pair of winding groups (see Figure 1-2).

Figure 1-2: Section Through a Shell Form Winding Group with a High-Low Coil Configuration
(arrows illustrate mechanical forces)

In addition to the control of total force magnitude available in a shell form design, the
unit stresses on the winding insulation structures are kept at a low level. The major
winding force is perpendicular to the face of the pancake coils, and each coil is supported
by spacer blocks on its adjacent pressboard washers. Between spacers, the windings act
as uniformly loaded beams, and the total winding force is transmitted through the group
by compression of the spacer blocks. The shell form design uses large pancake coils;
thus a large number of spacer blocks are available to absorb the total force, and the unit
stresses in the pressboard are relatively low.
The total force magnitude in a shell form design can be reduced considerably with
multiple high-low coil arrangements. Even with this advantage it is essential to have a
rugged mechanical structure to withstand the ultimate forces encountered during thru-
fault conditions. In the ABB Shell Form-Form Fit design the major components of force
are taken by well-braced structures completely outside the winding. The close-fitting
Form Fit tank and the core assembly combine to restrain the total forces acting on the
winding. For the portions of the winding that are above and below the core, heavy steel
structural members welded to the tank provide the restraint for the forces. The bracing
structures are completely outside the winding and can be reinforced without any
compromises in winding design.
The ABB Shell Form-Form Fit design offers a combination of controlled maximum stress,
inherent stability, and high mechanical strength to withstand the forces produced by
system thru-faults. The use of the Form Fit tank as the major structural support makes

21
up to a 20 % reduction in total weight and as much as 40 % reduction in oil volume in
ABB shell form large power transformers possible (see Figure 1-3).

Figure 1-3: Partial Cutaway of a Shell Form Transformer Showing Support Structure for Core and
Coils

1.5.1.3 T HERMAL CAPABILITY

A transformer is a very efficient piece of apparatus; however, energy is generated by
losses in the core and coils during normal operation. This energy is in the form of heat,
which increases approximately as the square of the load current and must be dissipated
to prevent deterioration of the insulation system. The oil in the transformer serves as a
medium for transmitting this energy from the core and coils to a heat exchanger, where it
is dispersed to the atmosphere.
The HV and LV coils in an ABB Shell Form Transformer are arranged vertically in the
tank and pressboard insulation washers containing spacer blocks in a pre-designed
pattern are located on either side of each coil. The spacer block pattern provides ducts
on both sides of the conductor through which the oil travels from the bottom to the top of
the tank. The core in a shell form transformer is a stack of narrow-width steel punchings.
Oil flowing on both sides of the core adequately cools this area; therefore, oil ducts within
the magnetic circuit are not necessary.
The oil flow in the transformer tank during self-cooled operation is supported by the
temperature differential between the oil at the top and bottom of the tank. This
temperature differential, or thermal head, is approximately 12 °C for a shell form
transformer (see Figure 1-4).

22
 Figure 1-4: Partial Cutaway of a Shell Form Transformer Illustrating OA (Self-Cooled) Cooling
Action

As the load on a transformer increases, the energy generated by the losses in the coil
system will increase in proportion to the square of the increase in load. Forced cooling is
applied to dissipate this additional energy and allow the transformer to operate at the
increased load and within temperature guarantees.
ABB applies both pumps and fans for forced cooled ratings on shell form transformers.
The pumps augment the circulation of oil that exists due to the thermal head, and since
the coils are positioned vertically, no barriers are necessary to direct the oil flow. The
additional oil flow provided by the pumps virtually eliminates the oil temperature
differential in the transformer and reduces the winding hottest spot temperature as much
as 10°C. The fans direct the airflow over the heat exchanger at a high velocity, thus
improving energy transfer to the atmosphere.
The addition of fans alone to a typical radiator bank will significantly increase its energy
dissipation; fans used in conjunction with pumps to provide forced air and forced oil
cooling will further increase the cooling capability of the same radiator bank. The forced
cooling can be operated continuously for heavily loaded transformers, or it can be
actuated in stages as the load increases.
Forced oil-forced air cooling is the most efficient method of increasing the capacity of a
transformer. This method of cooling coupled with the inherent thermal characteristics of
the ABB Shell Form Transformer design offer the highest thermal capability in large
power transformers.

23
Figure 1-5: Partial Cutaway of a Shell Form Transformer Illustrating FOA (Forced-Cooled) Cooling
Action

1.5.1.4 DIELECTRIC CHARACTERISTICS

The effect of overvoltage and system surge conditions on the windings of a transformer
is determined by the characteristics of the particular coil and insulation system. As this
voltage surge enters the transformer winding, the initial voltage distribution will be
directly determined by the capacitance networks of the coil and winding system (see
Figure 1-6). Oscillations may develop as the surge progresses through the coil system,
which for certain designs may be amplified by the natural oscillation in these systems to a
value greater than the initial crest. This overvoltage condition may concentrate at some
point in the winding, such as the first several turns at the line end of the winding or
around a tap section, and stress the turn-to-turn insulation in these areas.

Figure 1-6: Equivalent Inductance-Capacitance Network of a Shell Form Winding Section

The coil assembly of an ABB Shell Form Transformer consists of a relatively few
“pancake” coils with a broad cross-sectional area and a narrow coil edge (see Figure
1-6). Since the capacitances between coils and from coil to ground are directly
proportional, respective to the cross-sectional area of the coil and the area of its edge,

24
the shell form coil system has a high coil-to-coil and a low coil-to-ground capacitance.
When the ratio of coil-to-coil capacitance to the coil-to-ground capacitance is high, as it is
in a shell form transformer design, the voltage distribution with rapidly rising voltage
surges is more nearly uniform.

Figure 1-7: Shell Form Transformer - Cross-Section of Line End Coils within the Core Iron

The turn-to-turn voltage stresses due to the initial application of the surges are thereby
reduced in the shell form design insulation system, and the succeeding oscillations
developed in the winding are also reduced. The large inherent capacitance of the shell
form design causes the natural period of the winding oscillation to be relatively long,
thus allowing the voltage surges to decay to a low value before the winding
oscillations can develop to a significant magnitude.
The insulation structures between coils, between coils and core, and between winding
groups are made of high dielectric strength oil-impregnated sheets. Oil spaces are
provided with a precise relationship to the coil and pressboard structures to control
voltage stress concentrations. Specially formed insulation pieces are used over the coil
edge where the voltage stress is highest. This insulation is stressed in puncture rather
than creep for additional strength.

25
The pancake coils in a shell form transformer are arranged to terminate at the top of the
transformer where line end and tap connections can be made with a short lead. The
magnitude of circulating currents induced by high fields is minimized in an ABB Shell
Form Transformer because of the short lead length and unique subdivided lead
construction.
The inherent design characteristics of ABB Shell Form Transformers assure their reliable
operation. The performance of ABB Shell Form Transformers is verified by exclusive
modeling techniques prior to manufacture.

1.5.2 CORE FORM

1.5.2.1 DESIGN F EATURES
Core Form construction (see Figure 1-8) utilizes a series of cylindrical windings
stacked on a steel core. The core is at ground potential; therefore, the lowest voltage
winding is located adjacent to it, and the higher voltage windings are separated from the
core in order of voltage. The highest voltage winding is on the outside of the assembly.
The windings are supported laterally by laminated winding tubes and properly selected
conductor tension. Vertical support for the coils is provided by a plate type pressure
ring and lock plate assembly restrained by channel end frames.

26
 Figure 1-8: Partial Cutaway of a Core Form Transformer

Cooling of the core and coil assembly is accomplished by oil circulation through ducts
between the coils and also ducts within the core. The oil flow from the bottom to the top
of the tank is supported by the thermal head or temperature differential from the bottom
to the top of the transformer. The oil passes through a heat exchanger, where it cools
before reentering the transformer at the bottom.
The individual turns in the coil are insulated with high-density cellulose tape. Oil spaces
are provided between the disc sections of the coil with laminated spacer blocks. The oil
spaces between coils are maintained by vertical spacer rods.
1.5.2.2 MECHANICAL STRENGTH
The coil system of a core form transformer consists of cylindrical type windings placed
on a vertical steel core. The forces created by thru-fault currents tend to separate these
windings. The forces on the outer (or HV) winding push the winding out and place the

27
conductors in tension. The force on the inner (or LV) winding acts to compress the
winding, and the stress is transmitted to the winding tube (see Figure 1-9).

Figure 1-9: Section Through a Core Form Winding Group with an Expanded View of One Coil and
Spacers (arrows illustrate mechanical forces)

If the electrical centers of the coils are displaced by taps or an unequal winding
arrangement, a vertical force is introduced, which tends to telescope the windings. The
vertical forces can exceed 800,000 pounds per phase during the thru-fault conditions.
The forces in a core form transformer increase with transformer size; therefore, the
mechanical properties of winding tubes, vertical spacers, and radial spacers are critical to
the mechanical strength of the design. The tensile strength of the HV winding conductor is
also a very important consideration.
The vertical forces that act to telescope the windings are transmitted through radial
spacers to the pressure rings and then to the core end frames at the end of the winding.
These forces are transmitted through the winding across the narrow face of the
conductor, resulting in a high per-unit stress on the conductor and spacers. The vertical
forces tend to compress the spacer material, and over a period of time will cause
looseness between the disc sections of the coils. Preventing this will require some
means provided to maintain compression on the winding.
On ABB Core Form designs, the horizontal and vertical forces occurring during thru-
fault conditions are calculated during the design of a transformer, and the support
structure is designed accordingly. The coils are pre-stressed at the time of assembly to
maintain the vertical dimensional tolerances and the tightness of the coils.

28
Figure 1-10: Partial Cutaway of a Core Form Transformer Showing the Support Structure for Core
and Coils

1.5.2.3 T HERMAL CAPABILITY

The energy generated by the losses in the core and coil system of a core form
transformer is transmitted to the heat exchanger by the circulation of oil through ducts
between the coils and ducts within the core. The oil flow is supported by the thermal
head in the tank. The HV winding in a core form transformer is made up of a series of
disc sections positioned horizontally on the winding tube. The oil must travel through
both horizontal and vertical ducts to properly cool the conductors. Typically, the LV coil
construction is a helical winding that uses insulated rectangular or transpose conductors
and is cooled by oil flow through ducts on either side of the coil. The core has a relatively
large cross-sectional area and is located inside the coil assembly where heat is
concentrated; therefore, ducts must be provided within the core to allow oil circulation
for cooling.
The plate type pressure rings, which are located at each end of the coil assembly, tend to
block the flow of oil through the coil assembly; therefore, ducts and barriers must be
provided to direct the oil flow to the inner windings.
Forced cooling is applied to core form design by adding high velocity fans to the heat
exchanges to increase energy dissipation. The oil circulation is supported by the
thermal head in the transformer tank (see Figure 1-11).
If pumps are added for forced oil circulation, baffles must be provided to direct the oil
flow, otherwise the greater part of the oil volume will move upward in the area between
the HV winding and the tank wall. The barriers used to direct forced oil flow will impede

29
the flow during self-cooled operation. Transformer designs with continuous forced
cooling, such as generator step-up units, can advantageously utilize the baffled
arrangement.

(a) (b)
Figure 1-11: Partial Cutaway of a Core Form Transformer Illustrating (a) ONAN (Self-Cooled)
Cooling Action; and (b) OFAF (Forced-Cooled) Cooling Action

ABB uses a patented bypass valve on the core form design, which allows the proper
thermosiphon action to function during self-cooled operation. It will also properly direct
forced oil flow so that pumps can be used to an advantage during forced-cooled
operation.
1.5.2.4 DIELECTRIC CHARACTERISTICS
Overvoltage and system surge conditions can cause severe stresses on the insulation
system of core form transformers if the coil system is not arranged to distribute the
voltage surge uniformly across the winding. The initial distribution of a voltage surge is
determined by the ratio of the capacitance networks of the winding.
Transformers designed for service with system ratings of 69 kV or below generally utilize
a continuously wound HV coil made up of a column of disc sections separated by
horizontal oil ducts. The ratio of coil-to-coil capacitance to coil-to-ground capacitance will
be relatively low for this type of coil; however, additional insulation can be added in
critical areas to withstand any voltage surges.
Core form transformers used where system voltages are above 69 kV employ a variety
of winding configurations to increase the coil-to-coil capacitance, thus improving the
voltage surge distribution. HV coils for ABB Core Form Transformers in these voltage
ratings are mechanically similar to the continuous wound coils, except the turns are
interleaved to obtain a high series capacitance and a uniform voltage surge distribution.

30
Transformers rated above 100 MVA would require several conductors in parallel in order
to carry the current in the HV coils, and the winding procedure would also be very
complex.

The taps in a core form winding are brought out near the center of the coil in order to not
displace the electrical center of the coil. The tap leads are generally brought to a
switching mechanism at the top of the core and coil assembly (see Figure 1-12).
When underload taps are required, a small regulating winding is often employed. If tap
sections are placed in the HV coil, thyrister devices are used between the coil sections to
reduce the turn-to-turn voltage stresses.

Figure 1-12: Partial Cutaway of a Core Form Transformer Showing Coils, Insulation, and Tap
Leads

31
1.6 BUSHINGS [2]
Bushings may be classified by design as follows:
Condenser type:
a) Oil-impregnated paper insulation, with interspersed conducting (or condenser)
layers of oil-impregnated paper insulation continuously wound with interleaved
lined paper layers
b) Resin-bonded paper insulation, with interspersed conducting (condenser layers)
c) Resin-impregnated paper insulation. Bushing in which the major insulation is
impregnated with a curable epoxy resin

Non-condenser type:
a) Solid core or alternate layers of solid and liquid insulation
b) Solid mass of homogeneous insulating material (e.g. solid porcelain)
c) Gas-filled

Bushings may be further classified as either having a test tap, potential tap (also
referred to as capacitance, voltage tap) or not.

Condenser bushings facilitate electric stress control through the insertion of floating
equalizer screens made of aluminium or semi-conducting materials. The condenser
core in which the screens are located decreases the field gradient and distributes the
field along the length of the insulator. The screens are located coaxially resulting in the
optimal balance between external flashover and internal puncture strength.

Bushings, as with other electrical equipment, are bound by industry standards, which
vary between international, regional and national standards for the electrical and
mechanical performance of bushings. The international IEC standard has a broad global
acceptance but it cannot address specific regional issues. For this reason regional
standards deal with application issues such as atmospheric and seismic conditions or in
some cases the interchangeability of products among different manufacturers.

The rest of this section covers general information for bushing designed under
ANSI/IEEE standards and will focus mainly on condenser type bushings. Similar design
criteria are used under IEC standards.

Parts of the section related to bushings are excerpts from the ABB Instruction Manual
[3]

1.6.1 DESIGN AND CONSTRUCTION OF CAPACITANCES IN CONDENSER BUSHINGS

COMPLYING WITH THE IEEE STANDARDS [4]
ABB condenser bushings (e.g. Type “O Plus C”, Type AB) are designed for transformer
and oil-filled circuit breaker applications. These bushings meet all applicable
dimensional requirements of the IEEE Standard C57.19.01 and meet or exceed all
applicable electrical and mechanical requirements of the IEEE Standard C57.19.00.
They are also manufactured to meet the E.E.M.A.C. Standard.

32
A condenser bushing is essentially a series of concentric capacitors between the center
conductor and the ground sleeve or mounting flange. As per the IEEE Standards
C57.19.00 and C57.19.01, condenser bushings rated 115 kV and above are provided
with C1 (main) and C2 (tap) capacitances. The C1 capacitance is formed by the main
oil/paper insulation between the central conductor and the C1 layer/foil, which is inserted
during the condenser winding process. The C2 capacitance is formed by the tap
insulation between the C1 and the C2 layers. The C1 layer/foil is internally connected to
the voltage tap stud whereas the C2 layer/foil is permanently connected to the grounded
mounting flange. Under normal operating conditions, the C1 layer/foil is automatically
grounded to the mounting flange with the help of the screw-in voltage tap cover that
makes a connection between the tap stud and the mounting flange. The C2 insulation
under normal operating condition is therefore shorted and not subjected to any
voltage stress.

When such a bushing is used in conjunction with a potential device, the voltage tap is
connected to this device. Under this condition, the C1 and C2 capacitances are in series
and perform like a voltage or potential divider. The voltage developed across the C2
capacitance is modified by the potential device and is used for operation of relays,
and other instruments. Also, the voltage tap can be used for measuring the power factor
and capacitance of C1 and C2 insulation of the bushing. In addition, this tap can be used
for monitoring the partial discharge during factory tests and insulation leakage current
(including partial discharge) during field service operation. For condenser bushings with
potential taps, the C2 capacitance is much greater than the C1 capacitance and may be
10 times as much. Figure 1-13 shows the construction details of a typical condenser
bushing with voltage rating 115kV and above.

33
 CENTER CONDUCTOR

Figure 1-13: Design Details of a Typical Condenser Bushing, 115kV and Above

Condenser bushings rated 69 kV and below are provided with C1 capacitance as

per the IEEE Standards. This capacitance, which is considered the main capacitance,
is formed by the oil/paper insulation between the central conductor and the C1
layer/foil, which is inserted during the condenser winding process. The C1 layer/foil
is internally connected to the test tap. These bushings have an inherent C2
capacitance, which is formed by the insulation between the C1 layer and the mounting
flange. This insulation consists of a few layers of paper with adhesive, an oil gap
between the condenser core and the mounting flange, and the tap insulator. Under
normal operating conditions, the C1 layer/foil is automatically grounded to the mounting
flange with the help of the screw-in test tap cover that makes a connection between the
test tap spring and the flange. The C2 insulation under normal operating conditions is
therefore shorted and not subjected to any voltage stress. The test tap is used for
measuring the power factor and capacitance of C1 and C2 insulation of the bushing. In
addition, this tap is sometimes used for monitoring partial discharges during factory
tests and insulation leakage current (including partial discharge) during field service
operation. For condenser bushings with power factor taps, the C2 capacitance is
typically of the same order as the C1 capacitance. See Figure 1-14 for condenser design
and test tap details.

34
 Voltage
Equalizers Oil Impregnated
Paper

C1 Layer Foil

CENTER CONDUCTOR
Test Tap

Mounting
Flange
(Grounded)

C1 C2
Figure 1-14: Design Details of a Typical Condenser Bushings, 69 kV And Below

For both constructions the condenser is housed in a sealed cavity formed by the upper
and lower porcelain insulators, the high-strength, one-piece flange, and the metal or
glass expansion domes. This cavity along with the condenser is evacuated and then
filled with highly processed transformer oil for a very low moisture content and low
bushing power factor. This low moisture content and low power factor is maintained
throughout the life of the bushing by permanently sealing the bushing cavity. Spring-
loaded center clamping hardware is used to apply sufficient clamping pressure to seal
the bushing cavity during manufacturing. The upper and lower insulators, mounting
flange, flange extension, spring assembly, sight bowl, lower support, and clamping nut
form an oil-tight shell to contain the condenser and insulating oil. The sealing between
components is accomplished with oil-resistant “O-rings” in grooves and/or oil-resistant
flat fiber reinforced gaskets. This seal is never broken. A dehydrated nitrogen gas
cushion above the oil allows thermal expansion of the oil in the sealed cavity. The oil
level in the bushing can be monitored by visual inspection of the sight bowl.

The mounting flange and flange extension are high-strength, corrosion-resistant

aluminum. The lower support is designed to accept a variety of optional terminating
devices, such as standard threaded studs, NEMA blades, or draw rod system. The
upper insulator is one-piece, high-quality porcelain with sheds designed for maximum
performance. ABB condenser bushings are designed to meet or exceed “Heavy Creep”
requirements as described in IEEE Std C57.19.01-2000. Figure 1-15 shows a cutaway
view of a 138kV type ABB condenser bushing.

35
 Figure 1-15: Cutaway View of ABB Type AB Bushing 138 kV of Bushing Capacitances

1.6.2 BUSHINGS VOLTAGE TAP

ABB bushings rated 115 kV and higher (e.g. Type O Plus C) have a small housing
containing a voltage tap outlet just above the mounting flange. The terminal in the tap is
grounded by means of a spring clip in the tap cover. This tap is connected to one of the
inner foil electrodes of the condenser. In the factory, the voltage tap is tested at 20 kV,
50/60 HZ for 1 minute. Under normal operation, this tap is grounded. If the voltage tap is
used in conjunction with a potential/monitoring device, the voltage between the tap and
ground should be limited to 6 kV. While the purpose of the tap is to provide connection
to a bushing potential device, it also provides a convenient means for making
connections for measuring power factor and capacitance by the UST (Ungrounded
Specimen Test) method.

Many bushing users make it a practice to measure the UST power factor and
capacitance at the time of installation. We endorse this practice, and it is discussed in
more detail under the heading of “Maintenance.” When a connection is to be made to
the voltage tap, either for use with a potential device or for power factor measurement,

36
open the housing by removing the tap cover (item 19 in Figure 1-16). Assemble the
potential device connection or proceed with the power factor measurement.

After the power factor measurement is completed and if there is no connection to a

potential device, remove the test connection and close the housing by replacing the tap
cover. Be certain the cover is on tight. If the voltage tap is used for a connection to a
potential device, after the connection is assembled, remove the filler plug (Item 17,
Figure 1-16) and fill the chamber with clean, dry transformer oil. Leave an expansion
space of approximately one quarter of an inch at the top of the chamber when you fill it.
Coat the threads on the filler plug with a suitable sealer and replace the plug in the filling
hole. Be certain the plug is tight.

Figure 1-16: Sectional View of Bushing

37
WARNING:
DO NOT APPLY VOLTAGE TO THE BUSHING WITH THE VOLTAGE TAP COVER
REMOVED, EXCEPT WHEN USING THE BUSHING WITH A POTENTIAL DEVICE
OR WHEN MEASURING POWER FACTOR. IF THE TAP IS NOT GROUNDED, THE
VOLTAGE MAY EXCEED THE INSULATION DIELECTRIC STRENGTH, RESULTING
IN A FLASHOVER. THE VOLTAGE ON THE TAP MUST NOT EXCEED 5 kV WHEN
MEASURING POWER FACTOR. FAILURE TO FOLLOW THESE GUIDELINES
COULD RESULT IN SEVERE PERSONAL INJURY, DEATH, OR PROPERTY
DAMAGE.

1.6.3 CONNECTIONS

1.6.3.1 INTERNAL ELECTRICAL CONNECTIONS

The method used in making connections between a bushing and the apparatus on
which it is mounted will depend upon the type of connection used in the apparatus.
1.6.3.2 DRAW LEAD CONNECTED BUSHINGS
Bushings with current ratings of 800 amperes are generally designed with a hollow
conductor through which a flexible cable can be pulled. The cable is considered a
component of the apparatus on which the bushing is mounted and is not supplied with
the bushing.
1.6.3.3 BOTTOM CONNECTED BUSHINGS
Bushings rated 1,200 amperes and higher are designed to carry the current through the
center conductor. A circuit breaker interrupter or transformer terminal may be connected
to the lower support of the bushing.

1.6.4 LIQUID LEVEL INDICATION

The oil level in the bushing is adjusted in the factory to the normal level at approximately
25 °C. Unless there is subsequent mechanical damage to the bushing, which results in
loss of oil, the filler level should be satisfactory for the life of the bushing. Since
fluctuations in oil level will necessarily occur with changing temperatures, the column of
oil in the bushing is topped with a compressible cushion of nitrogen gas to fill the gas
space above the oil. The actual oil level can be seen on a bushing equipped with a sight
glass or a prismatic oil level gage. As long as the oil level can be seen, the level is at a
satisfactory height. When a low oil level is indicated, examine the bushing for possible
loss of oil, which could result in eventual electrical failure. A low level exists when the
pointer on a float type indicator is on “Low” or when the level has disappeared below the
sight glass or prismatic gage.

WARNING:
DO NOT OPERATE OR TEST A BUSHING WITH A LOW INTERNAL OIL LEVEL.
THIS COULD RESULT IN SERIOUS DAMAGE TO THE BUSHING, APPARATUS ON
WHICH THE BUSHING IS MOUNTED, AND/OR THE TESTING EQUIPMENT BEING

38
USED. OPERATION COULD RESULT IN SEVERE PERSONAL INJURY, DEATH, OR
PROPERTY DAMAGE.

1.6.5 PAINTING
The metal parts at the top end are painted for protection against the weather. Care
should be used to prevent scratching these painted surfaces. If the metal becomes
exposed, the area should be wiped with a commercial safety solvent and then wiped
dry. The cleaned area should then be coated with suitable outdoor gray enamel paint.

39
1.7 ON-LOAD TAP CHANGERS [5]
1.7.1 INTRODUCTIONS
There are some differences between tap-changers used under IEEE standards and tap-
changers used under IEC standards. The main differences are listed in Table 1-1 .

Table 1-1: IEC and IEEE Tap Changer Differences

Standard IEC IEEE

Designation OLTC LTC

Diverter switches Arcing switch

Tap Selection and Acing
Control Methods
Selector switch Arcing tap switches

Current Limiting Method Mainly resistor type Resistor and reactor type

The tap (regulation) winding in a load tap changing transformer is used to adjust the
number of transformer winding turns, usually to keep a constant voltage on the
secondary side of the transformer. If many electrical steps are required a plus/minus
connection or a coarse/fine connection is used. A plus/minus connection enables the
tapped winding to either add or subtract its voltage from the main winding. A coarse/fine
connection enables a coarse winding to be added to the regulating winding. The switch
that makes this connection is named change-over selector.

Figure 1-17: Different tap-changer connections.

On-load tap-changers must also be able to switch between the different positions
without interrupting the current flow. Different designed practices are used under IEC

40
and IEEE guidelines to achieve this smooth transition. The methods are outlined in the
sections below.

1.7.2 NORTH-AMERICAN PRACTICES 1

1.7.2.1 GENERAL DESCRIPTION OF LTCS

The tap or regulation winding in a load tap changing transformer is used to adjust the
number of transformer winding turns, usually in the secondary or low-voltage winding
and hence the transformer ratio. A regulating winding is commonly a layer type. A
reversing switch, located inside the LTC mechanism, enables the regulation winding to
either add or subtract its voltage from the low-voltage winding. Most LTCs have 16
mechanical tap positions, generally described as 32 electrical steps (16 above neutral
and 16 below). The usual range of regulation is ±10 % of the rated line voltage.
Although LTCs are built with other numbers of steps and ranges of regulation, the 32-
step, ±10 %, tap changing under load equipment has become a standard for many
types of transformers.

Voltage change must be provided smoothly and efficiently without interrupting the
secondary current flow, up to and including full load at the maximum nameplate rating,
plus any additional overload. When changing tap positions, the LTC mechanism must
“make before break” to avoid opening the secondary circuit. This causes the taps to be
connected together each time the LTC makes a voltage step. Electrically, this is a short
circuit in which a circulating current flows. The method used to limit this circulating
current defines the basic differences between the two types of LTC: reactance and
resistance types.

Both types use stationary and moving contacts. In some designs, the moving contacts
are located on an arm or shaft in the center of the fixed contacts and move over the
fixed contacts in a circular fashion. As the moving contacts make connection with each
fixed contact, a tap change is made.
1.7.2.2 REACTANCE TYPE LTCS
Reactance type LTCs use a preventive auto transformer, usually housed in the main
transformer tank and connected in series with the main low-voltage winding and
regulation windings. The preventive auto transformer is always connected in the circuit
and experiences circulating current each time a voltage step is made. The capacity of
the preventive auto transformer must be equal to the top nameplate rating of the
transformer multiplied by the step percentage of the LTC, plus sufficient capacity to
account for the circulating current during operation in the bridged position. Location and
construction of the preventive auto transformer can vary significantly between different
manufacturers and in different applications. In most cases, it is located in the main
transformer tank, sometimes on top of the main coil and core assembly. However, if the

1
Portions of this section are reprinted with permission from Electrical World Magazine, June
1995, copyright by The McGraw-Hill Companies, Inc. with all rights reserved. This reprint
implies no endorsement, either tacit or expressed, of any company, product, service, or
investment opportunity.

41
preventive auto transformer fails, the entire transformer must be taken out of service,
and the main core and coil assembly may be contaminated with carbon and copper
particles. A costly transformer repair may be the result. To reduce this possibility, the
preventive auto transformer can be located in a separate tank or compartment.

Reactance type LTCs are designed to operate continuously in the bridged position, thus
the need for the preventive auto to carry the full load current plus the circulating current.
However, a major shortcoming of the reactance type LTC is that the inherent inductance
of the preventive auto transformer increases the arcing time as the fixed and moving
contacts separate. Three different methods minimize the effect of this arcing and extend
contact life for as long as possible between overhauls.
1.7.2.3 ARCING CONTROL METHODS

1.7.2.3.1 Arcing Tap Switch

The arcing tap switch has tandem moving contacts, known as wipe contacts,
responsible for both breaking the arc and carrying the main current. Arcing takes place
on both edges of the wipe contacts, while the center of the same contacts carries the
load current during normal operation. The wiping action of these contacts is designed to
remove carbon buildup on the main contact and improve current carrying surface (see
Figure 1-18). Because the tap change operation is performed under oil, and no other
device is present to reduce contact wear and coking, the contamination of oil in this type
of LTC mechanism is much more severe than any other arcing-in-oil mechanism.

Figure 1-18: Arcing Tap Switch Reactance LTC

1.7.2.3.2 Arcing Switch and Tap Selector

The arcing switch-and-tap selector type has separate arcing and main current carrying
contacts. Arcing occurs on transfer switches located on a separate shaft from the main
current carrying contacts (see Figure 1-19). The two shafts are sequenced by a series
of gears, which are precisely aligned so that all arcing occurs on the transfer switches
and none on the main contacts.

42
 Figure 1-19: Arcing Switch-and-Tap Selector Reactance LTC

1.7.2.3.3 Drive Mechanism for Reactance Type LTCs

Reactance type LTC systems use direct-drive mechanisms. Direct-drive mechanisms
on reactance type LTC mechanisms use highly specialized gearing-and-scroll or dual-
slope cams to control the operating speed of the contacts and switches. When driven by
a motor, speed and positioning are controlled by gears and limit switches. Motor failure,
loss of power or control problems can cause the operation to stop before the tap change
is complete. The result is improper contact positioning, requiring immediate and
corrective action to avoid failure.

If the LTC is operated manually, movement must be fast and complete to limit contact
arcing. In a vacuum diverter LTC, the tap-selector contacts, diverter switches, and
vacuum bottles are connected by an extensive motor-driven gear train. Limit switches
stop the motor when a proper continuous operating position has been reached. In the
case of drive failure, it is possible for the mechanism to stop in an off-tap position so
that only one-half of the preventive auto transformer is in the circuit to carry the
circulating current. Most manufacturers state that, if this occurs, the mechanism must be
returned to a normal operating position as soon as possible or the transformer load
must be reduced to one-half of the nameplate rating. This off-tap position can also occur
in the arcing switch-and-tap selector type of reactance LTC.

Several users require that the preventive auto transformer be sized twice as large as
the normal center-tapped auto transformer and an alarm be included to avoid damage
from this condition. Manual operation of the vacuum diverter LTC, while energized, is
not recommended. If a vacuum bottle failed during a manual tap change, there would be
no way to stop the tap change from being completed, possibly damaging the
transformer and injuring the operator.
1.7.2.4 VACUUM INTERRUPTER TYPE LTCS
The vacuum interrupter-and-tap selector type is a significant improvement over other
reactance LTC mechanisms. In effect, the arcing current is diverted from the main

43
contacts into a vacuum bottle via two diverter switches. Because the arcing contacts are
housed in the vacuum bottle, there is no arcing to contaminate the oil (see Figure 1-20).
Minor arcing can occur in the switches that divert the current to the vacuum bottle.
Concentric drive shafts house the main current carrying contacts, diverter switches, and
vacuum bottles. These drive shafts operate in a precisely timed sequence so that
changes in the tap selector contacts only occur when no current is flowing. The tap
selector contacts usually last for the life of the transformer, since they are not burdened
with arcing and the associated contact wear.

Vacuum bottle switching eliminates multiple re-strikes and sustained arcing that occurs
in other types of reactance LTCs. The vacuum interrupter-and-tap selector is generally
good for 500,000 operations. This compares with 50,000 to 150,000 operations for the
other two reactance type LTC mechanisms. However, the complicated mechanical
interlocking and precise timing required is critical to proper operation.

Figure 1-20: Vacuum Interrupter Reactance Type LTC

1.7.2.5 RESISTANCE T YPE LTCS

Resistance type LTCs place resistors in the circuit to limit the circulating current during
the time that the tap change is taking place. The principal difference between resistance
type LTCs and reactance type mechanisms is that the resistance type never operates
continuously in the bridged position. The high-speed resistor transition type LTC (used
principally in the US) moves directly from one full-cycle position to the next, using the
impedance of the resistor to limit circulating currents for less than 60 milliseconds. The
rotating arm of the LTC mechanism carries both moving and arcing contacts, which are
electrically separate. The moving contact carries the main current, while the arcing
contacts carry the arcing current that occurs during a tap change (see Figure 1-21).

Because of the absence of inductance in the circuit, the arc is extinguished on the first
current/voltage zero. The high speed of the mechanism also contributes to the absence
of both re-strike and extended arcing. Arcing is limited to five or six milliseconds, which
is the average time to reach a current zero after contact separation. However, because
the bridged position is not used for continuous operation, the high-speed resistor

44
transition LTC needs 17 fixed contacts and 16 regulator winding conductors to provide
the electrical tap positions.

There is a second type of resistance LTC known as the resistive diverter. This type is
primarily used in Europe, where it is applied to the high-voltage transformer winding.
The main contacts of this mechanism are usually housed in the main transformer tank,
while the arcing contacts are housed in their own compartment.
Regulating Winding

Transition
Resistors

Moving Main Current

Carrying Contact

Figure 1-21: Resistance Type LTC

1.7.2.6 DRIVE MECHANISMS FOR RESISTANCE TYPE LTCS

Resistance type LTC systems use stored-energy drive mechanisms. The high-speed
resistive transition LTC mechanism uses the motor to charge a spring. The spring
cannot release its energy until it is fully charged, at which point the tap change is
made. Motor failure, loss of power, or control problems cannot leave the LTC
mechanism in an undesirable contact position.
1.7.2.7 FAILURE MECHANISMS FOR LTCS
From an analysis of failure statistics it is known that LTC failures can be grouped
under the following systems:

Electrical connections
Insulation system
Control system
Mechanical system

The typical failure mechanisms under each group are discussed below.

1.7.2.7.1 Electrical Connections

In an LTC, there are electrical connections that will not be opened during the lifetime
of the unit. In addition, there are switching contacts that will be opened and closed on

45
a frequent basis. The contact surfaces of the switching contacts are typically covered
with silver or an alloy of tungsten and copper. Because of the friction during the
switching, small particles will rub off the contact and move around in the oil. If many
particles come together, they are able to build a chain, which can create a short circuit
across contacts. Furthermore, these particles change the electrical fields within the
LTC and can cause partial discharges.

As the contact material becomes depleted, the underlying copper surface of the contact
becomes exposed. The copper and silver can react with oxygen in the oil or bond with
organic components that are present in some LTCs to form copper or silver oxides.
These materials form stable films on the surface of the copper and silver contacts,
resulting in an increase in resistance and in contact temperature. The increase in
temperature increases the deposition rate of the oxides and can lead to coking failures.

Coke, a black carbon material, is a by-product of oil degradation and is generated when
hydrocarbon-based insulating oils are subjected to extreme heat and arcing. The
presence of water contributes to the formation of the film as well as metal oxides on all
surfaces. The coking process tends to compound in nature. A point source of heat
begins the process. The resulting coke forms a carbon film resistor on the contact
surface, increasing mating resistance and heat by virtue of the higher I2R power loss.
The added heat anneals the spring material that holds the mating surfaces together,
releasing contact pressures and further adding to the problem. Eventually, the coke
formation prevents the contacts from moving, and a major failure can occur when the
LTC is required to make a change [6].

1.7.2.7.2 Insulation System

Usually the insulation system of a LTC consists of oil and solid insulation materials,
which depending on the construction, could be made of cardboard, fiberglass, or
epoxy resin. For the most part, only the insulation capability of the oil is of concern. It
is well known that oil degradation is highly dependent on temperature. Depending on
the brand of oil, the degradation of oil can start even under normal operating
conditions with a temperature over 60 °C. The rate of degradation significantly
increases at temperatures above 80 °C. As the oil degrades, CO, CO2, H2, and
hydrocarbon compounds like CH4, C2H6, C2H4, and C3H6 are generated. In addition, the
insulation capability of the oil decreases.

But the main destructive agent for the oil is hotspots, which are caused by joints or
contacts that have developed high-resistance surfaces and interfaces. The
temperature can go well over 150 °C on the connection surface. A by-product of the
hotspot degradation is the generation of soot particles in the oil. In addition, the
generation of some of the hydrocarbon compounds (C2H6, C2H4, and CH4) is greatly
enhanced by the presence of hotspots in the LTC.

The oil will also be destroyed by the high temperature of arcs, which occur during
normal switching operations. Partial discharges can be created by moving particles in
the oil as well as rough surfaces. As mentioned in the preceding section, at high

46
temperatures, oxygen and sulfur in the oil will react with copper and silver to form metal
oxides and sulfides on joints and contacts.

Excessive amounts of moisture in the oil will decrease the electric strength of the oil
and enhance the possibility of discharge activity.
1.7.2.7.3 Control System
The switching of the LTC is controlled and monitored by a system of relays and
RTUs. A failure of any of these components will lead to a failure of the LTC to
operate.
1.7.2.7.4 Mechanism
The force to switch the LTC is generated by a motor and transmitted by gears to the
contacts. The motor and the gears will age with time or develop their own set of
functional problems. For example, binding in the gears or the shafts that hold the
switches and contacts can slow down the switching sequence or prevent the
mechanism from moving. These problems as well as material or assembling failures
can cause a failure of the LTC.

1.7.3 EUROPEAN PRACTICES

1.7.3.1 RESISTANCE T YPE OLTCS

Resistance type OLTC’s exist in two main types: diverter switch type and selector
switch type. In both cases, transitions resistors are used to:

To carry the current during the switching operation when the main contact is
moving from one position to another
Reduce the circulation current that will start with the switching operation when
one loop in the regulation winding is short circuited

The arcs during the switching operation are normally extinguished at the first
current/voltage zero.

The high-speed resistive transition OLTC mechanism uses the motor to charge a
spring. The spring cannot release its energy until it is fully charged, at which point the
tap change is made. Motor failure, loss of power, or control problems cannot stop the
OLTC mechanism in an undesirable contact position because this critical part is
controlled exclusively by the springs.

The high speed of the mechanism also contributes to the absence of both re-strike and
extended arcing. The average arcing time is five to six milliseconds, which is the
average time to reach a current zero after contact separation. The time for a high-
speed resistor type OLTC to switch from one position to another position is
approximately 40-70 milliseconds. Loading of the springs and preparation for a new
switching operation takes between 2.5-6 seconds.

47
1.7.3.2 DIVERTER SWITCH OLTC
The diverter switch OLTC consists of a diverter switch and a tap selector. The diverter
switch, which breaks the arcs, is placed in a glass fiber (previously bakelite) cylinder.
This cylinder is tightly sealed to prevent the arcing products from entering the
transformer tank. The tap selector, which makes the connection to the tap (regulating)
winding, is placed under the diverter switch. Figure 1-22 shows the layout of a typical
diverter switch tap changer and Figure 1-23 shows a complete switching sequence
between taps.

Figure 1-22: An ABB diverter switch tap changer of type UC.

48
 Selector arm V lies on tap 6 and Selector contact H has moved in The main contact x has opened
selector arm H on tap 7. The the no-current state from tap 7 to and the arc has extinguished.
main contact x carries the load tap 5. The load current passes through
current. the resistor Ry and the resistor
contact y

The resistor contact u has The resistor contact y has The main contact v has closed,
closed. The load current is operated and the arc has resistor Ru is bypassed and the
shared between Ry and Ru. The extinguished. The load current load current passes through the
circulating current is limited by passes through Ru and contact u. main contact v. The on-load tap-
the resistor Ry plus Ru. changer is now in position 5.

Figure 1-23: Example of a switching sequence for a diverter switch type OLTC

1.7.3.3 SELECTOR SWITCH OLTC

Selector switch OLTC’s have only one compartment where both the breaking of arcs
and the connection to the different taps are made. This compartment is tightly sealed to
prevent arcing products from entering the transformer main tank and Figure 1-25 show
a layout and a switching sequence for a typical selector switch tap changer.

49
 Figure 1-24 : Selector switch tap-changers of UZ and UBB type

50
 UZ design with fixed contacts The main contact H is carrying The transition contact M2 has
in a circle and the main contact the load current. The transition made on the fixed contact 1,
surrounded by the transition contacts M1 and M2 are open, and the main contact H has
contacts at the top. resting in the space between been broken. After that the arc
the fixed contacts. has extinguished. The
transition resistor contact, M2,
carries the load current.

The transition contact M1 has The transition resistor contact The main contact H has made
made on the fixed contact 2. The M2 has broken at the fixed on contact 2. The main contact
load current is divided between contact 1, and the arc has H is carrying the load current.
the transition contacts M1 and extinguished. The transition
M2. The circulating current is resistor and the transition
limited by the resistors. contact M1 carry the load
current.

Figure 1-25 : Example of a switching sequence for selector switch tap-changers

1.7.3.4 T IE-IN RESISTORS

The change-over selector is only operated when it is not carrying current. However, due
to capacitive coupling to the surrounding windings, tank or core, the free floating tap
winding might develop a voltage that could create a dangerous arc on the change-over
selector contacts. This arcing will normally not affect the DGA in the transformer tank. If
the voltage over the selector is too high, a tie-in resistor is needed to reduce it. Figure
1-26 shows a tap changer layout that used a tie-in resistor to control arcing.

51
 The change-over selector is With a tie-in resistor the voltage With a switch that is only closed
moving and the tap winding is over the change-over selector at the time of the change-over
free floating. High voltages can can be reduced. There will, selector movement, the tie-in
appear over the change-over however, be extra losses due to losses can be avoided.
selector. the current in the tie-in resistor.

Figure 1-26: Tie-in connections

1.7.3.5 FAILURE MECHANISMS FOR OLTCS

From an analysis of failure statistics it is known that OLTC failures can be grouped
under the following systems:

Electrical connections
Insulation system
Control system
Mechanical system

The typical failure mechanisms under each group are discussed below.
1.7.3.5.1 Electrical Connections
The contacts where the breaking takes place are typically of copper/tungsten material.
At each operation, the arcing will carbonize some oil and a small amount of the contact
material will also end up in the oil. The maintenance criteria of the OLTC are set to
avoid these products since they tend to lower the dielectric withstand voltage. If proper
maintenance is not performed or if too much moisture enters the OLTC, the dielectric
strength of the oil in the OLTC can reach a dangerous level. If a contact remains in one
position for a long time (several months or years), the normal wiping action which
cleans the contact surfaces during normal operation of the tap selector contacts does
not occur. Consequently, the temperature in the contact might increase and led to
growth of carbon particles on the surface of the contact. This will cause the
temperature of the contact to increase and progressively worsens the situation. The
final result is the formation of coke on the contacts. This can lead to the generation of
free gas, and potentially to a flashover, which may catastrophically damage the
transformer.

52
In extreme cases, the carbon growth (sometimes referred to as pyrolytic carbon
growth) between and around the contacts can bind the contacts together. This
condition can cause mechanical damage if an attempt is made to operate the tap-
changer. Depending on the design, this may be a potential problem especially for the
change-over selector in on-load tap-changers.
1.7.3.5.2 Insulation System

The insulation system of an OLTC consists mainly of oil and solid insulation materials.
Depending on the construction, the solid insulation material could be made of
fiberglass, epoxy resin or bakelite. In the diverter and selector switches, the oil will be
degraded by the arcs even during normal switching operations. The condition of the oil
and electrically stressed surfaces in the solid material will be influenced by the arcing
products. Tap selectors are normally placed in the transformer tanks and therefore
share oil with the main winding insulation. Since no arcs are typically generated during
tap selection, there is no concern for the generation of arc-decomposition products that
may degrade the oil. However, excessive amounts of moisture in the oil will decrease
its electric strength and enhance the possibility of discharge activity.
1.7.3.5.3 Motor Drive Mechanism

The switching of the OLTC is performed from the OLTC motor device. This cabinet
contains relays and switches. A failure of any of these components can lead to a
malfunction of the control system for the OLTC. A fault in the motor drive mechanism
will not lead to a tap-changer failure.
1.7.3.5.4 Mechanism

A motor is used to drive the shaft system and gears that will load the spring battery and
also operate the tap selector. It is essential that the shaft system is correctly
coordinated with the tap-changer, else severe failures can result. If the gear box is
jammed, it can result in the motor protection stopping the motor from operating. If the
wear in the gear box is abnormal, it can prevent the tap-changer from operating.

53
1.8 STREAMING ELECTRIFICATION
Inside a power transformer, the insulation between high-voltage parts (high and low-
voltage coils) and grounded parts (tank walls and iron core) is provided mainly by paper,
pressboard, and low conductivity oil. In transformers with forced-oil cooling (OFAF), the
oil is circulated by pumps in a closed circuit and acts additionally as a coolant for the
power apparatus. Several factors have been shown to influence the likelihood of
streaming electrification in transformers. These include the electrostatic charging
tendency of the oil, the oil flow velocity, the conductivity of the oil, the insulation
temperature, and the moisture content of the insulation.

At any liquid-solid interface, and also at the contact surface between pressboard
insulation and transformer oil, an uneven charge distribution can be observed. The
uneven charge distribution is caused by the difference in adsorption rate of the solid
surface for positive and negative ions in the liquid. In a transformer, the solid surface
adsorbs typically more negative ions, forming a charge layer trapped within the
pressboard. The corresponding positive charges form a mobile, diffuse layer extending
into the liquid. The positive ions in the liquid are subjected to two counteracting forces:
the electrostatic force keeping the ions close to their negative counterparts in the solid
and the agitation of the fluid diffusing the ions to regions of lower ion concentration.
Apart from the diffusion process, there is also the macroscopic flow of the liquid
entraining the ions [7].

When the low-conductivity oil shears over the pressboard surface, it entrains the
diffused positive part of the electric double layer, while the solid retains the
corresponding negative charges on its surface. This process is called streaming
electrification, where the entrained ions form a streaming current. The entrained
charges may recombine with other countercharges in the liquid, be deposited on a
remote solid surface, flow along with the liquid, or undergo a combination of all these
processes. The accumulation of uni-polar charges on an insulated part of the structure,
a process referred to as static electrification, generates a potentially dangerous voltage
buildup. When the corresponding electric field surpasses a certain threshold, electrical
discharges may occur, damaging the system.

The damage can range from deterioration of the transformer oil to flashover between
high- and low-voltage coils or between an AC coil and ground, the latter most likely
leading to costly repair or replacement [8]. Figure 1-27 shows a graphical depiction of
the process of streaming electrification as described above.

54
 Figure 1-27: Streaming Electrification Model in Power Transformers [9]

1.8.1 CHARGING TENDENCY AND ITS EFFECT OF STREAMING ELECTRIFICATION

One of the key determinants of the risk of streaming electrification failure is the
electrostatic charging tendency (ECT) of the oil. This is defined as the amount of charge
generated per unit volume of oil as it flows though a specific filter and is measured in
microcoulombs ( C/m3). In a transformer, it provides an indication of the capability of oil
to generate charges as it flows past the surface of the cellulose in the cooling duct. It
has been found that the use of oils with high ECT in a transformer result in a higher
level of charge density in the transformer. This increases the risk of streaming
electrification failure.

The ECT is measured by forcing a specified volume of oil through a specified filter. As
the oil flows through the filter, charge separation occurs. The charge collected on the
filter is measured by an electrometer and is used to calculate the ECT. The changing
tendency of new oils is typically in the range of 0-150 C/m3. The charging tendencies
of oils in “normal” field units have been measured in the range of 5-200 C/m3.

55
Table 1-2 provides recommended limits of ECT for oils used in transformers in service.
The values provided in the table are to be used only as guidelines in determining the
risk of failure from streaming electrification. While most of the recorded streaming
electrification failures were in transformers with ECT values greater than 500, there
have been a few reported cases of failures in which the ECT was below 200. This points
to the varied number of conditions and mechanisms that can lead to a streaming
electrification failure. For example, if low-charging tendency oil is in a transformer that
has high flow velocities, and the transformer insulation is cold (as in a startup), sufficient
charge separation and accumulation can occur and increase the potential for streaming
electrification failure. On the other hand, in a transformer with normal flow velocities,
high-charging tendency oil at warm insulation temperatures would have reduced
potential for charge separation and accumulation. The risk of streaming electrification
failure would therefore be lower than the previous example.

Perhaps the most important factor that determines the level of charge separation in a
transformer is the flow velocity in the insulation ducts. The flow velocities in a large
power transformer vary depending on the design of the insulation ducts, the number of
pumps, and the volume flow rate of the cooling pumps. It is desirable to maintain as low
a flow rate as possible without affecting the cooling efficiency of the transformer. For
large power transformers that are a part of the installed base of inherited ABB
transformers, ABB design engineers have the capability to determine the flow velocities
in the cooling ducts to maintain the required cooling efficiencies. If a given transformer
is found to be susceptible to streaming electrification failure, ABB can make
recommendations for achieving the proper cooling efficiencies while minimizing the risk
of streaming electrification.

Table 1-2: Limits for Charging Tendency in Service Transformers

ECT ( C/m3) Potential for Streaming Electrification

<250 Normal
250-400 Moderate to High
>400 High

1.8.2 MITIGATION STRATEGIES FOR STREAMING ELECTRIFICATION

It is assumed that streaming electrification does exist to some extent in all transformers
with forced-oil cooling and especially those with directed flows. The goal is to determine
how these transformers can be safely operated in a way that will keep the effects of
streaming electrification under check. Several observations in a project [10] by ABB for
EPRI have been made as to the causes of the electrification process and modifications
to minimize these causes:
The charge generation process that aggravates the electrification process is
increased with flow rate and temperature. Charge relaxation, which
counterbalances the generation processes, is, on the other hand, enhanced
primarily by temperature. The result is that the potential for charge buildup is
increased at low temperatures, when the generation processes are dominant. As

56
the temperature increases, the relaxation processes are faster and eventually
overtake the generation processes. Beyond this point, the transformer can be
assumed to be out of danger with regard to charge buildup and eventual failure
of the insulation system.
The streaming electrification process is highly dependent on the charging
tendency of the insulating oil. High-charging tendency oils are likely to increase
the electrification characteristics by several times. The more high-charging an oil,
the more charges are generated under flow conditions. So, at low temperatures
there is more likelihood of extreme charge buildup, which can lead to damaging
discharges in the transformer. However, once the relaxation processes are
accelerated by temperature, these dangers subside as more charges relax than
are generated.
It was observed that the primary source of charge generation was inside the
winding ducts. The lower plenum, which has washers extending into the oil space
and also the entrance regions to the ducts, were presumed to generate some
charges are well. This was evidenced by high levels of charge density and
streaming currents that were measured in the upper plenum oil space than what
was measured in the lower plenum oil space.
It was also observed that the more open and leakage ducts there were in the
high-low voltage insulation of the transformer, the more charges were separated
in the ducts. This indicates that it may be possible to alter the design of the ducts
of a transformer so that there are fewer ducts open without sacrificing cooling
capability.
The height of the lower plenum oil space was found to play a very important role
in the level of charge generation that occurs in the ducts and more importantly at
the tips of the washers and the entrance regions to the ducts. It appears the local
eddy effects generated in the lower plenum become diffused as the height of the
oil space is increased. There is therefore less charge sheared from the insulation
structures extending into the oil space. This may be a possible change to a
problem transformer that may help alleviate the dangers of streaming
electrification.
It appears impurities that cause the charging tendency of the oil to increase can
be absorbed or loosely bonded to the cellulose fibers. Retrofitting with low-
charging oil after draining the high-charging oil may not be sufficient to reduce
electrification in the transformer. Perhaps, before oil retrofitting can be effective,
the cellulose insulation must be “washed” with oil that has a high degree of
solubility for impurities. This will hopefully dislodge most of the impurities from the
cellulose. Retrofitting with low-charging oil may then be effective.
Perhaps the most important observation was that the electrification process can
be controlled via modifications of the operational processes of the transformer.
Charge density measurements revealed a tremendous decrease in charge
accumulation in the upper plenum beyond 50 °C, even under full pumping
capabilities. The transformer can therefore be operated under reduced oil flow

57
 rates until the temperature is above this critical temperature. At this point, full oil
flow can be added without significant increases in charge densities and also any
dangers due to streaming electrification. The same procedure will be needed for
the reverse cycle.
ABB further recommends that utilities should ensure that all winding temperature
gauges are operational and properly calibrated; that the cooling controls operate
properly and are set in the AUTOMATIC position for operation. Also, the utility
should have in place operating procedures that prevent the running of all the
pumps when the oil temperature is below 50 °C. The charging tendency of the oil
should also be tested along with the other oil quality tests.
Several oil manufacturers recommend a chemical approach to solving this issue.
They focus on reduction of the ECT by using additives (inhibitors). This technique
could lead to a reduction in the risk of static electrification, especially for old
transformer designs.

58
2 A PRACTICAL APPROACH TO ASSESSING THE RISK OF
FAILURE OF POWER TRANSFORMERS

2.1 BACKGROUND

Transformer risk assessment is one of the main branches of transformer diagnostics. It

is related to strategic planning of technical and economical activities, i.e. how to
manage the transformer asset with available resources. The importance and need of
strategic planning is elaborated elsewhere in this handbook. However, in short it is
related to the inherent conflict between a desire of operating the transformer fleet at
lower cost and the requirement to retain the requested availability and reliability. A
consequence of this desire is a trend of operating the transformers harder (higher,
increasing loads) and for a longer period of time and at reduced costs (including
reduced costs for maintenance and expertise). The transformer fleet will become older
and many units will suffer an increasing risk of not being able to fulfill their function –
either by a technical malfunction or by being obsolete in another way. In most western
countries the average age of the transformer fleet is around 30-40 years, which is in the
range where the technical failure rate is expected to increase.

With the continuing ageing of transformers, it has become important to understand the
factors that contribute to elevated levels of risk of failure. The goal is that if these factors
are understood, then a risk of failure profile can be developed for each unit in an
organization’s fleet of transformers. This information allows the organization to target
appropriate strategies for mitigation, repair, upgrading, replacement, etc. for the correct
set of transformers as identified by the risk of failure profiles.

This section presents the general approach in a transformer risk assessment that
considers several factors, including condition indicators, known design capabilities, and
operational characteristics of a transformer. From these factors, a probable likelihood of
failure is calculated for each transformer. Together with the relative importance of each
unit to the power system, a prioritized strategy can be developed for transformers in a
fleet.

2.2 LIFE MANAGEMENT PROCESS

Transformer risk assessment is a part of an overall unit oriented transformer life

management process. This process has the following major ingredients:

1. A screening process to identify units for further scrutiny.

2. Condition analysis and more or less detailed design evaluation of individual units.
3. Life assessment decisions and their implementation (life extension via upgrading,
relocation, replacement etc.).

59
The risk assessment is used in the fleet screening process and its primary purpose is to
rank the transformers with respect to the risk. This allows us to prioritize the
transformers for follow-up corrective actions such as detailed design or condition
assessment, diagnostic evaluation, inspection, repair, or replacement. Another benefit
of a risk assessment is that the results (or scores) of the evaluation can provide the
basis for an intelligent estimation of the statistical technical risk of failure of the various
units.

2.2.1 RISK ASSESSMENT

In its true sense a risk consists of two different aspects – a probability of an occurrence
(e.g. a failure) during a time interval and the consequence of the occurrence. The
probability of a failure is the individually adjusted hazard function or failure rate. This
function depends on various technical factors – from design, service and diagnostics.
The consequence represents the severity of a failure and is determined essentially from
various costs of undelivered energy or power, costs of repair etc. It can also be
dependent on other factors such as strategic and environmental aspects etc.

In order to estimate a “true” adjusted individual failure rate, common statistical

distributions are used – but modified using models that depend on the score of the
technical risk. The ABB approach to fleet risk screening involves both risk aspects
mentioned above. However, the functional forms of these aspects are very complex
and it is difficult to determine them in an exact manner. Hence, in a first step, relative
parameters are used to map the original parameters. The technical risk (of a failure)
gives a value or score that depends on (or is a good estimator of) the individual failure
rate. The (relative or economic) importance is a measure of the negative consequences
of the failure. The result of the combined evaluation of the technical risk and importance
in a risk management investigation is normally presented in either of two ways:

As a Risk Index defined as a normalized product of the technical risk and relative
importance as shown in Figure 2-1.
In a two-dimensional diagram exemplified in Figure 2-2 and Figure 2-3 with the
technical risk and the relative importance on the two axes (Preferably the true
probability of failure and the true costs should be used but according to above
these parameters are difficult to determine).

60
 Technical Risk*Relative Importance

Risk Index

Transformer Units

Figure 2-1: Risk Index for a Number of Transformers

Figure 2-2: Risk Management Approach to Identify Transformers at Risk

61
 Technical Risk

B
Very Urgent

C A Urgent

Priority

Normal

100

Relative Importance

Figure 2-3: An Alternative Diagram for Risk Identification

The Risk Index represents the statistically expected cost due to a failure for the unit
under scrutiny. In this sense the product is related to the insurance premium to be paid
by the utility for keeping the unit in operation. In Figure 2-1 the Risk Index compares the
expected economical consequences of a failure for the different transformers belonging
to a utility. Discrimination between groups of units is clearly seen.
However, using a two-dimensional diagram is probably a better way to present the
results of a risk assessment. The two diagrams, Figure 2-2 and Figure 2-3, display the
outcome of analyses for two example fleets of transformers that have diverse risk of
failure characteristics as well as diverse relative importance. In the diagrams, each
transformer in the fleet is assigned a technical risk of failure and a relative importance
and is then displayed on the risk management plot. Those that fall in the (various
degrees of the) Red Zone are transformers with a combination of high risk of failure
and/or higher importance for the system. These are classified as Urgent (or very
Urgent), or those requiring immediate action. The next transformers are those in the
Yellow (Priority) Zone. Action would normally be taken on these transformers as soon
as the Urgent transformers have been taken care of. The transformers in the Normal
category would typically not require anything other than normal basic maintenance
unless circumstances move either the risk of failure or importance to a higher value into
the Yellow or Red Zone.

The intent of risk management is to move the identified transformers to areas of lower
risk. For example, a transformer can be moved from the Urgent zone to the normal zone
by reducing the expected technical risk of failure. (The arrows A in the figures exemplify

62
this case). The process of reducing the expected risk may begin with a detailed life
assessment study to identify ways of reducing the risk of failure. In the process, some of
the original assumptions regarding the risk of failure may also be modified to obtain a
more accurate view of the risk of failure. Actual methods for reducing the risk of failure
may include refurbishment of the transformer or accessories, moving the transformer to
an area with lower incidents of faults on the feeder lines, or it could involve system
changes like modifying reclosing practices or trimming trees in a right of way.
Another strategy of risk management involves reducing the relative importance of a
transformer. This is illustrated in the figures by case B. This strategy might involve
moving a higher-risk transformer to a less critical location. It might also include adding a
parallel spare transformer to reduce the impact of a failure.
Ideally, the actual strategies would include both types of solutions to reduce the risk of
failure and reduce the criticality of the application; exemplified by the case C.

2.2.2 LAYOUT OF THE EVALUATION PROCEDURE

Our risk assessment procedure focuses on the transformer functionality or
suitability-for-use [11]. We address various aspects that might jeopardize or
negatively influence this suitability-for-use.

Influential aspects on the suitability for

use of the transformer

Technical suitability Non-technical suitability

Accessories Main tank Economical Environ-

Strategic
incentives reasons mental
reasons
Mechanical Electric Thermal
suitability suitability suitability

Figure 2-4: Various directions of a transformer evaluation

Technical aspects include not only the traditional paper ageing aspects, but also other
aspects related to short-circuit strength, electric integrity, thermal degradation and
accessory failures. The focus on transformer functionality is fundamental. The aspects
that are addressed are linked to situations that are potentially dangerous to the
transformer operation. As can be seen in Figure 2-4, there are essentially four aspects
that are considered in determining the technical risk of failure of a given transformer:

63
 Mechanical aspects: This involves the risk of short circuit failure, which is based
on assessment of the short circuit strength of the windings and clamping
structure and the incidence and magnitude of short circuit through fault events.
Thermal aspects: This involves the winding thermal condition and is based on
the condition of the paper insulation. Aged, brittle insulation is more likely to fail
under the mechanical stress conditions. Also, metal parts at high temperature
could pose a risk to the transformer.
Electric aspects: This involves the risk of dielectric failure and is based on the
assessment of the dielectric withstand capability of the transformer insulation
system (oil, paper, etc.) and the electrical stress imposed by the power system
and naturally occurring events.
Accessory failures: Failures of a transformer accessory such as a bushing,
pump, or tap changer may cause a failure or loss of service of the transformer.

Each of these factors will be explained in more detail later. As for the consequences or
importance of a failure, the various cost factors mentioned above (undelivered power,
environmental costs etc) should be evaluated. This is an exercise for the utility or the
utility and ABB working together. Most often the utility ranks its transformer fleet with
respect to the relative importance of the various units and assigns an evaluation value
between 0 and 10 or 0 and 100.

2.2.3 EVALUATION PROCEDURE

Estimating the technical risk of failure of a transformer is a complex issue involving
analysis of historical failure data, knowledge of design issues, and interpretation of
diagnostic test results. The evaluation procedure also involves the selection of suitable
data to be used, rules and overall structure. ABB has methods of different complexity for
the evaluation. The ABB approach, [12,13,14,15,16] relies heavily on deep knowledge
in design, transformer manufacturing, service and transformer diagnostics.

The data used for reasoning when evaluating a large number of transformers in a fleet
screening must be based on easily available information in order for the evaluation to be
economically reasonable. The data for reasoning is then pre-processed data based on
various influential factors such as DGA, dissipation factor, oil condition, time-in-
operation, size, etc.

As illustrated in Figure 2-5, there are essentially two procedures used in algorithms for
combining the data for reasoning.

64
 Data for reasoning I.
Overall unstructured
method
Rules
w1
w2
Total Score
(Technical Risk)

wN

II.
Data for reasoning Rules Subgroup evaluation Method structured
along possible risks
Rules
Mechanical
Score wM
wE
Electric Total Score
Score (Technical Risk)
wT
Thermal
Score w..

Etc.

Figure 2-5: Procedures for obtaining the technical risk value for a transformer

Method I is an unstructured method while method II is structured according to different

external stress modes – mechanical stresses, thermal stresses, electric stresses,
auxiliary stresses etc.

In method I the total score is obtained through a formula applied directly to the data for
reasoning. Examples of such a formula are a weighing formula or a knockout criterion.
In the latter case the Total Score is determined only by the parameter having the worst
(maximum) influence. In method II the influential factors and data for reasoning are
combined in such a way that first an evaluation of the various subgroups are made and
then the risk scores of these subgroups are combined to a total evaluation.

The structure of method II can be extended beyond the “influential factor” procedure to
include a more detailed analysis involving design data and calculations and more
condition assessment measurements. This is a more precise risk of failure estimate
than performed with influential factors. It focuses on specific knowledge of the
transformer design and condition, in addition to the statistical and historical parameters.

65
The reasoning rules are based on known transformer relationships. This is the method
used in the Mature Transformer Maintenance Program (MTMPTM) offered by ABB.

In this evaluation a more pertinent statement of the condition and risk in connection with
various transformer stresses can be obtained, for example, regarding short-circuit
strength, dielectric strength, insulation ageing, tap changer status and loadability. The
more detailed design and condition ranking is for practical reasons applied only to a
reduced number of transformers since it requires more input data.

For an evaluation performed according to the structured method II, not only can a total
ranking be performed but also separate rankings according to the different types of
stress. The subgroup ranking can be made either when the data for reasoning is
obtained from influential factors or when it comes from more detailed
calculations/analyses.

A final step in a ranking procedure is to scrutinize the evaluation for parameters having
a large or significant single impact on the result – even if the total risk for the particular
transformer is calculated to be low. Knowledge of such parameters is used to direct the
engineering mitigation work.

2.2.4 PROBABILITY OF FAILURE – INDIVIDUAL FAILURE RATE

The evaluation described above yields an estimation of the technical risk in a relative
scale. Sometimes an absolute assessment of the individual failure rate of a unit is
desired. A first approximation to this is achieved by combining the technical risk with
statistical failure rate models as shown in Figure 2-6. This can be done on component
(influential factor), on subgroup level and on total risk level.

STATISTICAL (RELATIVE)
FAILURE RATE MODEL TECHNICAL RISK

MODIFICATION
MODEL =
f (Technical Risk)

INDIVIDUAL
FAILURE RATE

Figure 2-6: Combination of a statistical failure rate function with a technical parameter value to
obtain an estimation of the individual failure rate of the addressed transformer

66
2.3 ASSESSMENT OF THE TECHNICAL RISK OF FAILURE BY
CATEGORY (MTMPTM PROGRAM)
The algorithms for technical risk of failure, as discussed above, are based on influential
factors related to the individual subcategories [17,18,19]. The total technical risk is then
determined either directly from these influential factors or from a combination of the
assessed risks for the subcategories. To aid in the understanding of the risks for the
fleet of transformers, the relative risks for each of these categories will be briefly
presented.

2.3.1 MECHANICAL ASPECTS

One of the more common types of failures in power transformers is a winding failure
caused by the forces associated with a through-fault. As part of the risk of failure
analysis, each of the transformers in the fleet is evaluated for the potential risk of short
circuit failure. The influential risk factors that may be considered as part of the short
circuit risk include the transformer design, the dielectric and thermal condition of the
windings, the reclosing practices, and the average number of through-faults
experienced by the transformer in a given year. For example, it is typically the case that
transformers having a higher incidence of through-faults have the highest relative risk of
short circuit failure. These transformers are generally located in substations feeding
distribution lines.

2.3.2 THERMAL ASPECTS

An important factor in the risk of a short circuit failure is the condition of the paper
insulation. An aged transformer with brittle insulation and/or loose windings is more
likely to experience a failure under the same through-fault conditions than another
transformer of the same design that does not have brittle insulation or loose windings.
This principle is incorporated into the risk of failure analysis by the thermal winding risk
factor. Typical influential factors are the temperature, the age of the transformer
insulation, the relative compositions of produced carbon oxides, the load profile and the
MVA size.
Another thermal risk factor is hot spots in metallic materials such as core or current
carrying contacts. This risk is determined from DGA.

2.3.3 ELECTRIC ASPECTS - RISK OF DIELECTRIC FAILURE

The risk of dielectric failure involves both design and condition issues. Both design
knowledge and the historical information are used in this evaluation as well as the
diagnostic test data. Conditions such as the dissipation factor (tan , power factor) of
the insulation, oil quality results, the amount and distribution of dissolved gases in oil,
and design of the over voltage protection may be used in the evaluation of the dielectric
risk.

2.3.4 ASPECTS RELATED TO ACCESSORY FAILURE

Accessory failure refers to the loss of service of the transformer due to either the failure
or operational breakdown of an accessory. The accessories considered in this analysis
include oil coolant pumps, tap changers and bushings. The risk of accessory failure is

67
based on the type of accessory and the diagnostic evidence from DGA, power factor
(tan results, or other analyses.

In addition, a “Random failure risk” is included in the assessment. This risk is related to
external causes not associated with the design or condition of the transformer itself. It
takes into account other types of failure risks not accounted for in the other factors. The
parameters affecting random failure can be: the type of transformer, the location, cases
where a transformer must be removed from service to de-gas the oil, loading practice
etc. This type of risk also includes transformers at risk for streaming electrification due
to the design type, potential high oil velocity, and/or cooling operation philosophy.

2.3.5 TOTAL TECHNICAL RISK OF FAILURE

The total technical risk (or individual failure rate) is obtained either directly from method
I in Figure 2-5 or (better) according to method II from a combination of each of the risk
categories discussed above. The risk of failure is determined for each of the
transformers in the fleet.

Figure 2-7 shows a histogram of failure rates for over 200 power transformers. An
indication of the relative importance of each of the transformers is also calculated based
on the replacement cost for the transformer or the criticality of the transformer to system
reliability. In order to develop a priority for addressing mitigation strategies for the
transformers, a plot of the risk of failure vs. the importance is shown in Figure 2-8.

40
35
30
Number of Units

25
20
15
10
5
0
0.125
0.625

1.125
1.625
2.125
2.625
3.125

3.625
4.125
4.625
5.125

5.625
6.125

Total Failure Risk

Figure 2-7: Total Risk of Failure of Transformers

68
 100

80

Relative Importance
A
60

40

20
B

0
0.0 1.0 2.0 3.0 4.0 5.0 6.0

Probability of Failure

Figure 2-8: Categorization of Risk (Technical Risk or probability of failure and relative importance)
Profiles for Power Transformers

In this chart, the transformers are grouped into three categories: Urgent (red), Priority
(yellow), and Normal (green). For each transformer in the Urgent or Priority regions
(these are considered the abnormal regions), a more detailed analysis is made to
identify which risk factors were prominent in placing it in that category. For those factors
that are flagged, the sub-factors are analyzed to determine which underlying
parameters triggered the abnormal status. All such sub-factors are summarized as the
reasons for the transformer being classified in a particular abnormal category. This
detailed analysis is then used as the basis of recommendations for mitigation actions.

As an example, consider the transformer labeled A in Figure 2-8. Ninety-six percent of

the total risk was contributed by the relative risk of accessory failure. The underlying
factor for the high accessory risk factor was traced to a conditional factor associated
with a leaking high-voltage bushing. On the other hand, the unit labeled B is at risk due
to several factors. It has increased potential for through-fault failure due to its design
and the high incidence of through-faults at the substation. In addition, its LTC is at risk
for failure due to the type of LTC and the presence of certain combustible gases in the
selector switch compartment. The same unit is also at risk of dielectric failure since the
kV breakdown of the oil is low and the high-low insulation power factor is higher than
1%.

The histogram in Figure 2-7 is also suitable when comparing the evaluation of a single
transformer with the evaluation of previously evaluated units. For instance, a new
transformer with the risk evaluation value 3 belongs to the upper 10 % most risky units
of all units evaluated so far.

69
2.4 RISK MITIGATION
For all of the transformers identified in the Urgent or Priority category, recommended
risk mitigation actions are suggested based on the underlying factors that support the
high-risk evaluation. In some cases, immediate action such as replacement of an
offending bushing or inspection of a tap changer can be taken to correct the situation.
For other cases, additional diagnostic testing is needed to better evaluate the risk to
determine the most appropriate maintenance and risk mitigation actions. In such cases,
the evaluation is taken further to include also condition assessment and design
assessment if possible

One important risk management area is to identify spare transformers for the Urgent
and Priority transformers in the system. The risk of failure ranking is used to identify
which transformers to begin with. In many cases, especially those where design issues
such as short circuit strength are involved, it may be more appropriate to replace a high-
risk transformer with a new unit and keep the older transformer as a spare in order to
reduce the risk and improve the system reliability.

For a great number of the transformers that have been analyzed, the greatest risks of
failure are (1) risk of accessory (bushing, tap changer, pump, etc.) failure, (2) failure due
to through-fault currents caused by close-in faults on the transmission system, and (3)
risk of dielectric failure due to various causes.

2.5 SUMMARY
In this section we have discussed the principle and methods for the risk assessment of
power transformers that takes into consideration various risk factors that together
present a comprehensive risk profile for a given transformer. Each of these risk factors
is assessed based on certain condition indicators and/or the design and/or the
application of the transformer. This results in a quantitative and repeatable assessment
of the risk of failure. The risk of failure is used in conjunction with the relative importance
of each transformer to classify the overall risk of each transformer. By understanding
the underlying reasons for the risk classification of each transformer, the appropriate
mitigation actions can be prescribed. Because of the quantitative nature of the analysis,
mitigation options can be evaluated to determine the most cost effective means of
reducing risk of failure of a given transformer. So far, this method of risk assessment
has been performed on a large number of transformers, including industrial
transformers, generator step-ups, and power transformers of various voltage classes
and MVA sizes.

70
3 DIAGNOSIS OF TRANSFORMERS
Power transformers are of prime importance for electrical power systems. The condition
of a power transformer is crucial for its successful operation and, as a consequence, for
the reliability of the power system as whole.

During transportation or installation or under service operation, a power transformer is

exposed to transient and steady-state stresses that can affect its condition as well as its
useful life. In addition, transformers are subjected to a natural ageing process under
service conditions.

The detection of incipient faults which may be caused by insulation weakness,

malfunction, defects or deterioration is of fundamental importance. So is the estimation
of the ageing condition of the power transformer insulation and its main accessories.
This may allow the operators to plan adequate corrective actions at an early stage.
Diagnostic techniques are usually used as a means to detect fault and ageing condition
in power transformers in the field. Conventional and advanced off-line diagnostic
methods may be applied periodically or whenever necessary to help detect incipient
faults. In some cases, modern on-line monitoring systems may be applied to
continuously monitor the condition of the transformer and/or its accessories.

3.1 DIAGNOSTICS METHODS FOR POWER TRANSFORMERS AND

ACCESSORIES [20]
A set of modern diagnostics methods is available and applied for oil filled power
transformers and accessories. In this book, both general and advanced diagnostic
methods are presented in a summarized format.

General diagnostic methods include the analysis of oil quality (physical, chemical and
electrical properties, as well as dissolved gases), core and core insulation analysis,
winding and insulation analysis and analysis of the condition of the accessories.

In addition, there are advanced diagnostic methods that address the thermal, electrical
and mechanical condition of a transformer. Thermal assessment techniques are well
established and are typically used to analyze the condition and remaining life of the
transformer insulation. Electrical assessment includes partial discharge (PD) analysis,
which is a powerful tool used to detect incipient faults in the transformer insulation.
Mechanical assessment includes frequency response analysis (FRA), which is applied
to detect changes in transformer winding dimensions due to deformations,
displacements, shorted turns, etc. Other methods are presented in the proceeding
sections.

3.1.1 DIAGNOSTIC METHODS FOR POWER TRANSFORMERS

Power transformers are considered to include generator step-up transformers,
transmission step-down transformers, system inter-tie transmission transformers, and

71
DC converter transformers, together with such associated equipment as shunt, series,
and saturated reactors. Power transformers may be equipped with on-load and/or de-
energized tap changers.

Power transformers are used to reduce the costs of power transmission by transforming
the voltage at which current is transmitted. Shunt and series reactor components are
similar to transformers but need to absorb reactive power and limit fault currents
respectively.

The insulation system of a power transformer is a combination of cellulose based

material impregnated with mineral insulating oil. The following cellulose materials are
normally used:

Kraft paper used as a turn-turn insulation;

Kraft-based high density transformer board used for winding spacers and
mechanical supports; and
Kraft-based medium to high density transformer board used as major insulation
between windings and from windings to ground.

Kraft paper can also be converted into flexible creped paper and used for insulating
conductors and leads. Mineral insulating oil is used as an impregnating fluid for
dielectric and cooling purposes.

Since the mid 1960s, thermally-upgraded Kraft paper has been used as turn-to-turn
insulation in transformers. In more recent years, natural esters (vegetable oils) are
being used as insulating fluids in power transformers.
3.1.1.1 STRESSES ACTING ON POWER T RANSFORMERS
The major stresses acting on a power transformer, either individually or in conjunction,
are:

MECHANICAL stresses between conductors, leads, and windings due to short-

time load overcurrents, fault currents mainly caused by system
short circuit and inrush currents while under energization conditions
THERMAL stresses, due heating or local overheating, associated to short-time
overload currents and leakage flux when loading above nameplate
rating, or due to malfunction of the cooling systems
DIELECTRIC stresses, due to system overvoltages, transient impulse conditions,
or internal resonances within the windings

A definitive analysis of the subject of diagnostic tests on power transformers must take
into account that the majority of diagnostic indicators are sensitive to all three
fundamental stresses acting on the transformer. Therefore, the general interpretations
of the outputs of the diagnostic indicators, including the localization of faults, can be
problematic for a reliable evaluation of the risk of failure. The experience and
interpretation capabilities of transformer experts are crucial for a successfully diagnosis.

72
The situation is also complicated because dielectric failure is often the final stage
consequent to the mechanical and/or thermal stresses, especially if moisture and/or oil
deterioration have already placed the transformer in a hazardous condition. This fact
underscores the importance of assessing the service stresses (overvoltages,
overcurrents, temperature, etc.) jointly with a detailed knowledge of the design
technology and materials.

The interpretation of the values and trends of the diagnostics tools must therefore be
tailored to different units in order to avoid unjustified alarms.
3.1.1.2 DETERIORATION F ACTORS AND F AILURE MECHANISMS
Deterioration of the paper-oil insulation is caused by thermal stresses and is
accelerated by the presence of moisture, oxygen, or high acidity compounds in the oil.
The insulation is unlikely to exhibit a lower dielectric strength after deterioration, but it is
more subject to rupture under mechanical stress, leading to dielectric failure as a
consequence.

Few transformers fail due to old age; they usually fail as a consequence of:

Short circuit faults

Local overheating due to circulating currents, current unbalance or the effects of
leakage flux
Insulation failure under electric stress (dielectric failure), perhaps as the final stage
of a scenario involving previous short-circuit faults and/or local overheating, and
Accessory failures (bushings, tap changers, coolers, surge-arresters, etc.).

Faults can be classified as developing in one of three time scales:

An immediate fault where electrical breakdown occurs within seconds of a short

circuit, system overvoltage, lightning impulse surge or any other transient
phenomena in the system interacting with the transformer;
A local fault developing over days, weeks, or months;
A deterioration of HV insulation over a period of months or years.

Diagnostic techniques have been introduced mainly to detect the presence of small
local faults and to monitor their development over time on a period of weeks or months.
They provide evidence to plan for further investigation and remedial work to take place
on a planned basis, rather than as an emergency.
3.1.1.3 DIAGNOSTIC METHODS
Table 3-1 presents the diagnostic techniques used most widely for power transformers,
together with their field of application, present status, effectiveness, and specific
references. Diagnostic techniques may give information on detection of incipient faults
as well as about the specific source or location in a transformer structure.

73
 Table 3-1: Most Important Diagnostic Techniques Used for Power Transformers

SERVICE PROVEN
STATUS OF THE
CONDITIONS EFFECTIVENESS OF
PROBLEMS DIAGNOSTIC TECHNIQUES DIAGNOSTIC REFERENCE
OF THE 3 THE DIAGNOSTIC
2 TECHNIQUE 4
EQUIPMENT TECHNIQUE

1. Excitation Current OFF-S A M

2. Low-voltage impulse OFF-S A L
21
MECHANICAL 3. Frequency response analysis OFF-S A H
22
4. Leakage inductance measurement OFF-S A M/H
23
5. Capacitance OFF-S A H

GAS-IN-OIL ANALYSIS
A H 24, 25
6. Gas chromatography ON
A M 26
7. Equivalent Hydrogen method ON

OIL-PAPER DETERIORATION
THERMAL 8. Liquid chromatography-DP ON B M/H
27
method
9. Furan Analysis ON B M/H

HOTSPOT DETECTION
10. Invasive sensors ON B L 28
11. Infrared thermography ON A H

OIL ANALYSIS
12. Moisture, electric strength, ON A M 29
resistivity, etc.
13. Turns ratio OFF-S A L
DIELECTRIC PD MEASUREMENT
14. Ultrasonic method ON B M/H 30, 31
15. Electrical method ON B M/H 32

16. Power Factor and Capacitance OFF-S A H

33
17. Dielectric Frequency Response OFF-S A H

3.1.2 DIAGNOSTIC METHODS FOR BUSHINGS

Bushings provide insulated terminals carrying current into and out from power
apparatus, such as transformers, reactors, circuit breakers and HVDC valve halls. They
additionally serve as mechanical supports for external bus and lines, as well as for
internal supports, such as circuit breaker contacts.

2
OFF-S = equipment out of service at site, OFF-L = equipment out of service in laboratory, ON = equipment in service
3
A = generally applied, B = development stage
4
H = high, M = medium, L = low

74
Bushings are constructed to numerous design considerations, but commonly consist of:
Center conductor
Mounting flange
Insulation (solid, fluid, plastic, or in combination) between conductor and flange

The core may consist of only two terminals:

the bushing center conductor; and
the mounting flange/ground sleeve system

In a bushing having a non-condenser body design the electric voltage will be distributed
logarithmically between the conductor and the flange. In a bushing having a condenser
body design, it may include strategically placed conducting wrappings or layers to
uniformly distribute the voltage stresses in the core. Most high-voltage bushing designs
use the condenser principle.

The insulation system may be:

Dry: bulk porcelain, gas, or air
Wound paper and wound paper with conducting layers

The wound paper core may be:

Oil-immersed, in porcelain
Oil-impregnated, oil-immersed
Resin-bonded, either oil or gas-immersed
Resin-impregnated, oil-immersed
3.1.2.1 STRESSES ACTING ON BUSHINGS
Apparatus bushings are subject to the effects of internal apparatus voltage, current,
temperature, and contamination but are also subject to external atmospheric and
environmental conditions as well as mechanical stresses.
3.1.2.2 DETERIORATION F ACTORS AND F AILURE MECHANISMS
Bushing insulation integrity degrades in normal service from internal moisture, internal
PD and tracking from external corona, flashover and tracking from ageing, and from
physical damage. Despite the intention that outdoors bushings be hermetically sealed
devices, inadvertent ingress of moisture resulting from defective gasket seals and
physical strain or damage is a major cause of insulation deterioration.

Internal PD and tracking can be a symptom and result of internal moisture

contamination, physical shrinkage of plastic or compound fillers, system overvoltage or
marginal designs where there is inadequate stress distribution. External surface
contamination effects can be minimized by proper housekeeping and/or by use of
coatings. Bushing insulation systems do not usually deteriorate due to time alone,
except where they have been subjected to unusual service conditions, such as
excessive temperature or operation at voltages above the nameplate rating over long
periods of time.

75
3.1.2.3 DIAGNOSTIC METHODS
Bushings are ideally suited for field-testing by dielectric diagnostics to detect and
analyze defects or deterioration resulting from the conditions previously described.
Bushings are commonly field tested when new to confirm factory test data and to
monitor for shipping damage, and then periodically tested following system disturbances
or apparatus failures and routine outages.

Table 3-2 reports the diagnostic techniques used most widely on bushings alone or
installed together with their field of application. The present status and effectiveness of
the techniques and specific references for further description of the method are also
provided.

Table 3-2: Most Important Diagnostic Techniques Used for Bushings

PROVEN
SERVICE STATUS OF THE
DIAGNOSTIC EFFECTIVENESS OF
PROBLEMS CONDITIONS OF DIAGNOSTIC REFERENCE
TECHNIQUES 5 6 THE DIAGNOSTIC
THE EQUIPMENT TECHNIQUE 7
TECHNIQUE
Moisture Capacitance/Power Factor OFF-S A H 34, 35, 36, 37
Tap voltage ON A M 34, 35, 36, 37
DC resistance OFF-S A L 34, 37
Hot-collar OFF-S A H 37
Corona Partial discharge (PD) OFF-S B M/L 37
Radio-influence voltage ON B M 37
Ageing (RIV)
Capacitance/Power Factor OFF-S A H 34, 35, 36, 37
DC resistance OFF-S A L 34, 37
Short-circuited Capacitance/Power Factor OFF-S ON/OFF-S A H 34, 35, 36, 37
condensers Tap voltage A M 34, 37
Internal surface PD/RIV OFF-S A M/L 34, 37
leakage Capacitance Power Factor OFF-S A M 34, 37
AC dielectric loss OFF-S A H 37
Poor Infrared scanning ON A H 37
connections

3.1.3 DIAGNOSTIC METHODS FOR SURGE ARRESTERS

Surge arresters are used as protective devices to limit the amplitude of possible
overvoltages in the electrical network. However, most of the time they are expected to
function as insulators. According to service experience, most of the trouble caused by
surge arresters comes from the deterioration of this "insulator function."

The majority of arresters in service are still of the so called conventional type, i.e. made
of the series combination of active gaps and non-linear silicon carbide (SiC) resistors,
encapsulated in a porcelain housing. For this type, the withstand voltage relies mainly
on the gaps, spacers, and the external grading rings used in higher voltage applications.

5
OFF-S = equipment out of service at site, OFF-L = equipment out of service in laboratory, ON = equipment in service
6
A = generally applied, B = development stage
7
H = high, M = medium, L = low

76
A very important feature is that the voltage distribution across the several gaps in series
is controlled by "grading" non-linear resistances and also sometimes by internal
capacitors.

Nowadays, Metal Oxide Varistors (MOV) are able to perform the voltage clamping
function as well as the insulator function: several tens of non-linear zinc oxide (ZnO)
varistors are connected in series, and gaps are no longer needed in MOV arresters.
3.1.3.1 STRESSES ACTING ON SURGE ARRESTERS
In addition to the obvious electric stress, arresters are also exposed to substantial
thermal stress. Sizeable temperature increase is caused by normal duty operation or by
external potential redistribution due to pollution or salt in combination with rain or fog.

In the latter case, internal discharges may also occur, generating reactive species that
can cause internal surface deterioration in the arrester.

Mechanical stresses are normally taken entirely by the porcelain insulator, whereas the
active arrester parts are well protected.
3.1.3.2 DETERIORATION F ACTORS AND F AILURE MECHANISMS
The insulator function of arresters can be deteriorated in several ways:

Moisture ingress: Condensation and corrosion inside the arrester can affect the
dielectric withstand of insulating pieces and surfaces, and the spark-over characteristics
of the gaps can also be affected. Tightness is a must for good performance of arresters.
Heavy external pollution: The surface currents on heavily contaminated housings,
especially for multi-unit arresters, affect the voltage distribution and may create
important temperature rises, jeopardizing the grading system of conventional arresters
or the blocks in MOV arresters.
Discharges inside the arresters: Decomposition products resulting from gas
discharges in the arrester can impair the chemical stability and the dielectric surface
properties of the internal parts, especially of the varistors.
Varistor deteriorations: ZnO blocks in MOV arresters, as well as grading resistors in
SiC gapped type arresters, may suffer from changes of their characteristics during
service. This results in higher leakage currents and losses. For conventional arresters,
the final stage of deterioration is sparking at service voltage; for MOV arresters, the final
stage is thermal runaway.
Grading capacitor deterioration: Less frequent than grading resistor deterioration, but
essentially the same effect.
Gap deterioration by arrester duty: Spark-over characteristics will be affected.

The failure rate of arresters depends on the keraunic level (number of thunderstorm
days/year), the system voltage, and the margin used in the selection of the rated
voltage. For healthy and well-designed arresters, the failure rate should not be higher
than about 1/1,000 per year.

77
Once a particular category of arresters (make, environment, age) suffers from one of
the above-mentioned problems, the failure rate becomes much higher. Diagnostic
techniques are then necessary to make decisions on the replacement policy. Otherwise
diagnostic techniques are not likely to be more intensively used than just being included
in the maintenance programs.
3.1.3.3 DIAGNOSTIC METHODS
Table 3-3 summarizes the diagnostic techniques used most widely for surge arresters,
together with their field of application, present status, effectiveness, and specific
references.

Table 3-3: Most Important Diagnostic Techniques Used for Surge Arresters

PROVEN
STATUS OF EFFECTIVE-
SERVICE
THE NESS OF THE
PROBLEMS DIAGNOSTIC TECHNIQUES CONDITIONS OF REFERENCE
8 DIAGNOSTIC DIAGNOSTIC
THE EQUIPMENT 9 TECHNIQUE
TECHNIQUE 10

CONVENTIONAL SURGE ARRESTERS

- Visual inspection
ON A L
External pollution - Measurement of external leakage
ON ? L
current
Heating of grading
-Thermovision ON A M
resistors
- Leakage current under controlled
OFF-S A H 38
Deterioration of voltage
OFF-S A H 38
grading system - Watt loss under controlled voltage
OFF-S A H 38
- 60 Hz spark-over voltage
METAL-OXIDE SURGE ARRESTERS

- Visual inspection
ON A L
External pollution - Measurement of external leakage
ON ? L
current

- Leakage current
ON A L
- Harmonic decomposition of leakage
Deterioration of ON A M 39
current
varistor blocks ON A H 38
- Peak of resistive current
ON B H 40
- 3rd harmonic of resistive current
OFF-L A H
- Reference voltage

8
OFF-S = equipment out of service at site, OFF-L = equipment out of service in laboratory, ON = equipment in service
9
A = generally applied, B = development stage
10
H = high, M = medium, L = low

78
3.2 GENERAL DIAGNOSIS TOOLS
3.2.1 OIL QUALITY ASSESSMENT

3.2.1.1 FACTORS AFFECTING THE HEALTH AND LIFE OF POWER T RANSFORMERS 11

The three main components subject to deterioration and contamination in a transformer
are the paper, which is used for conductor insulation; the pressboard, which is used for
the major insulation and winding support; and the insulating oil. Water, air or gas
bubbles, particles of different origin, oxygen, and oil ageing products are agents of
degradation. The presence of these elements in the transformer can directly reduce the
dielectric strength of the insulation system or result in acceleration of the rate of ageing
of the insulation system.

The level of possible contamination of a transformer over years depends on its design,
especially on the effectiveness of the oil preservation system, and sources of
contamination. Detection of possible sources of contamination in the particular
transformer is a critical step of its condition assessment. The CIGRE working group
12.18 has suggested some possible sources of typical contamination that are listed in
Table 3-4. The objects of primary concern should be transformers that have poor
sealing, worn-out oil pump bearings, sources of overheating, aged oil and free-breathing
transformers operating with variable load.

Table 3-4: Sources of Typical Contamination of Power Transformers

Contaminant Source Storage Mode

Water Entering as a Direct exposure of the insulation to air during installation Most of the water is stored in the thin
Vapor and inspection. structure that operates at oil bulk
Ingress by viscous movement of wet air through temperature (20-30% of the total
unsealed oil expansion systems (conservator tanks) and insulation mass).
through loose or cracked gaskets (at flange Presence of “wet zones” (typically bottom
connections). part of insulation of outer winding).
As a byproduct of the ageing of the insulation system Concentration in the vicinity of hotspots
Bound water-in–oil.
Liquid Water Damaged water heat exchangers. Typically on the bottom parts of the tank
When the transformer is under less than atmospheric and coolers.
pressure because of bad gaskets and loose connections Diffusion into the oil.
(the top seal of draw-lead bushings, the seals in Temperature migration.
explosion vents, leaks through poor sealing of nitrogen Movement of ice by oil flow.
blanketed transformer).
Condensation in the coolest regions.
Particles From manufacturing process Migration in oil.
Dress and test dirt Sediment under effect of gravity, oil flow
Oil ageing and particularly effect of electrical and
Wear of aged cellulose electromagnetic field that attracts the
Overheating of metals (carbon) conductive particles and stimulates
Carbon from OLTC depositing them on the winding surfaces,
Wear of the pump bearings pressboard barriers, and bushing
porcelain.

11
This section is extracted by permission from CIGRE WG12.18 – Brochure N° 227, 2003 ‘Life
Management of Transformers’, CIGRE, Paris

79
Processes of insulation deterioration involve slow diffusion of water, gases, and ageing
products, and therefore affect basically only a part of the insulation structure, the so
called “thin structure” (conductor insulation, pressboard sheets, etc.), which comprises
typically 40-60 % of the total mass.

The moisture distribution is a function of the system moisture content, thermal

distribution, and also the dimensions of the cellulosic insulation structures. Parts of the
insulation that are in contact with less heated layers of bulk oil may have notably higher
moisture content.

Hydrolysis is a dominant mechanism of insulation ageing decomposition at normal

operating temperature. Accordingly, adsorbed moisture and oil ageing products (acids
particularly) have to be considered in order to estimate the degree of ageing. The
heated mass of conductor insulation (hotspots) that is subjected to accelerating
decomposition due to elevated temperature and contributes to formation of by-products,
comprises typically 2-10 % of the total mass of transformer insulation. Those heated
zones are usually inaccessible for visual inspection or sampling. However, water and
acids affect the outer layers of insulation, which are quite accessible for inspection.
Information about thermal distribution across the winding is vital to assess the ageing
state of insulation.

Based on these observations, a review of the methods used to assess the level of
contamination in the insulation of transformers is presented below.
3.2.1.2 METHODS FOR ASSESSING THE Q UALITY OF T RANSFORMER OILS

3.2.1.2.1 Dielectric Breakdown Strength (BDV)

This test measures the voltage at which the oil electrically breaks down. The test gives
a good indication of the amount of contaminants (water, dirt, oxidation particles, or
particulate matter) in the oil. The property is measured by applying a voltage between
two electrodes under prescribed conditions under the liquid. There are two ASTM
procedures: D-877, which specifies a test cup equipped with one-inch diameter vertical
electrodes that are 0.100 inch apart; and ASTM D-1816, which specifies a test cup
equipped with spherical electrodes spaced either 1 mm or 2 mm apart. This cup
includes a stirrer and is therefore sensitive to small amounts of particulates. In the latest
IEEE guide for acceptance and maintenance of insulating oils in equipment, it is stated
that the preferred method for assessing the dielectric breakdown of transformer oil is the
ASTM D-1816 (Note: this is at least 2000 or newer) method. This is because the
electrode configuration of the D-1816 method more closely approximates transformer
application. Moreover, the method provides a higher sensitivity to the presence of
particles and moisture that are detrimental to the operation of transformers.
3.2.1.2.2 Interfacial Tension (IFT)
This test (ASTM D-971-99a) is used to determine the interfacial tension between the oil
sample and distilled water. The oil sample is put into a beaker of distilled water at a
temperature of 25 °C. The oil should float because its specific gravity is less than that of
water. There should be a distinct line between the two liquids. The IFT number is the

80
amount of force (dynes) required to pull a small wire ring upward a distance of 1 cm
through the water/oil interface. A dyne is a very small unit of force equal to 0.000002247
pound. Good clean oil will make a very distinct line on top of the water and give an IFT
number of 40 to 50 dynes per centimeter of travel of the wire ring.

As the oil ages, it is contaminated by tiny particles (oxidation products of the oil and
paper insulation). These particles extend across the water/oil interface line and weaken
the tension between the two liquids. The more particles are present, the weaker the
interfacial tension and the lower the IFT number. The IFT and acid numbers together
are an excellent indication of when the oil needs to be reclaimed. Low IFT numbers are
an indication of highly contaminated oil, which can lead to sludging. If such oil is not
reclaimed, sludge will settle on windings, insulation, etc., and cause loading and cooling
problems.

There is definitely a relationship between the acid number, the IFT, and the number of
years in service. The accompanying curve (see Figure 3-1) shows the relationship and
is found in many publications (this chart was originally published in the AIEE
transactions in 1955). Notice that the curve shows the normal service limits both for the
IFT and the acid number.

3.2.1.2.3 Acid Neutralization Number

The acid number (acidity) is the amount of potassium hydroxide (KOH) in milligrams
(mg) that it takes to neutralize the acid in 1 gram (g) of transformer oil. The higher the
acid number, the more acid that is in the oil. New transformer oils contain practically no
acid. Oxidation of the insulation and oil forms acids as the transformer ages. The
oxidation products form sludge and precipitate out inside the transformer. The acids
attack metals inside the tank and form soaps (more sludge). Acid also attacks cellulose
and accelerates insulation degradation. Sludging has been found to begin when the
acid number reaches 0.40. At this point it is necessary to reclaim or replace the oil. The
acid number is measured using the latest version of ASTM method D974.

Figure 3-1 shows a plot of the relationship between acid number and interfacial tension
as a function of the number of normal years of service for a transformer.

81
 Figure 3-1: Interfacial Tension, Acid Number, and Years in Service

3.2.1.2.4 Power Factor

Power factor indicates the dielectric loss leakage current of the oil. A high power factor
indicates deterioration and/or contamination by-products such as water, carbon, or other
conducting particles; metal soaps caused by acids; attacking transformer metals; and
products of oxidation. The test method for power factor is the latest version of ASTM
D924, and the measurement is typically performed at 25 °C and 100 °C. Some ionic
contaminants can often pass undetected at 25 °C but will reveal their presence as
unacceptably high readings in the 100 °C test. ABB recommends always measuring the
oil power factor at both suggested temperatures. A high power factor at 25 °C and a low
power factor at 100 °C typically indicate the presence of moisture, since the moisture
will evaporate at 100 °C. On the other hand, a high power factor reading at both
temperatures or only at 100 °C typically indicates the presence of contaminants.
3.2.1.2.5 Test for Oxygen Inhibitor
Moisture is destructive to cellulose and even more so in the presence of oxygen. It is
therefore important to mitigate the effects of the presence of oxygen in transformer oil.
Oxygen inhibitors are the key to minimizing the effects of oxidation of oil. The two most
common inhibitors used are 2-6 ditertiary butyl para-cresol (DBPC) and ditertiary butyl
phenol (DBP). The first choice of attack by oxygen in the oil is the inhibitor molecules.
This keeps the oil free from oxidation and its harmful by-products. However, as the
transformer ages, the inhibitor is used up and needs to be replaced. Oxygen inhibitor
content is measured using the latest version of ASTM method D2668.
3.2.1.2.6 Furan Analysis
2-Furfuraldehyde and some related substances, all belonging to a group of chemical
compounds called furans, are formed when paper degrades. High furan content or a
high production rate may indicate a high rate of paper degradation. When DGA results
are not conclusive, furan analysis may aid the interpretation and give a more accurate

82
diagnosis. Section 3.3.2.2 provides a detailed discussion about analysis of furans in
transformers.
3.2.1.2.7 PCB Content
Environmental legislation often requires that oil contaminated with PCB is given special
treatment. For this reason service providers may sometimes refuse to handle oil that
has not been proven to be PCB-free. There may also be strict rules for the disposal of
PCB-containing oil.
3.2.1.2.8 Corrosive Sulphur
In recent years there have been a significant number of failures, in different types of
equipment, due to the formation of copper sulphide in the cellulosic insulation. Also,
other problems due to the action of corrosive sulphur compounds in oil have been
reported. It has become apparent that commonly accepted tests for corrosive sulphur
used in oil specifications (ASTM D1275 (copper strip) or DIN 51353 (silver strip)) are not
adequate. Several oils that have passed these tests have caused copper sulphide
formation in real life and in some cases have resulted in failure of the transformer.

New tests have been developed that have higher sensitivity and are more relevant for
the failure mechanisms involved. A new more severe copper strip test has been
introduced (ASTM D1275 method B), and a covered conductor deposition test (“CCD”)
has been developed to identify oils that may cause copper sulphide precipitation in
cellulosic insulation. A simplified version of the latter test is presently under
consideration as a new IEC standard test for corrosive sulphur.
3.2.1.3 MOISTURE IN T RANSFORMER INSULATION SYSTEMS [41]
The presence of moisture in a transformer deteriorates the transformer insulation by
decreasing both the electrical and mechanical strength. In general, the mechanical life
of non-upgraded Kraft paper insulation is reduced by the presence of moisture; the rate
of thermal deterioration of the paper is proportional to its water content [42]. Recent
studies performed by SINTEF Energy Research have shown that if normal life is defined
as ageing under dry, oxygen-free conditions, a moisture content of 1 % in non-upgraded
Kraft insulation can reduce life expectancy to 30 % of normal life. For 1 % moisture
content in thermally upgraded Kraft insulation, the life expectancy is approximately 60 %
of normal life. If the moisture content increases to 3-4 %, the life expectancy of the non-
upgraded Kraft insulation will drop to approximately 10 % of normal life expectancy and
thermally upgraded Kraft insulation will drop to approximately 25 % of normal life
expectancy [43].

Electrical discharges can occur in a high-voltage region due to a disturbance of the

moisture equilibrium causing a low partial discharge inception voltage and higher partial
discharge intensity [44]. Water in mineral oil transformers also brings the risk of bubble
formation when water from the surface of the cellulosic insulation migrates into the oil
and increases the local concentration of gases in the oil [45]. In the upcoming sections
we discuss the presence of water in the main components of insulation system: oil and
paper.

83
3.2.1.3.1 Transformer Oil
Mineral transformer insulating oils are refined from predominantly crude oils. The
refining processes could include solvent extraction, dewaxing, hydrogen treatment, or
combinations of these methods to yield mineral insulating oil that meets the
specification. It is mainly a mixture of hydrocarbon compounds of three classes:
alkanes, naphthenes, and aromatic hydrocarbons. These molecules have little or no
polarity. Polar and ionic species are a minor part of the constituents, but their presence
may greatly influence the chemical and electrical properties of the oil. Polar compounds
found in transformer oil usually contain oxygen, nitrogen, or sulfur. Ionic compounds are
typically organic salts found only in trace quantities.

Insulating oils, such as transformer oil, have a low affinity for water. However, the
solubility increases markedly with temperature for normally refined naphthenic
transformer oil. Water can exist in transformer oil in three states. In practical cases,
most water in oil is found in the dissolved state. Certain discrepancies in examining the
moisture content using different measurement techniques suggest that water also exists
in the oil, tightly bound to oil molecules (bound moisture), and especially in deteriorated
oil. When the moisture in oil exceeds the saturation value, there will be free water
precipitated from the oil in suspension or drops. Moisture in oil is measured in parts per
million (ppm) using the weight of moisture divided by the weight of oil (g/g).
3.2.1.3.2 Relative Humidity
Relative humidity can be defined in terms of the moisture –mixing ratio r versus the
saturation mixing ratio rs, %RH r rs which is a dimensionless percentage. Relative
humidity for air is the water vapor content of the air relative to its content at saturation.
Relative humidity for oil is the dissolved water content of the oil relative to the maximum
capacity of moisture that the oil can hold (the saturation limit). The higher the %RH, the
closer the oil is to saturation. In a transformer, it is preferable to keep the %RH below
10-20 %, depending on voltage class (see Figure 3-2 for moisture content curves at
different %RH).

84
 Figure 3-2: Relative Humidity Curves for Transformer Oil 12

NOTE: Below 30 °C, the curves are not very accurate.

3.2.1.3.3 Paper (Cellulose)

The following four terms are often used interchangeably in the context of solid
transformer insulation: pressboard, paper (or Kraft paper), transformer board, and
cellulose. Although in the context of particular transformer insulation they may indicate
different parts, e.g., paper tape, paper cylinders, transformer board cylinders, angle
rings, blocks, etc. In the context of moisture equilibrium, they all generally refer to
electrical-grade paper insulation manufactured from unbleached sulfate cellulose,
basically consisting of a long chain of glucose rings. Insulation paper used in
transformers can be completely dried, degassed, and oil impregnated. Insulation paper
can be manufactured to different densities, shapes, and other properties for different
applications.

Water in paper may be found in four states: adsorbed to surfaces, as vapor between the
cellulose fibers, as free water in capillaries, and as absorbed free water in the body of
the insulation. The paper can contain much more moisture than the oil. For example, a
150 MVA, 400 kV transformer with about seven tons of paper can contain as much as
223 kg of water. If it is assumed that such a transformer contains 80,000 liters of oil and
assuming a 20 ppm moisture concentration in oil, the total mass of moisture in the oil is
about 2 kg. This amount is much less than the moisture in the paper. The unit for
moisture concentration in paper is typically expressed in percent, which is the weight of
the moisture divided by the weight of the dry oil-free pressboard.

12
From IEEE Std 62-1995

85
3.2.1.3.4 Where Does the Water Come From
Moisture can be in the insulation when it is delivered from the factory. If the transformer
is opened for inspection, the insulation can absorb moisture from the atmosphere. If
there is a leak, moisture can enter in the form of water or humidity in air. Moisture is
also formed by the degradation of insulation as the transformer ages. Most water
penetration is the flow of wet air or rainwater through poor gasket seals due to pressure
differences caused by transformer cooling. During rain or snow, if a transformer is
removed from service, some transformer designs cool rapidly and the pressure inside
drops. The most common moisture ingress points are gaskets between bushing
bottoms and the transformer top and the pressure relief device gasket. Small oil leaks,
especially in the oil cooling piping, will also allow moisture ingress. With rapid cooling
and the resultant pressure drop, relatively large amounts of water and water vapor can
be pumped into the transformer in a short time. It is important to repair small oil leaks.
The small amount of visible oil is not important in itself, but it indicates a point where
moisture will enter the transformer.

It is critical for life extension to keep transformers as dry and as free of oxygen as
possible. Moisture and oxygen cause the paper insulation to decay much faster than
normal and form acids, sludge, and more moisture. Sludge settles on windings and
inside the structure, causing transformer cooling to be less efficient; therefore, the
temperature rises slowly over time. Acids cause an increase in the rate of decay, which
forms more acid, sludge, and moisture at a faster rate [46]. This is a vicious cycle with
increasing speed, forming more acid and causing more decay.
3.2.1.3.5 Moisture Equilibrium between Oil and Paper in Transformers
Since there is more water in the cellulose than in the oil and a significant part of the
protection of the transformer relies on the integrity of the cellulose insulation, it is
important to know the moisture in the cellulose. Unfortunately, this cannot be measured
directly without obtaining a sample of pressboard or paper from inside the transformer.
Methods have been developed to estimate the moisture of the cellulose insulation from
the moisture in the oil, based on the partitioning of water between the oil and the
cellulose under certain conditions. When the transformer is in equilibrium operation, this
provides a quick way of examining the moisture content in paper to predict future failure
by measuring the moisture in oil. A set of moisture equilibrium curves is shown in Figure
3-3. The original curves have been modified to include the insulation moisture limits for
different voltage classes of transformers. Given the average oil temperature of the
transformer and the measured moisture content of the oil, the moisture content of the
cellulose can be estimated from the chart in Figure 3-3. It can also be determined if the
moisture content is excessive and action is required.

Unfortunately, during regular operation of a transformer, the moisture in the oil and the
cellulose are never in equilibrium. Moisture constantly migrates from the cellulose into
the oil as the transformer load increases and the windings “heat” up. The reverse occurs
when the load is reduced and the transformer windings “cool” down. Equilibrium is
especially difficult to establish at low transformer temperatures. The situation improves
somewhat as the transformer oil temperature gets above 50 °C. It is important for users
of these curves to understand they may not be getting a true measure of the moisture in

86
the insulation. Advanced methods, such as the Dielectric Frequency Response (DFR)
analysis allow the direct measurement of moisture in the cellulose insulation. This
method is described in 3.3.3 of this handbook.

5.0 o o o o o
0C 10 C 20 C 30 C 40 C

4.5

4.0
o
50 C
3.5
% Moisture in Paper

IEEE C57.106-2002
3.0 Insulation Moisture
o
60 C Limits
2.5
69kV
70oC
>69kV - <230kV
2.0 230kV

o
80 C
1.5

1.0 90oC
o
100
0.5

0.0
0 5 10 15 20 25 30 35 40 45 50
Moisture in Oil (PPM)

Figure 3-3: Oommen Curves for Low Moisture Region of Moisture Equilibrium for Paper-Oil
Systems [47]. (Note: Moisture limits from C57.106-2002 and shown in Table 3-6 have been inserted
into the equilibrium plots.)

In order to obtain the average temperature of the transformer, it is advisable to measure

the temperature of the oil at the top and bottom oil sampling valves and then take an
average. It is also advisable to use a calibrated thermometer for these measurements
instead of relying on the readings of the temperature gauges.

The data from the moisture equilibrium curves and the recommended limits for moisture
in the solid insulation can be combined into a chart that gives the maximum allowed
equilibrium moisture in the oil at any given temperature and each voltage range. This
chart is shown in Figure 3-4. The chart indicates, for example, that at 60 °C the moisture
content in a 145 kV transformer at equilibrium should be no more than 30 ppm, whereas
for a 69 kV transformer the limit is approximately 65 ppm. Based on the measured
moisture in oil, the temperature, and the voltage class of a transformer, this chart can be
used to provide some indication of the moisture condition of a transformer.

87
 Maximum Recommended Moisture in Oil Based on
Recommended Maximum Moisture in Cellulose

100
69kV
90 >69 - <230kV
Moisture Limit in Oil (ppm)

80 230kV
70
60
50
40
30
20
10
0
0 10 20 30 40 50 60 70 80 90 100
o
T emp ( C)

Figure 3-4: Maximum Recommended Moisture in Oil versus Temperature

3.2.1.3.6 Cautions in Estimation of Moisture Using Moisture Equilibrium Curves

As discussed above, the moisture content of the oil samples taken from transformers
can be measured using the Karl Fischer method. The moisture in the board is read from
the equilibrium curves by projecting the measured moisture in oil onto the
corresponding measurement temperature curve. There is potential for significant errors
in this method at low temperatures and for low oil moisture contents due to the
steepness of the equilibrium curves in this region. For example, if the measured
moisture in oil is 10 ppm, and considering a measurement error of ±2 ppm, the moisture
can range from 3.2 - 4.0 % by weight at 20 °C and between 0.8 - 1.1 % at 60 °C (see
Figure 3-5). The spread is smaller at higher temperatures and much worse at lower
temperatures. If this method is to be used, the temperature of the insulation must be at
least 50 °C in order to get reliable results. There is also always the question about
whether the transformer is ever in equilibrium during normal operation. If there are
concerns about the moisture content of the insulation, it is advisable that advanced
diagnostic methods, such as dielectric frequency response, be used.

88
 5.0 o o o o o
0 C 10 C 20 C 30 C 40 C

4.5

4.0
o
50 C
3.5
% Moisture in Paper

3.0
o
60 C

2.5

o
70 C
2.0
o
80 C
1.5
o
90 C

1.0 100
o

0.5

0.0
0 5 10 15 20 25 30 35 40 45 50
Moisture in Oil (PPM)

Figure 3-5: Moisture Estimation Using Equilibrium Curves

3.2.1.4 LIMITS FOR MEASUREMENT OIL QUALITY PARAMETERS [48]

The following tables (Table 3-5 and Table 3-6) show the various limits for assessing
moisture in a transformer as set forth in the IEEE Std. C57.106-2002. These limits can
be used as guidelines in making maintenance decisions about transformers. For
example, if the %RH of water in the oil is greater than 30% and the corresponding
moisture in the cellulose is greater than the limit specified for the voltage class, the
transformer insulation may need to be dried. It would be advisable in this situation to
contact ABB. Since a dry-out is an expensive process, advanced diagnostic methods,
such as Dielectric Frequency Response Analysis (DFR), can be applied to directly verify
the insulation moisture measurement. An independent assessment of a fresh sample of
oil would also be made to reassess the original diagnosis.

Table 3-5: General Guidelines for Interpreting Data Expressed in Percent Saturation

Percent Saturation, Water-in-Oil . Condition of Cellulosic Insulation

0-5 Dry insulation
6-20 Moderate—wet, low numbers indicate fairly dry to
moderate levels of water in the insulation. Values
toward the upper limit indicate moderately wet
insulation.
21-30 Wet insulation
>30 Extremely wet insulation

89
Table 3-6: Recommended Maximum Limit of Water Content in Mineral Insulating Oil of Operating
Gas Blanketed, Sealed, or Diaphragm Conservator Transformers a
b
Average Oil Suggested Maximum Water Contents in mg/kg and Percent Saturation
Temperature 50°C 60°C 70°C
c c c
mg/kg % saturation mg/kg % saturation mg/kg % saturation
69 kV 27 15 35 15 55 15
>69 - <230 kV 12 8 20 8 30 8
230 kV and 10 5 12 5 15 5
greater
NOTES
1 - These values are, by necessity, approximate but are adequate for maximum water-in-oil guides.
2 - The oil sample should, if practical, be taken when the load and oil temperatures have been relatively
constant for 48 h. The intent is to obtain a sample when the moisture content in the transformer is close to
equilibrium. If the load and/or ambient are variable, the oil temperature can be maintained relatively constant
by controlling the amount of cooling in operation. If you are confident that the temperature gauges are in
calibration, then record the top oil temperature at the time that the sample is taken. For Oil Natural Air
Natural (ONAN) and Oil Natural Air Forced (ONAF) ratings, subtract 10 °C from the top oil to obtain the
average oil temperature. If you are unsure of the gauge accuracy, record the actual sample temperature and
add 5 °C to approximate the average oil temperature.
3 - The above values are based on the following approximate percent by weight of water in solid insulation
values (see IEEE Std 62-1995):
69 kv 3% maximum
>69 - <230 kv 2% maximum
230 kv and greater 1.25% maximum
4 - Saturation values (mg/kg) at 100% saturation:
50 °C - 175 / 60 °C - 245 / 70 °C - 335
a) The data in this table is from sealed transformers and may also apply to free-breathing type transformers.
b) Calculated from formulas 1 and 2 in Clause 44 from Bruce, C. M., Christie, J. D., and Griffin. Paul [49]
c) Equivalent measurement is parts per million, ppm.

Table 3-7 and Table 3-8 are the recommended limits for oil quality tests performed on
new and service aged transformers (always refer to the latest IEEE standards for the
current suggested limits). Note that these are the suggested limits for acceptable
conditions. If any measurements are beyond the suggested limits, it is advisable to take
another sample to confirm the first result. If the results are confirmed, it is recommended
you contact ABB for advice on further action. Table 3-9 provides some guidelines on
actions to be taken based on the results of oil quality measurements.

Table 3-7: Test Limits for New Mineral Insulating Oil Received in or Processed for New Equipment

Test and Method Value for Voltage Class

69 kV >69 - <230 kV 230 kV - <345 345 kV and above
kV
a
Dielectric strength ,
ASTM D1816-97,
kV minimum,
b
1 mm gap : 25 30 32 35
b
2 mm gap : 45 52 55 60
Dissipation factor (power factor),
ASTM D924-99e1,
25°C, % maximum: 0.05 0.05 0.05 0.05
100°C, %maximum: 040 040 0.30 0.30
Interfacial tension,
ASTM D971-99a, mN/m 38 38 38 38
minimum:
Color,
ASTM D1500-98, ASTM units 1.0 1.0 1.0 0.5
maximum:
Visual examination, Bright and Bright and clear Bright and clear Bright and clear
ASTM D1524-94 (1999): clear

90
Test and Method Value for Voltage Class
69 kV >69 - <230 kV 230 kV - <345 345 kV and above
kV
c c c c
Neutralization number (acidity), 0.015 0.015 0.015 0.015
ASTM D974-02,
mg KOH/g maximum:
Water Content, 20 10 10 10
ASTM D1533-00, mg/kg
d
maximum :
Oxidation inhibitor content when
specified,
ASTM D2668-96,
Type I oil, % maximum: - 0.08
Type I oil, % minimum: - 0.0
Type II oil, % maximum: 0.3 0.3
Type II oil, % minimum: >0.08 >0.08
Total dissolved gas, 0.5% or per 0.5% or per manufacturer’s
e
ASTM D2945-90 (1998): manufacturer’s requirements
e
requirements
a) Oil dielectric testing in accordance with ASTM D877-00 has been replaced by ASTM D1816-97.
b) Alternate measurements of 0.04 in and 0.08 in respectively for gaps.
c) This value is more stringent than the ASTM D3487 requirement.
d) Equivalent measurement is parts per million, ppm.
e) This value should be obtained from a sample collected 24 to 48 hrs after the transformer is filled and applies only to
transformers with diaphragm conservator systems.

Table 3-8: Suggested Limits for Continued Use of Service-Aged Insulating Oil

Test and Method Value for Voltage Class

69 kV >69 - <230 kV 230 kV and above
a
Dielectric strength ,
ASTM D1816-97, kV minimum,
b
1 mm gap : 23 28 30
b
2 mm gap : 40 47 50
a
Dissipation factor (power factor) ,
ASTM D924-99e1,
o
25 C, % maximum 0.5 0.5 0.5
o
100 C, % maximum 5.0 5.0 5.0
Interfacial tension,
ASTM D971-99a,
mN/m minimum 25 30 32
Neutralization number (acidity),
ASTM D974-02,
mg KOH/g maximum 0.20 0.15 0.10
Water content Refer to Table 3-6
Oxidation Inhibitor Content,
ASTM D2668-96,
Type II Oil 0.09% minimum, if in original oil.
a) Older transformers with inadequate oil preservation systems or maintenance may have lower values.
b) Alternate measurements of 0.04 in and 0.08 in respectively for gaps.

91
 Table 3-9: Maintenance Guidelines for In-Service Oils [50]

o
Power Factor Results at 25 C Suggested Action

0.5% Acceptable
>0.5% but 1.0% Investigate. Oil may require replacement or clay treatment.
>1.0 but 2.0% Investigate. Oil may cause failure of equipment. Oil may require replacement or
clay treatment.

2.0% Remove from service. Investigate. Oil may require replacement or clay treatment.

Neutralization (mg KOH/gm) Suggested Action

Results

<0.05 Acceptable

0.05 but <0.15 Clay treat or replace at convenience. For 345 kV, clay treat or replace oil in
immediate future.

0.15 but <0.50 Clay treat or replace oil in immediate future.

0.50 Replace oil.

IFT (dynes/cm) Results Suggested Action

25 Acceptable
22 but <25 Clay treat or replace at convenience. For 345 kV, clay treat or replace oil in
immediate future.

16 but <22 Clay treat or replace oil in immediate future.

<16 Replace oil.

3.2.1.5 MOISTURE AND BUBBLE EVOLUTION IN T RANSFORMERS

Water in a transformer reduces the insulation capability in the active part. Water affects
the electric strength, power factor, ageing, losses, and mechanical strength of the
insulation [51,52].

Not only does moisture in the cellulose decrease the breakdown strength of the
insulation system and increase the ageing process, there is also potential danger due to
enhanced chances of partial discharge activity and eventual breakdown of the
insulation.

Bubbles in a transformer may arise from several causes: 1) excessive gas generation
from faults, 2) nitrogen supersaturation in the case of gas-blanketed units, and 3)
gas/vapor release from overload conditions, particularly for paper insulated systems
such as large and medium power transformers. In experiments on gas evolution
performed at ABB [53, 54, 55], the following key observations were made:

o Bubble evolution temperature decreased exponentially with increasing moisture

content.
o Bubble evolution temperature decreased significantly with increasing gas content
of oil at high moisture levels in the cellulose insulation.

92
The studies revealed that bubble evolution in paper-wrapped windings under overload
conditions is significantly influenced by the moisture in paper which tends to be released
as bubbles. At low moisture levels in paper, systems with low gas content and gas
saturated systems behave somewhat similarly. It appears the dissolved gas is not the
determining factor for bubble generation. Indeed, the data showed that bubble
evolution from overload conditions may not happen below 200oC in very dry
transformers, regardless of the gas content. A service aged transformer with two
percent moisture may release at 140oC when overloaded. An empirical mathematical
relationship to predict bubble evolution temperature [56] is shown graphically in Figure
3-6.
200

Values are calculated

for 1 atmosphere

180

Gas
C

Conte nt
o

0%
Bubble Evolution Temperature,

1%
160 2%
Ze ro ga s conte nt 3%
systems 4%
5%
6%
7%
8%
140
9%

Observe d for N2
sa turated system s

120

100
0.0 1.0 2.0 3.0 4.0
% Moisture in Coil

Figure 3-6: Bubble Evolution Temperature vs. Moisture Content in Paper and Gas Content in Oil

If the loading guidelines suggested by IEEE Std C57.91 for transformers under various
load conditions are superimposed on Figure 3-6, some rather critical decisions can be

93
made for what transformers can be operated under what load conditions. The resulting
chart is shown in Figure 3-7.

200

190

180

Zero gas
Normal Life Expectancy Loading
C

170 conte nt
o
Bubble Evolution Temperature,

Planned Loading Beyond

Nameplate
160
Long-term Emergency Loading

Short-term Emergency Loading

150 Observe d for
N2 sa turated
syste ms

140

130

120

110

100
0.0 1.0 2.0 3.0 4.0
% Moisture in Coil

Figure 3-7: Loading Guidelines Based on Moisture Content of Cellulose Insulation

The loading guidelines shown in Table 3-10 can be derived from Figure 3-7 and IEEE
C57.91. The table should be read as follows: a transformer with approximate gas
content of 9 % and moisture content of up to 2.0 % can be operated under long-time
emergency conditions so long as the hottest spot temperature never exceeds 140 °C.
However, it should never be operated under short-term emergency conditions. Another
important observation is that transformers with insulation moisture content greater than
0.8 % may be exposed to significant risk of failure if operated under short-term
emergency loading conditions.

94
 Table 3-10: Loading Limits Based on Moisture Content

Cellulose Moisture
Hottest Spot
(%) Overload Level
Temperature Overload Type
Zero gas N2 saturated with 40°C Ambient
(°C)
content system system
120 3.9 3.3 Normal Loading 0%
130 2.9 2.6 Planned O/L Beyond N/P 6%
Long Time Emerg.
140 2.2 2.0 12%
(1-3 mo.)
Short-Time Emerg.
180 0.8 0.8 40%
(½ -2 hrs)

A word of caution should be given here regarding the preceding discussion. It is our
experience that an accurate determination of the transformer hotspot temperature,
especially on older transformers, can only be made after an updated engineering
calculation using modern design programs. Relying on readings from hotspot gauges or
on test reports may result in significant underestimation (or in some cases
overestimation) of the true hotspot temperatures. Also it is important to get a proper
measure of the moisture content of the paper insulation before subjecting a transformer
to overload conditions. At present, the Dielectric Frequency Response method (see
section 3.3.3.4) is the most accurate means of estimating the moisture content of the
paper insulation in transformers.

For most transformers, especially those that are continuously loaded, a more significant
effect of moisture in the insulation is the increased ageing associated with the moisture
in the cellulose insulation. Ageing calculations given in IEEE Std C57.91 assume dry,
oxygen-free insulation. Dry insulation is assumed to be approximately 0.5% moisture or
less. Field measurements done by ABB have demonstrated that most transformers in
the utility network have moisture levels higher than this. Since the ageing rate of
insulation is dependent on the temperature, the moisture level in the insulation, and the
oxygen level in the oil, the actual ageing rates are often much higher than might be
assumed for normal dry transformers.

95
 3.2.2 DISSOLVED GAS IN OIL ANALYSIS (DGA) [57]

3.2.2.1 INTRODUCTION
For many years the method of analyzing gasses dissolved in the oil (DGA) has been
used as a tool in transformer diagnostics. The method has been used for several
purposes: to detect incipient faults; to supervise suspect transformers; to test a
hypothesis or explanation for the probable cause of failures or disturbances which have
already occurred; and to ensure that new transformers are healthy. DGA could also be
used as part of a scoring system in a strategic ranking of a transformer population.

What is said about DGA for transformers is also applicable to reactors, instrument
transformers and bushings. It is worth noting that DGA is a fairly mature technique and
is employed by several ABB transformer companies around the world either in own
plant or in co-operation with affiliated or independent laboratories. In assessing
dissolved gases in oil, the rate of increase of different gases during a time interval is
the most important indicator of the health of the unit. The actual gas levels may of may
not be of consequence for the operation or the health of the transformer.

The idea behind the use of dissolved gas analysis is based on the fact that during its
lifetime, all oil/cellulose insulated systems generate decomposition gases under the
influence of various stresses - both normal and abnormal. The gases that are of interest
for the DGA analysis are shown in Table 3-11.

Table 3-11: Dissolved Gases in Mineral Oil-filled Transformers

Gas Symbol Comments

Hydrogen H2
Methane CH4
Ethylene C2H4
Ethane C2H6
Acetylene C2H2
Propene C3H6 Not used under ANSI/IEEE
standards
Propane C3H8 Not used under ANSI/IEEE
standards
Carbon monoxide CO
Carbon dioxide CO2
Oxygen O2
Nitrogen N2
Total dissolved combustible TDCG (=H2+CH4+C2H4+C2H6 +C2H2+CO)
gases

All these gases except oxygen and nitrogen may be formed during the degradation of
the insulation. The amount and the relative distribution of these gasses depend on the
type and severity of the degradation and stress.

96
Over the years several different schemes have been proposed as evaluation schemes
for DGA. Several of these techniques are presented in the IEEE Standard C57.104 and
IEC Publication 60599.

A number of faults can not be detected by DGA. One example is faults that are not in
contact with the oil. Other examples are faults in which only very small energies are
released or in which the energy is spread over a large surface or large volume. Such
faults are typically associated with sporadic discharges or weak discharges.
3.2.2.2 PROCEDURE
The procedure for performing DGA consists of essentially four steps:
- Sampling of oil from the transformer
- Extraction of the gases from the oil
- Analysis of the extracted gas mixture through gas chromatography.
- Interpretation of the analysis according to an evaluation scheme.
3.2.2.3 SAMPLING
Suitable locations for sampling are valves in the cooler/radiator circuit. Because of
design limitations it may not always possible to take samples from these locations.
Other places from which to draw samples are the cover, bottom valve, the conservator
and from the Buchholz relay. In addition, care must be taken to make sure the sample is
not exposed to the atmosphere and that gases are not lost during sampling or
transportation to the laboratory. For more general information about sampling of gases
refer to the latest version of IEC Standard 60567 or ASTM Standard 3613. Figure 3-8
shows the sampling methodology used by ABB.
3.2.2.4 EXTRACTION
The removal of the gases from the oil can be accomplished by various methods:
- Partial degassing (single-cycle vacuum extraction)
- Total degassing (multi-cycle vacuum extraction)
- Stripping by flushing the oil with another gas.
- The head-space technique in which gases are “equalized” between a free gas
volume and the oil volume.
3.2.2.5 ANALYSIS
After extraction the gas mixture is fed into adsorption columns in a gas chromatograph
(GC) where the different gases are adsorbed to various degrees and reach the detector
after different periods of time. In this way the gas mixture is separated into individual
chemical compounds and their concentrations are calculated in volume gas at standard
temperature and pressure (STP) per oil volume and expressed in parts per million
(ppm).

It should be emphasized that this extraction and analysis may involve analytical errors.
It may therefore be difficult to directly compare results from two different laboratories.
One should not jump from one lab to another but instead try to stick with one well-
reputed lab.

97
 SAMPLING OF OIL FOR GAS ANALYSIS
Important things to consider :
The syringe piston must be clean at use.
Used hoses shall not be returned to ABB Transformers.
Please remember to note the number of the syringe in the questionaire.

Connect the hose and T-piece to the syringe according to the Connect the hose from the sampling valve to the T-piece.
picture.

Put the hose with the T-piece in a bucket and open the valve Turn the handle on the syringe valve as in the picture and
on the transformer. Flush min. 3 times the valve and hose suck carefully in about 15 ml. of oil into the syringe.
volume. Let the oil flow during the sampling.

Hold the syringe so that the valve points upwards and press Suck carefully 20 ml of oil into the syringe. No air bubbles
the air and oil out. No airbubbles should be left. shall be seen in the syringe.
Close the valve on the syringe by turning the handle on the
Setfo/ta 980903 KR syringe valve as in the picture.
Gasanalys provtagning engelsk

Figure 3-8: ABB Method for Sampling Oil for Gas Analysis

98
3.2.2.6 INTERPRETATION
In order to properly interpret the results of the gas analysis, it is necessary to determine
the gas production rate for the period under consideration, i.e. how much the gas levels
have changed over a given time period. The absolute gas levels seldom give a sufficient
good basis for the interpretation.
3.2.2.7 AIR
Oxygen (O2) and nitrogen (N2) come from the air. Air contains about 20% oxygen and
about 80% nitrogen. The levels in the oil could be respectively 30,000 and 80,000 ppm
at air saturation. Oxygen and nitrogen have different solubility in oil. It is unusual to
measure oxygen levels below 1,000 ppm and nitrogen levels below 2,000 ppm. The air
content may be used to check the sampling procedure. The air content must not jump
up and down between subsequent samples. If that is the case, one can suspect that the
samples have not been taken with sufficient accuracy. The oxygen level could decrease
at high temperatures of the oil. Oxygen is also consumed during periods of strong
ageing of oil and cellulose.

A small amount (up to 200 ppm) of carbon dioxide, CO2 may also come from air, but
only if the oil is saturated with air (around 10%).
3.2.2.8 GAS SPECTRUM – TYPES OF F AULTS

3.2.2.8.1 Hot Metal Surface

The hydrocarbons: methane (CH4), ethylene (C2H4), ethane (C2H6), propene
(C3H6), propane (C3H8) etc. are mainly produced from hot oil.
Acetylene (C2H2) is not produced until temperatures close to 1000 ºC. One
example is glowing spots due to circulating currents in the core.
The oil boils at around 320 ºC. This means that that it is difficult to obtain a stable
temperature on a metal surface above this temperature limit.
The oil starts to degrade already at 80-100 ºC, even if the degradation rate is
very slow.
One needs a higher temperature to form for example C2H4 and C3H6 than CH4 and
C3H8. More gas is formed at higher temperatures. This together makes it possible to use
ratios between hydrocarbons to get an estimation of the temperature around the fault.
3.2.2.8.2 Examples of Hot Metal Surfaces
The following are examples of situations in a transformer that could result in hot metal
faults:
A bolted joint which has lost totally or partly its clamping force
A very high resistance between the cleats and leads and the bushing.
A damaged draw rod or a wrongly assembled draw rod that makes a bad contact
at the connection.
Bad contact in soldered or welded leads.
When there is a current running in the draw rod of the bushing.
Sliding contacts from the selector that becomes hot with time.
Currents due to stray fluxes in the tank.

99
 Inadvertent grounds that create circulating currents.
Increased resistance of the selector contacts for the tap changer.
Circulating currents in the core. A low resistance between different core steel
packages or to metallic parts or to high burrs on the sheets.
Induced currents due to non compensated currents in the core window.
Currents in metal pieces which should have been insulated or which have
damaged insulation. Consider which joints there are in the unit, core bolts, etc.
Closed loops for currents because of damaged insulation between parallel
conductors.
The insulation of the steel band around the core becomes damaged.
3.2.2.9 OVERHEATED CELLULOSE
Carbon oxide (CO) and carbon dioxide (CO2) come mainly from hot cellulose. They are
produced at moderate temperatures (< 150 ºC) with the ratio CO/CO2 = 0.3.
3.2.2.9.1 Examples of Overheated Cellulose
Overheated conductor insulation
Insulated multiple grounds which conduct a high current
Parallel conductors with common covering which come into electrical contact with
each other
Conductors for the cleats and leads
Winding conductors, obstructed cooling, loosened/wrongly positioned oil guiding
ring
Overcurrents because of leakage fields
Circulating currents in the yoke bolts
Any of the conditions in the “Hot metal surface” list that involve surfaces that are
covered with cellulose.
3.2.2.10 ELECTRICAL FAULTS
Electrical faults mainly produce hydrogen (H2) and acetylene (C2H2). For a low energy
partial discharge, hydrogen is the main gas that is generated. For a high energy partial
discharge, acetylene and other hydrocarbons may also be found.
3.2.2.10.1 Examples of Electrical Faults
When a joint used for equalizing a potential becomes lose, one end can be at a
floating potential with partial discharges. Sometimes this fault can include
overheating of the cellulose
Continuous strong partial discharges between parallel conductors with a certain
potential difference. A strong partial discharge will sooner or later lead to a
flashover
Break in a soldered connection which cause partial discharges
Floating potential, shielding ring, toroids
Partial discharges between turns/conductors which are next to each other
Partial discharges due to inadequate impregnation or air bubbles enclosed in the
insulation.

100
3.2.2.11 FACTORS AFFECTING GAS CONCENTRATION IN TRANSFORMERS

3.2.2.11.1 Type and Brand of Oil

Recently it has been shown that different oils show different gassing patterns. In
particular, some additives, for example oxygen inhibitors, influence the gassing pattern.
3.2.2.11.2 Oxygen
It has long been known that the concentration of oxygen has an impact on the ageing of
materials. The ageing of both the solid and liquid insulation materials has an impact on
the gassing rate. It has been experienced that the factor of the gassing with/without
oxygen is dependent on temperature.
3.2.2.11.3 Load
An increase in the load gives directly an increase in the temperature. A higher
temperature gives a higher gassing rate.
3.2.2.11.4 Oil Preservation Systems
Presently, state-of-the-art gas analysis is done mostly on oil samples taken from
transformer units. The interpretation of gas analysis results is based on gas-in-oil
composition. Under identical conditions, a transformer with gas space allows part of the
gases to be distributed into the gas space. Therefore, the gas concentration in oil would
be less than the total gas generated. The three main types of oil preservation systems
are illustrated in Figure 3-9. It is readily seen that only Type II comes close to preserving
all the gases in the oil. While both Types I and II are sealed systems, Type III allows
gases to be lost to the atmosphere.

Figure 3-9: Oil Preservation Systems for Power Transformers

If there are increasing levels of nitrogen, oxygen, and carbon dioxide in a conservator
type transformer, there is a possibility the tank has a leak or the oil may have been
poorly processed. In this case, it is advisable to check the diaphragm or bladder for
leaks and to check for oily residue around the Buchholz relay and other gasketed
openings. There should be fairly low nitrogen and especially low oxygen in a

101
conservator type transformer. With time some air could leak through the bladder and
raise the oxygen and nitrogen levels.
3.2.2.11.5 Gas Mixing
Concentration of gases in close proximity to an active fault will be higher than in the
DGA oil sample. As distance increases from a fault, gas concentrations decrease. Equal
mixing of dissolved gases in the total volume of oil depends on time and oil circulation. If
there are no pumps to force oil through radiators, complete mixing of gases in the total
oil volume takes longer. With pumping and normal loading, complete mixing equilibrium
should be reached within a few days and will have little effect on DGA if an oil sample is
taken then or long after a problem begins.
3.2.2.11.6 Temperature
There is an old chemist's rule of thumb stating that a small increase in temperature (5-
15 ºC) can yield a two or threefold increase in gassing rate. The basic explanation of
this phenomenon is found in the well-known Arrhenius equation, which holds true for
most chemical reactions.

Gas production rates increase exponentially with temperature and directly with volume
of oil and paper insulation. Temperature decreases as the distance from the fault
increases. Temperature at the fault centre is highest, and oil and paper there will
produce the most gas. As distance from the fault increases the temperature decreases,
and the rate of gas generation also decreases. Because of the volume effect, a large
heated volume of oil and paper can produce the same amount of gas as a smaller
volume at a higher temperature. It is impossible to tell the difference by just analyzing
the DGA. It is important to note that the ambient temperature directly influences the
gassing rate. If there is a fault, the higher the ambient temperature, the higher would be
the gassing rate.

A gas generation chart [58] [59] is shown in Figure 3-10. Note that temperatures at
which gases form are only approximate. Moreover, the figure is not drawn to scale and
is only to be used for purposes of illustrating temperature relationships, gas types, and
quantities as fault temperature vary in a transformer. These relationships represent
what generally has been proven in controlled laboratory conditions using a mass
spectrometer.

The vertical band at left side of the chart shows what gases and approximate relative
quantities are produced under partial discharge conditions (low energy discharge
events). The total hydrogen produced by a partial discharge in oil could be as much as
75% of the total gases, the remaining part being composed of small percentages of
hydrocarbons, in decreasing order C2H2 > CH4 > C2H4 > C2H6. With paper or
pressboard added to the system, some CO is also produced. Discharges in cellulose
alone produce CO and H2 in large quantities, in approximately equal quantities.

Various gases begin forming in a transformer at specific temperatures. From Figure

3-10 we can see relative amounts of gas as well as approximate temperatures.
Hydrogen and methane begin to form in small amounts around 150 °C. Methane (CH4),

102
ethane (C2H6), and ethylene (C2H4) production peaks at certain temperatures and
declines as temperature increases beyond the peak. At about 250 °C, production of
ethane (C2H6) starts. At about 350 °C, production of ethylene (C2H4) begins. This
suggests that low temperature thermal faults will produce virtually no ethylene, but
plenty of ethane and methane. Acetylene (C2H2) starts above 700 °C. This indicates that
a thermal fault of greater than 700 °C can produce trace amounts of acetylene. Larger
amounts of acetylene may only be produced above 900 °C and by internal arcing.

The C2H4/C2H6 ratio is a good indicator of the hotspot temperature for mild to moderate
cases of overheating. The following expression is generally used as an approximation of
the oil decomposition temperature in terms of the C2H4/C2H6 ratio [60]:

C2 H 4
T ( o C ) 100 150
C2 H 6

Figure 3-10: Combustion Gas Generation versus Temperature

3.2.2.11.7 Gas Solubility in Oil

Transformers with gas space above oil have the possibility of distribution of gases
between the liquid and gas space. These gases, except for the nitrogen in the gas
space and trace amounts of oxygen, are generated during transformer operation and
afterwards distribute between the oil and gas space according to the laws of distribution.
In a closed system, if gas generation proceeds at a slow rate, and mixing is effective,
equilibrium is attained soon. The deciding factors in gas distribution are the solubility of

103
the gas in the liquid medium and the prevailing temperature. The more soluble gases
would be found in a higher proportion in the oil than the less soluble ones. On the other
hand, the less soluble gases would be found in a higher proportion in the gas space.

The solubility of gases in oil varies with temperature and pressure. The solubility of all
transformer gases increase proportionally with pressure. The solubility of hydrogen,
nitrogen, carbon monoxide, and oxygen increases with temperature. The solubility of
carbon dioxide, acetylene, ethylene, and ethane decreases with increasing temperature.
The solubility of methane remains almost constant with temperature. Figure 3-11 shows
the distribution coefficient (or Ostwald coefficient) of gases at 1 atmosphere. These
coefficients are used to compute the gas space concentration corresponding to the
concentration in oil and vice versa.

10
Gas Solubility Coefficients in Oil

C2H6
C2H4
1 CO2 C2H2

CH4

O2
CO
N2
0.1 H2

0.01
0 20 40 60 80 100
o
Temperature ( C)

Figure 3-11: Gas Distribution Coefficients at 1 Atmosphere

From the chart it is clear that the solubility of acetylene in oil is much greater than that of
hydrogen in oil. Indeed at 25 °C and 1 atmosphere, the solubility of acetylene is 122 %
and that of hydrogen is 5.6 %. It is clear that transformer oil has a much greater
capacity for dissolving acetylene than hydrogen. It should be noted that gas from the
gas space is lost as the pressure in the gas space is released.
3.2.2.11.8 Other Factors
Below is a list of factors that are known to influence the gassing rate. However, there is
presently no consensus on how the individual factors affect the gassing rate.

104
- Temperature distribution in the oil and in the cellulose
Since the gassing is strongly temperature dependent, the temperature
distribution will be important for the gassing.
- Average winding temperature
When the temperature distribution is not exactly known, the average
winding temperature rise could be a good approximation.
Ambient temperature
Governs the absolute temperatures in the transformers
Oil production process
It has been shown that how the oil is manufactured can influence the
gassing. The oil production process could be more or less harmful to the
oil.
- Transformer history
What the transformer has gone through could be accumulated in the
insulation. The most common cases are when gasses are dissolved in the
cellulose and released at degassing or at temperature changes
- Repair
- Tests
No load loss test
Electrical tests
- Unaged insulation material
New cellulose has weak links in the material, which are cut early in the
ageing process, giving higher gassing rates in the beginning
- Type of cellulose insulation: The manufacturing processes and the ingredients in
the board have an influence on the gassing rate
Kraft, Insuldur, Thermally upgraded
Pressboard
- Low density, High density
- Laminated wood
- Different manufacturers
- Laminated polyester or casein glued board

- Type of design: Since the gassing is measured in ppm/day or ml/day, it is of

importance the volumes of oil, solid insulation and its ratio.
Size = rating
Oil volume
Solid insulation volume
- Design materials: It has been shown that many design materials have impact on
gassing. Most famous is perhaps inadequately cured epoxy inside radiators and
reactor cheeses that produces hydrogen. Another famous type of material is the
catalytic material. Among this group are zinc and stainless steel in transformers;
as well as core steel insulation. These materials also enhance hydrogen
production :

Glue, Epoxy
Paint

105
 Zinc
Stainless steel
- Phenomenon
Transport in and out of insulation: It has been shown that the solubility of
carbon oxide (CO) and carbon dioxide (CO2) is temperature dependent.
This means that the content of these gases will change when the
temperature changes. These gases will go out into the oil to a certain
extent when the oil gets colder.
“Sweating”: If the level of a particular gas in the solid insulation is high, it
could take a substantial amount of time before the gas in the insulation is
in equilibrium with the gas in the oil.
3.2.2.12 DGA INTERPRETATION METHODS

3.2.2.12.1 Key Gas Method of Interpreting DGA

In this method, one looks for the most prominent gas - the one which differs most from
an expected "normal" level (or change). For example, during overheating of cellulose
the main decomposition gases are CO and CO2. During a partial discharge or corona
activity, H2 is formed. If the partial discharges involve cellulose, carbon oxides will be
present as well. During a more severe electric discharge, for example arcing, C2H2 will
be produced. Normally H2 and smaller amounts of CH4 and C2H6 will also be produced
during arcing discharges. Further, if cellulose is involved in the fault, CO will be
produced. If oil is overheated, the hydrocarbons are the main gases produced –
normally the saturated hydrocarbons such as C2H6 at lower temperatures and
unsaturated hydrocarbons such as C2H4 at higher temperatures. At very high
temperatures, overheated oil will produce C2H2.

CO2, O2 and N2 can also be absorbed from the air if there is an oil/air interface or if there
is a leak in the tank. For Type I preservation systems that have a nitrogen blanket,
nitrogen in the oil may be near saturation. As described above, each key gas is
identified with a certain type of fault. There are four fault patterns that can be associated
with key gases as shown in Table 3-12. The key gas is frequently the predominant gas
in the mixture of generated gases in the oil, but occasionally another gas could be in
high concentration. Such variations are possible, because over a wide range of
temperatures each gas attains a maximum generation rate at a certain temperature.
Depending on the temperature present at the fault site, one gas or the other may be in
larger proportion.

It should be noted that small amounts of H2, CH4, CO2, and CO are produced by normal
ageing. Thermal decomposition of oil-impregnated cellulose produces CO, CO2, H2,
CH4, and O2. Substantial decomposition of cellulose insulation begins at only about
100°C or less. Faults will produce internal hotspots of far higher temperatures than
these, and the resultant gases show up in the DGA.

Table 3-12: Key Gas and Fault Type Guide

106
 Fault Pattern Key Gas Secondary Gases Possible Findings
Discoloration of paper insulation.
CH4 and C2H4 if the fault Overloading and/or cooling
Conductor CO2/CO (Carbon
involves an oil-impregnated problem. Bad connection in leads
Overheating Oxides) or tap changer. Stray current path
structure
and/or stray magnetic flux.
CH4 and smaller quantities of
Metal discoloration. Paper
H2 and C2H6. Traces of C2H2 if
Oil Overheating C2H4 (Ethylene) insulation destroyed. Oil heavily
fault is severe or involves carbonized.
electrical contacts.
Weakened insulation from ageing
and electrical stress. Pinhole
punctures in paper insulation with
CH4 and minor quantities of
Partial Discharge H2 (Hydrogen) carbon and carbon tracking.
C2H6 and C2H4 Possible carbon particles in oil.
Possible loose shield, poor
grounding of metal objects.
Metal fusion, (poor contacts in tap
changer or lead connections).
H2, and minor quantities of Weakened insulation from ageing
Arcing C2H2 (Acetylene) and electrical stress. Carbonized
CH4, C2H4
oil. Paper destruction if it is in the
arc path or is overheated.

3.2.2.12.2 Individual and Total Dissolved Key-Gas Concentration Method

A four-condition DGA guide to classify risks to transformers with no previous problems
has been developed in IEEE C57.104 [61]. The guide uses combinations of individual
gases and total combustible gas concentration. This guide is not universally accepted
and is only one of many tools used to evaluate transformers. The four conditions are
defined below:

Condition 1: Total dissolved combustible gas (TDCG) below this level indicates the
transformer is operating satisfactorily. Any individual combustible gas exceeding
specified levels in Table 3-13 should have additional investigation.
Condition 2: TDCG within this range indicates greater than normal combustible gas
level. Any individual combustible gas exceeding specified levels in Table 3-13 should
have additional investigation. A fault may be present. Take DGA samples at least often
enough to calculate the amount of gas generation per day for each gas (see Table 3-14
for recommended sampling frequency and actions).
Condition 3: TDCG within this range indicates a high level of decomposition of
cellulose insulation and/or oil. Any individual combustible gas exceeding specified levels
in Table 3-13 should have additional investigation. A fault or faults are probably present.
Take DGA samples at least often enough to calculate the amount of gas generation per
day for each gas (see Table 3-14).
Condition 4: TDCG within this range indicates excessive decomposition of cellulose
insulation and/or oil. Continued operation could result in failure of the transformer (see
Table 3-14).

If TDCG and individual gases are increasing significantly (more than 30 ppm/day), the
fault is active and the transformer should be de-energized when Condition 4 levels are
reached. A sudden increase in key gases and the rate of gas production is more
important in evaluating a transformer than the amount of gas. One exception is

107
acetylene (C2H2). The generation of any amount of this gas above a few ppm indicates
high energy arcing. Note however, that trace amounts (a few ppm) can be generated by
a very hot thermal fault (500 °C). One-time arcs caused by a nearby lightning strike or a
high-voltage surge can also generate acetylene. If C2H2 is found in the DGA, oil
samples should be taken weekly to determine if additional acetylene is being generated.
If no additional acetylene is found and the level is below the IEEE Condition 4, the
transformer may continue in service. However, if acetylene continues to increase, the
transformer has an active high-energy internal arc and should be taken out of service.
Further operation is extremely hazardous and may result in catastrophic failure.

Table 3-13 assumes that no previous DGA tests have been made on the transformer or
that no recent history exists. If a previous DGA exists, it should be reviewed to
determine if the situation is stable (gases are not increasing significantly) or unstable
(gases are increasing significantly). Deciding whether gases are increasing significantly
depends on the particular transformer.

Table 3-13: Dissolved Key Gas Concentration Limits in Parts Per Million (ppm)

CO CO2
H2 CH4 C2H2 C2H4 C2H6
Status (Carbon (Carbon TDCG
(Hydrogen (Methane (Acetylene (Ethylene (Ethane
Monoxide) Dioxide)
Condition 1 100 120 1 50 65 350 2,500 720
Condition 2 101-700 121-400 2-9 51-100 66-100 351-570 2,500-4,000 721-1,920
Condition 3 701-1,800 401-1,000 10-35 101-200 101-150 571-1,400 4,001-10,000 1,921-4,630
Condition 4 >1,800 >1,000 >35 >200 >150 >1,400 >10,000 >4,630
*
CO2 is not included in adding the numbers for TDCG because it is not a combustible gas

Compare the current DGA to earlier DGAs. If the production rate (ppm/day) of any one
of the key gases and/or TDCG (ppm) has suddenly gone up, gases are probably
increasing significantly. Refer to Table 3-14, which gives suggested actions based on
total amount of gas in ppm and rate of gas production in ppm/day.

Before going to Table 3-14, determine transformer status from Table 3-13; that is, look
at the DGA and see if the transformer is in Condition 1, 2, 3, or 4. The condition for a
particular transformer is determined by finding the highest level for any individual gas or
by using the TDCG. If the TDCG number shows the transformer in Condition 3 and an
individual gas shows the transformer in Condition 4, the transformer is in Condition 4.
Always be conservative and assume the worst until proven otherwise [62].

108
 Table 3-14: Actions Based on Dissolved Combustible Gas

Conditions TDCG Level or Highest TDCG Sampling Intervals and Operating Actions for Gas Generation Rates
Individual Gas Generation
(See Table 4) Rates Sampling Interval Operating Procedures
(ppm/Day)
Condition 1 720 ppm of TDCG or <10 Annually: Continue normal operation.
highest condition based 6 months for EHV
on individual gas from transformers
Table 3-13 10-30 Quarterly
>30 Monthly Exercise caution.
Analyze individual gases to find cause.
Determine load dependence.
Condition 2 721-1,920 ppm of TDCG <10 Quarterly Exercise caution.
or highest condition Analyze individual gases to find cause.
based on individual gas 10-30 Monthly Determine load dependence.
from Table 3-13
>30 Monthly

Condition 3 1,941-4,630 ppm of <10 Monthly Exercise extreme caution.

TDCG or highest Analyze individual gases to find cause.
condition based on 10-30 Weekly Plan outage.
individual gas from Table Call manufacturer and other consultants for advice.
3-13 >30 Weekly

Condition 4 >4,630 ppm of TDCG or <10 Weekly Exercise extreme caution.

highest condition based Analyze individual gases to find cause.
on individual gas from 10-30 Daily Plan outage. Call manufacturer and other
Table 3-13 consultants for advice.
>30 Daily Consider removal from service.
Call manufacturer and other consultants for advice.
NOTES:
1. Either the Highest Condition Based on Individual Gas or Total Dissolved Combustible Gas can determine the condition (1, 2, 3, or 4) of
the transformer. For example, if the TDCG is between 1,941 ppm and 2,630 ppm, this indicates Condition 3. However if hydrogen is
greater than 1,800 ppm, the transformer is in Condition 4, as shown in Table 3-13.
2. When the table says “determine load dependence,” this means, if possible, find out if the gas generation rate in ppm/day goes up and
down with load. Perhaps the transformer is overloaded. Take oil samples every time the load changes; if load changes are too frequent,
this may not be possible.
NOTES:
3. Either the highest condition based on individual gas or total dissolved combustible gas can determine the condition (1, 2, 3, or 4) of the
transformer. For example, if the TDCG is between 1,941 ppm and 2,630 ppm, this indicates Condition 3. However if hydrogen is greater
than 1,800 ppm, the transformer is in Condition 4, as shown in Table 3-13.
4. When the table says “determine load dependence,” this means, if possible, find out if the gas generation rate in ppm/day goes up and
down with load. Perhaps the transformer is overloaded. Take oil samples every time the load changes; if load changes are too frequent,
this may not be possible.
5. To get TDCG generation rate, divide the change in TDCG by the number of days between samples that the transformer has been
loaded. Down-days should not be included. The individual gas generation rate ppm/day is determined by the same method.

Sampling intervals and recommended actions: When sudden increases occur in

dissolved gases, the procedures recommended in Table 3-14 should be followed. Table
3-14 is paraphrased from Table 3 in IEEE C57.104-1991. The table indicates the
recommended sampling intervals and actions for various levels of TDCG in ppm. An
increasing gas generation rate indicates a problem of increasing severity; therefore, as
the generation rate (ppm/day) increases, a shorter sampling interval is recommended
(see Table 3-14).

Some information has been added to the table from IEEE C57-104-1991 as can be
inferred from the text. If the cause of the gassing can be determined and the risk can be
assessed, the sampling interval may be extended. For example, if the core is tested

109
with a M -meter and an additional core ground is found, even though Table 3-14 may
recommend a monthly sampling interval, an operator may choose to lengthen the
sampling interval since the source of the gassing and generation rate is known.

A decision should never be made on the basis of just one DGA. It is very easy to
contaminate the sample by accidentally exposing it to air. Mishandling may allow some
gases to escape to the atmosphere and other gases, such as oxygen, nitrogen, and
carbon dioxide, to migrate from the atmosphere into the sample. If you notice a
transformer problem from the DGA, the first thing to do is take another sample for
comparison.
3.2.2.12.3 Rogers Ratio Method
In interpreting gas analysis results, relative gas concentrations are found to be more
useful than actual concentrations. For most purposes, only five gas concentrations (H2,
CH4, C2H6, C2H4, and C2H2) are sufficient. According to the scheme developed by R.R.
Rogers [63] and later simplified by the IEC, three gas ratios define a given condition. It
is important to note that in developing the ratio analysis, Rogers considered gas
measurements from mostly conservator type transformers with open expansion tanks
(Type III transformers). Like the key gas analysis discussed above, this method does
not provide guaranteed answers, but is only an additional tool to use in analyzing
transformer problems.

The three-ratio version of the Rogers Ratio Method uses the following ratios:

R1 = C2H2/C2H4
R2 = CH4/H2
R3 = C2H4/C2H6

Note that the Rogers Ratio Method is for analyzing faults and not for detecting the
presence of faults. Its use requires the establishment of a problem based on the total
amount of gas (using IEEE limits) or increased gas generation rates. A good system to
determine whether there is a problem is to use Table 3-13 (latest version) in the Key
Gas Method. If two or more of the key gases are in Condition 2 and the gas generation
is at least 10% per month of the L1 limit (Table 3-17), there is a high likelihood of a
problem. If a gas used in the denominator of any ratio is zero, or is shown in the DGA
as not detected (ND), use the detection limit of that particular gas as the denominator.
This gives a reasonable ratio to use for diagnosis. A further refinement in applying the
ratio methods is to subtract gases that were present prior to any sudden gas increases.
This takes out gases that have been generated up to the point of analysis due to normal
ageing and prior problems. This is especially true for ratios involving gases that are
generated during normal ageing, H2, and the cellulose insulation gases CO and CO2
[64].

In using these ratios, it is advisable to never make a decision based only on a ratio if
either of the two gases used in that ratio is less than 10 times the amount the gas
chromatograph can detect [64]. This rule makes sure that instrument inaccuracies have
little effect on the ratios. If either of the gases is lower than 10 times the detection limit, it

110
is most likely that the transformer does not have the particular problem that this ratio
deals with. When a fault occurs inside a transformer, there will be more than enough
gases present to make the ratios valid. Detection limits for the key gases are shown in
Table 3-15. Table 3-15 also provides possible diagnoses based on the values of the
three ratios.
Table 3-15: Rogers Ratios for Key Gases

Gas Detection Limits 10 x Detection Limits

C2 H2 1 ppm 10 ppm
C2H2/ CH4/ C2H4/ C2 H4 1 ppm 10 ppm
Code Range of Ratios
C2H4 H2 C2H6 CH4 1 ppm 10 ppm
H2 5 ppm 50 ppm
C2 H6 1 ppm 10 ppm
<0.1 0 1 0
0.1-1 1 0 0
1-3 1 2 1
>3 2 2 2
C2H2/ CH4/ C2H4/
Case Fault Type Problems Found
C2H4 H2 C2H6
0 No fault 0 0 0 Normal ageing
Electric discharges in bubbles, caused by insulation voids, super
Low energy partial
1 1 1 0 gas saturation in oil or cavitation (from pumps), or high moisture in
discharge
oil (water vapor bubbles).
Same as above but leading to tracking or perforation of solid
High energy
2 1 1 0 cellulose insulation by sparking or arcing. This generally produces
partial discharge
CO and CO2.
Low energy Continuous sparking in oil between bad connections of different
3 discharges, 1-2 0 1-2 potential or to floating potential (poorly grounded shield etc);
sparking, arcing breakdown of oil dielectric between solid insulation materials.
Discharges (arcing) with power follow through; arcing breakdown of
High energy oil between windings or coils, between coils and ground, or load
4 1 0 2
discharges, arcing tap changer arcing across the contacts during switching with the oil
leaking into the main tank.
Insulated conductor overheating This generally produces CO and
Thermal fault less than
5 0 0 1 CO2, because this type of fault generally involves cellulose
150°C (see note 2)
insulation.
Thermal fault Spot overheating in the core due to flux concentrations. Items
temp. range below are in order of increasing temperatures of hotspots. Small
6 0 2 0
150-300°C hotspots in core. Shorted laminations in core. Overheating of
(see note 3) copper conductor from eddy currents. Bad connection on winding
Thermal fault to incoming lead or bad contacts on load or no-load tap changer.
7 temp. range 0 2 1 Circulating currents in core. This could be an extra core ground,
300-700°C (circulating currents in the tank and core). This could also mean
stray flux in the tank.
Thermal fault
8 temp. range over 0 2 2 These problems may involve cellulose insulation, which will
700°C (see note 4) produce CO and CO2.

Notes:
1. There will be a tendency for ratio C2H2 /C2H4 to rise from 0.1 to above 3 and the ratio C2H4 /C2H6 to rise from 1-3 to above 3
as the spark increases in intensity. The code at the beginning stage will then be 1 0 1.
2. These gases come mainly from the decomposition of the cellulose, which explains the zeros in this code.
3. This fault condition is normally indicated by increasing gas concentrations. CH 4/H2 is normally about 1, the actual value
above or below 1, is dependent on many factors, such as the oil preservation system (conservator, N 2 blanket, etc.), the oil
temperature, and oil quality.
o
4. Increasing values of C2H2 (more than trace amounts), generally indicates a hotspot higher than 700 C. This generally
indicates arcing in the transformer. If acetylene is increasing and especially if the generation rate is increasing, the
transformer should be de-energized as further operation is extremely hazardous.
General Remarks:
1. Values quoted for ratios should be regarded as typical (not absolute). There may be transformers with the same problems
whose ratio numbers fall outside the ratios shown at the top of the table.
2. Combinations of ratios not included in the above codes may occur in the field. If this occurs, the Rogers Ratio Method will
not work for analyzing these cases.
3. Transformers with on-load tap changers may indicate faults of code type 2 0 2 or 1 0 2 depending on the amount of oil
interchange between the tap changer tank and the main tank.

111
If samples from Type I transformers (N2 blanket) are compared to those from Type II
transformers (sealed conservator), it is necessary to make adjustments to gas
concentrations and consequently some gas ratios used for diagnostic purposes.
Fortunately, major adjustment is required only for the hydrogen concentration. Details of
the adjustment procedure were derived by Oommen [65]. The only gas ratio that needs
significant adjustment is the CH4/H2 ratio. The adjustment factor is 0.44 at 25 °C. This
means that a gas ratio obtained from measurement on a Type I transformer should be
multiplied by 0.44 to equate to a measurement on a Type II transformer. Since Rogers
developed his method based on sample from Type III transformers, there is some
uncertainty about strict enforcement of ratio codes to all types of transformers. With this
qualification, it may be pointed out that the ratio codes are of great value in diagnosing
transformer faults.

The severity of faults identified in transformers using the Rogers ratio patterns is shown
in Table 3-16. The level of urgency in correcting a problem will obviously depend on the
severity of the fault. While it may be sufficient to place a transformer with an overheating
conductor problem on a watch list, one with an arcing fault might require immediate
removal from service and subsequent investigation.

Table 3-16: Order of Severity of Transformer Faults

Increasing Order of Severity Fault Patterns Pattern # in Table 3-15

1 Normal 0
2 Conductor Overheating 5
3 Oil Overheating, Mild 6
4 Oil Overheating, Moderate 7
5 Oil Overheating, Severe 8
6 Partial Discharge, Low Energy 1
7 Partial Discharge, High Energy 2
8 Arcing 3,4

3.2.2.12.4 IEC Method

The IEC method (See IEC 60599 latest version is second edition 1999-03) uses five
different types of faults and three basic ratios. The method is very similar to the Rogers
Ratio above. The faults and ratios are as follows:

PD Partial discharges
D1 Discharges of low energy
D2 Discharges of high energy
T1 Thermal fault, T < 300 °C
T2 Thermal fault, 300°C < T < 700 °C
T3 Thermal fault, T > 700 °C
Basic ratios: C2H2/C2H4, CH4/H2 and C2H4/C2H6
3.2.2.12.4.1 Carbon Dioxide/Carbon Monoxide (CO2/CO) Ratio
The formation of CO2 and CO from oil-impregnated paper insulation increases rapidly
with temperature. Incremental (corrected) CO2/CO ratios less than 3 are generally
considered as an indication of probable paper involvement in a fault, with some degree

112
of carbonization. Normal CO2/CO ratios are typically in the range 5 - 9. Ratios above 10
generally indicate a thermal fault with the involvement of cellulose. If CO is increasing
around 70 ppm or more per month (generation limit from IEC 60599), there is probably
a fault. In order to get reliable CO2/CO ratios in the equipment, CO2 and CO values
should be corrected first for possible CO2 absorption from atmospheric air; and CO2 and
CO background generation (see 6.1 and clause 9 of IEC 60599). The background
generation result from the ageing of cellulosic insulation, overheating of wooden blocks
and the long term oxidation of oil. For example, if air-breathing equipment is saturated
with approximately 10% of dissolved air, there could be up to 300 l/l (ppm) of CO2 just
from the air. In sealed equipment, air is normally excluded but may enter through leaks.
The concentration of CO2 will be in proportion to the amount of air present. When
excessive paper degradation is suspected (CO2/CO < 3), it is advisable to ask for a
furanic compounds analysis or a measurement of the degree of polymerization of paper
samples, if this is possible.
3.2.2.12.4.2 IEC C2H2/H2 Ratio
In power transformers equipped with on-load tap changers (OLTC), the tap changer
operations produce gases corresponding to discharges of low energy in the main tank
(D1). If some oil or gas communication is possible between the OLTC compartment and
the main tank, or between the respective conservators, these gases may contaminate
the oil in the main tank and lead to wrong diagnoses. The pattern of gas decomposition
in the OLTC, however, is quite specific and different from that of regular low energy
discharges in the main tank.
3.2.2.12.4.3 IEC Recommended Method of Interpretation
a) Reject or correct inconsistent DGA values. Calculate the rate of gas increase since
the last analysis, taking into account the precision of the DGA results. If all gases are
below typical values of gas concentrations and rates of gas increase, report as
"Normal DGA/healthy equipment". If at least one gas is above typical values of gas
concentrations and rates of gas increase, calculate gas ratios and identify fault.
Check for eventual erroneous diagnosis. If necessary subtract last values from
present ones before calculating ratios, particularly in the case of CO and CO2. If
DGA values are above typical values but below 10 times the analytical detection limit,
see the section in IEC 60599 on “Uncertainty of ratios”.
b) Determine if gas concentrations and rates of gas increase are above alarm values.
Verify if fault is evolving towards final stage. Determine if paper is involved.
c) Take proper action according to the best engineering judgment.

It is recommended to:
1) Increase sampling frequency (quarterly, monthly or other) when the gas
concentrations and their rates of increase exceed typical values,
2) Consider immediate action when gas concentrations and rates of gas increase
exceed alarm values.

113
3.2.2.12.5 Duval Triangle Method for Diagnosing a Transformer Problem Using Dissolved Gas
Analysis [66]
Duval developed this method in the 1960s using a database of thousands of DGAs and
transformer problem diagnoses. This method has proven to be accurate and
dependable over many years and is now gaining in popularity. The method and how it is
used is described below. Before this method is applied, it is best to follow these steps:

1. First determine whether a problem exists by using the IEEE method above,
and/or Table 3-17 below. At least one of the hydrocarbon gases or hydrogen (H2)
must be in IEEE Condition 3, and increasing at a generation rate (G2) from the
table below, before a problem is confirmed. To use Table 3-17 below without the
IEEE method, at least one of the individual gases must be at L1 level or above
and the gas generation rate at least at G2. If there is a sudden increase in H2
with only carbon monoxide (CO) and carbon dioxide (CO2) and little or none of
the hydrocarbon gases, use the (CO2/CO ratio) below to determine if the
cellulose insulation is being degraded by overheating.
2. Once a problem has been determined to exist, use the total accumulated amount
of the three Duval Triangle gases and plot the percentages of the total on the
triangle to arrive at a diagnosis. Also, calculate the amount of the three gases
used in the Duval Triangle, generated since the sudden increase in gas began.
Subtracting out the amount of gas generated prior to the sudden increase will
give the amount of gases generated since the fault began. Detailed instructions
and an example are shown below.

a) Take the amount (ppm) of methane (CH4) in the DGA and subtract the
amount of CH4 from an earlier DGA, before the sudden increase in gas. This
will give the amount of methane generated since the problem started.
b) Repeat this process for the remaining two gases, ethylene (C2H4) and
acetylene (C2H2).
c) Add the three numbers (differences) obtained by the process of step b)
above. This gives 100 % of the three key gases generated since the fault.
d) Divide each individual gas difference by the total difference of gas obtained in
step c) above. This gives the percentage increase of each gas of the total
increase.
e) Plot the percentage of each gas on the Duval Triangle, beginning on the side
indicated for that particular gas. Draw lines across the triangle for each gas
parallel to the hash marks shown on each side of the triangle (see Figure
3-12). The triangle coordinates, corresponding to DGA results in ppm, can be
calculated as follows: %C2H2 = 100x/(x+y+z); %C2H4 =100y/(x+y+z); %CH4 =
100z/(x+y+z); where x = C2H2, y = C2H4, z = CH4.

114
The diagnostic regions in the triangle are defined as:

PD = Partial Discharge
T1 = Thermal Fault less than 300 °C
T2 = Thermal Fault between 300 °C and 700 °C
T3 = Thermal Fault greater than 700 °C
D1 = Low Energy discharge (Sparking)
D2 = High Energy discharge (Arcing)
DT = Mix of Thermal and Electrical Faults

Table 3-18 provides examples of the typical faults in transformers for each of the
diagnostic categories in the Duval analysis triangle. The table is derived from the IEC
draft 60599 (Edition 2) [64].

Figure 3-12: Coordinates and Fault Zones of the Duval Triangle

CAUTION:
Do not use the Duval Triangle to determine whether or not a transformer has a problem.
Notice, there is no area on the triangle for a transformer that does not have a problem.
The triangle will show a fault for every transformer whether it has a fault or not. Use the
key gas or TDCG methods to determine if a problem exists before applying the Duval
Triangle. The Duval Triangle is used only to diagnose what the problem is. As with other
methods, a significant amount of gas must already be present before this method is
valid.

115
 Table 3-17 : L1 Limits and Generation (G1, G2) Rate Per Month Limits

Gas L1 Limits G1 Limits G2 Limits

(ppm per (ppm per
month) month)
H2 100 10 50
CH4 75 8 38
C2 H2 3 3 3
C2 H4 75 8 38
C2 H6 75 8 38
CO 700 70 350
CO2 7,000 700 3,500
NOTE:
In most cases, acetylene (C2H2) will be zero, and the result will be a point on the right
side of the Duval Triangle. Compare the total accumulated gas diagnosis and the
diagnosis obtained by using only the increase-in-gases after a fault. If the fault has
existed for some time, or if generation rates are high, the two diagnoses will be the
same. If the diagnoses are not the same, always use the diagnosis of the increase in
gases generated by the fault, which will be the more severe of the two.

Table 3-18: Example of Faults from the Duval Analysis of Power Transformers

Fault Type Examples

Partial Discharges in gas-filled cavities in insulation, resulting from incomplete impregnation, high moisture in paper,
discharges gas-in-oil super-saturation or cavitation (gas bubbles in oil), leading to X wax formation on paper.
Discharges of Sparking or arcing between bad connections of different floating potential, from shielding rings, toroids, adjacent
low energy discs or conductors of different windings, broken brazing, closed loops in the core. Additional core grounds.
Discharges between clamping parts, bushing and tank, high voltage and ground, within windings. Tracking in
wood blocks, glue of insulating beam, winding spacers. Dielectric breakdown of oil, load tap changer breaking
contact.
Discharges of Flashover, tracking or arcing of high local energy, or with power follow through. Short circuits between low
high energy voltage and ground, connectors, windings, bushings, and tank, windings and core copper bus and tank, in oil
duct. Closed loops between two adjacent conductors around the main magnetic flux, insulated bolts of core,
metal rings holding core legs.
Overheating Overloading the transformer in emergency situations. Blocked or restricted oil flow in windings. Other cooling
less than 300°C problems, pumps valves, etc. Stray flux in damping beams of yoke.
Overheating Defective contacts at bolted connections (especially bus bar), contacts within tap changer, connections between
300 to 700°C cable and draw rod of bushings. Circulating currents between yoke clamps and bolts, clamps and laminations, in
ground wiring, bad welds or clamps in magnetic shields. Abraded insulation between adjacent parallel
conductors in windings.
Overheating Large circulating currents in tank and core. Minor currents in tank walls created by high uncompensated
over 700°C magnetic field. Shorted core laminations.
Notes:
1. X wax formation comes from Paraffinic oils (paraffin based); however, naphthenic oils are not immune to X wax formation
2. The last overheating problem in the table is for faults over 700°C. Recent laboratory discoveries have found that acetylene can
be produced in trace amounts at 500°C, which is not reflected in this table. Transformers that show trace amounts of acetylene
are probably not active arcing but may be the result of high-temperature thermal faults. It may also be the result of one arc, due
to a nearby lightning strike or voltage surge.
3. A bad connection at the bottom of a bushing can be confirmed by comparing infrared scans of the top of the bushing with a
sister bushing. When loaded, heat from a poor connection at the bottom will migrate to the top of the bushing, which will display
a markedly higher temperature. If the top connection is checked and found tight, the problem is probably a bad connection at
the bottom of the bushing.

116
3.2.2.12.6 ABB's Advanced Dissolved Gas Analysis Software (ADGA)

It has been ABB’s experience that the design and application of a transformer can
make it have its own unique gassing pattern. ABB has developed an internal
software package that combines DGA raw data, ratios, trending, key indicators, and
ABB's resident design expertise and transformer construction knowledge to interpret
the results. By combining ABB's design and manufacturing knowledge with the
analysis capabilities of the software, the analysis is able to offer greater analytical
depth than what is standard practice in the industry. The program has the ability to
pinpoint specific sources and causes of gas generation, rather than simply identify
general categories of gas generation.

Figure 3-13 shows the results of an analysis performed with this software. In addition to
the individual gas concentrations, the program requests the rate of generation of each
gas and a series of inputs relating to the type of oil preservation system and application
of LTC, etc. The results are a prioritized list of diagnoses and colour-coded pictorials of
the severity of each gas concentration and diagnostic ratio. The likely sources of the
fault can be obtained by activating an explanation screen.

Figure 3-13: Advanced DGA Analysis of Power Transformer Gas-in-Oil Sample

117
 3.2.3 ANALYSIS OF PARTICLES IN TRANSFORMER OILS [67]
Transformer manufacturers and utilities currently use particle contamination as another
means of monitoring oil quality in transformers. This is due to the increasing awareness
of the factors that influence the dielectric strength of oil. High-level particle
contamination is recognized as an important factor. The breakdown strength of
transformer oil is a function of the concentration, size, shape, and type of the particles
and the moisture level in the oil.

In performing these analyses, identification of the particles is important in determining

the source of particle generation in operating transformers. The chief sources of
particulate matter in transformers are cellulosic dust and fibers and residual dirt. Iron
particles, particles of copper, and other metals could exist from manufacturing
operations. The factory filtering and flushing operations remove most of these particles;
therefore, the particle level would be relatively low. Some undesirable conditions in
service, such as pump problems and electrical discharges, tend to generate particles;
therefore, the periodic monitoring of particle level should be considered part of the
preventive maintenance program. Many field problems are detected by electrical tests
and gas analysis, but a few, such as pump bearing, wear may not be apparent. Pump
bearing wear is of particular interest, because metallic particles generated could reduce
the dielectric strength of the transformer insulation.
3.2.3.1 OIL SAMPLING FOR PARTICLE ANALYSIS
In taking samples for particle count analysis, the oil should be taken from the bottom
valve of transformers via flexible tubing. At least a gallon of oil should be allowed to flow
out to purge the lines before the sample bottle is introduced into the flowing stream. The
bottle itself should be clean; ultrasonic cleaning is preferable in most cases. Large
bottles should be rinsed in the oil stream even if they are pre-cleaned. The bottles
should be capped soon after sampling. In spite of these precautions, particle
contamination from outside sources may not be completely eliminated. When high
counts are measured, a second sample should be taken to check for sampling errors.
3.2.3.2 PARTICLE COUNTING [68]
Particle size analyzers are used to perform particle counting in transformer oils. These
instruments use the principle of light blockage to estimate the size of each particle as it
passes through a micro cell in which a transverse light beam is applied. The cross-
sectional area of the particle is automatically estimated, and this area is converted to an
equivalent circle or ellipse. The measured particle size may be expressed as the
diameter of this circle or as the major diameter of the ellipse. Since most particles are
non-spherical, especially dust and wear particles, the ellipse approximation is preferred.
Until 2000, the optical particle counters used for transformer fluid analysis were
calibrated using a standard suspension of what is known as ACFTD (Air Cleaner Fine
Test Dust) particles in a hydrocarbon fluid (MIL 5606) of comparable viscosity. The
standard for calibrating particles based on the ACFTD method is ISO 4406-1999 [69].
The ACFTD calibration method has since been replaced with ISO 11171 [70] and ASTM
method D6786, which specifies a solution of Medium Test Dust (MTD). The conversion
of a sample of particle sizes from ACFTD to MTD methods is given in Table 3-19.

118
 Table 3-19: Particle Size Conversion

ACFTD Size ( m) MTD Size ( m)

1.0 4.2
3.0 5.1
5.0 6.4
10.0 9.8
15.0 13.6
20.0 17.5
50.0 38.2
100.0 70.0

Counting is done in the cumulative mode, i.e., for any specified size, the number of
particles above that size would be measured. The ACFTD method suggested reporting
cumulative particle counts 1, 5, 10, 15, 25, 50, and 100 m sizes. These
correspond roughly to the recommended sizes of 4, 6, 10, 14, 21, 38, and 70
m sizes for the newer MTD method.

The level of contamination in a unit is determined via a contamination code that

depends on the number of cumulative particles in a defined range per milliliter of oil.
The contamination code is determined with reference to the scheme given in ISO-4406-
1999. A segment of the scheme is shown in Table 3-20. To determine the
contamination code, particle levels at three sizes are used, 4 m, 6 m, and 14 m.
These roughly correspond to the ACFTD sizes of 1 m, 5 m, and 15 m.

Table 3-20: Particle Contamination Code

Number of Particles per Milliliter CODE Number
5000 to 10,000 20
2500 to 5000 19
1300 to 2500 18
640 to 1300 17
320 to 640 16
160 to 320 15
80 to 160 14
40 to 80 13
20 to 40 12
10 to 20 11
5 to 10 10

3.2.3.2.1 Normal and Abnormal Particle Count Levels

From experiments performed by ABB, it appears that units with greater than 150
particles above 5 m per milliliter of oil using the MTD method (or 150 particles above 3
m per milliliter of oil using the ACFTD) deserve further analysis and possibly inspection
if other tests prove positive. These values are not intended to be an upper limit on the
permissible particles in operating transformers. However, particle size analysis should
be supplemented by quantitative trace metal analysis as described below.

119
3.2.3.3 T RACE METAL CONTENT OF PARTICLES
The sources of particles with metallic content have already been mentioned. In this
section the technique used to measure trace metallic levels and the results obtained will
be described.
3.2.3.3.1 Method of Measurement
Several methods exist for trace metal analysis of oil samples. The most commonly used
methods at present are ICP (inductively coupled plasma) atomic absorption and atomic
emission and x-ray fluorescence. The atomic absorption technique is especially suited
for very low-level contamination levels (in the parts per billion ranges). Unlike emission
spectroscopy and x-ray fluorescence, it does not directly identify all the elements
present in a sample; the presence of each element has to be tested using standards in
atomic absorption spectroscopy, and a selection of metals is detected by ICP. Atomic
absorption is therefore a time consuming technique if several elements are to be tested.

In the AAS technique the sample is “atomized” in a flame or furnace at temperatures in

the 1,500-3,000 °C range. For each element to be tested, a separate hollow cathode
lamp of that element is mounted and energized to produce emission lines of
characteristic wavelengths. The emission beam is allowed to pass through the vapor. If
the vapor has atoms of the same element, these atoms would absorb energy from the
beam in proportional to the concentration of atoms.

The exact methodology has to be worked out for each type of analysis. If particles are
extremely fine, e.g., below 10 microns, the sample would be more or less
homogeneous, and the furnace technique could be used. Only micro liters of the oil
sample are required, and no sample preparation is needed. However, the reproducibility
of the furnace technique is not high when small samples are used, and many particles
suspended in oil are greater than 10 microns. This type of analysis would give metallic
content of both suspended and dissolved material. This procedure of only analyzing
suspended particles has been found to be reproducible and correlates well with units
having known sources of contamination.

Meaningful metal analysis can be confined to four elements: iron, copper, lead, and
zinc. All these elements could be analyzed easily by the flame technique using
air/acetylene flame. The selection of iron and copper needs no explanation. Lead and
zinc are elements normally found in the pump bearing alloy material. It must be pointed
out that lead and zinc could be present in oil from other sources such as solders, zinc
plated parts, and paints; also, the wear of the alloy may not produce particles of the
same composition. Lead and zinc are relatively low melting, and may be partly lost
during wear process and sample preparation. Iron oxide is a component of dirt, dust,
and impure clay; therefore, a bad sample could show excessive iron content.
3.2.3.3.2 Normal and Abnormal Metallic Content of Particles in Oil
The metallic content of particles is expressed, for convenience, in parts per billion (ppb)
units, which could be better stated as micrograms/ml. For very clean oil, the levels of
iron, copper, lead, and zinc approach detection limits, 1-2 ppb. An upper limit of 5 ppb is

120
typically observed for each element in clean oil with total particle count not exceeding
500.

Based on a limited study of 200 samples taken by ABB from both factory and field units,
the following levels appear normal for both factory and field units:

Iron: 10 ppb, max.

Copper: 20 ppb, max.

If levels greater than these are measured, further study may be required. Most units
with reported bearing problems show higher than average upper levels. Considerable
caution should be used in the application of these limits. First, the analysis technique
between laboratories should be standardized. Secondly, the total volume of oil in the
transformer should be taken into account. The oil volume in large power transformers
could vary from 10,000 to 30,000 gallons. If particles originate from general degradation
processes, the particle concentration would be uniform regardless of the size of the
transformer. If, however, particles originate from a localized mechanical problem, the
total oil volume would influence particle concentration. This is especially true of oil
samples tested from residual oil after flushing operations. Both the particle level and
metallic level could be higher than normal. However, such concentrated samples may
still be of value for metal identification. The levels suggested above do not correct for
transformer size.
3.2.3.4 DIAGNOSTIC EXAMPLES OF PARTICLE ANALYSIS

Table 3-21 shows some examples of problem diagnoses using particle analysis. They
show that excessive high particle levels may indicate wear and degradation. Also,
excessive copper content may be associated with pump bearing wear problems in some
cases. These examples and others reported in the open literature demonstrate that
particle level analysis coupled with AAS is a useful technique to monitor metallic
contamination in transformer oil.
Table 3-21: Diagnostic Examples of Particle Analysis
Case Description Total Visible METALLIC CONTENT (ppb) Comments
Particles Particles Iron Copper Lead Zinc
Pump failure from impeller 58,225 31 8.8 107.7 15.5 6.9 The excessive copper content
and thrust bearing wear. confirms the problem. Shiny metallic
Sample taken from bottom particles were visible. Pump bearing
of unit after pump failure. wear may not always produce such
high levels, but AAS and particle
counting could still be used to test
whether the problem exists.
Pump problem from radial 750 6 17.6 75 2.7 3.8 Visual inspection showed that the
bearing wear. Sample was rear radial bearing had frozen on the
taken from a unit with axle; the pump was, however, still
suspected pump problems. operating.
Pump motor winding short. 619 3 3.5 116.8 12.1 17.2 Although particle count is deceptively
This analysis was low, the metallic analysis showed
performed after a pump excessive copper content. The
winding failure in a factory shorting caused gas generation from
situation. oil decomposition in the pump
housing.

121
3.2.3.5 EFFECT OF PARTICLES ON DIELECTRIC STRENGTH OF INSULATING OIL13 [71]
The effect of particles on the dielectric strength of transformers has been characterized
to a large extent. Experimental investigations have been numerous and most of them
show a sizeable reduction of dielectric strength, especially if a large oil volume is used
and the voltage is applied over a long time period. Since investigations were mostly
carried out on bare electrodes, they are relevant only for the case of discharges initiated
in the oil. For discharges initiated at the electrode-to-oil interface, the effect of particle-
initiated discharges on the insulation is obviously considerable but has not yet been
characterized.

The measurement of the particle content in an oil sample has shown large
discrepancies when results from different laboratories are compared. Measurements on
different samples, carried out in a single laboratory, appear to be much more consistent
and successive measurements on the same sample have shown very good
repeatability. Particle counting is somewhat hampered by the very small volume of the
oil sample compared to the total oil volume of oil in the transformer. Depending on the
sampling valves and techniques used, it is possible to measure completely different
particle concentrations in the same transformer.

In spite of these difficulties, it is necessary to establish recommendations since the

detrimental effect of particles has been identified conclusively in a number of failures
either in the field or during factory acceptance tests. The experience of utilities and
manufacturers reveals that this type of failure is observed almost exclusively on EHV
transformers. This is believed to be linked to the smaller ratio of test voltage to service
voltage and the large oil volume found in EHV transformers. The most vulnerable part of
the transformer is the high-voltage bushing shield and high-voltage lead, especially if
the insulation is provided mainly by a large oil volume without subdividing barriers. This
effect appears to be enhanced when these components are located in a turret.
3.2.3.5.1 Current filtering practices on new transformers
It is quite normal for a newly manufactured transformer to have a significant content of
particles, mainly cellulose. It is now common to apply a filtering procedure to the oil,
before proceeding with the acceptance tests. This precaution applies mainly to EHV
transformers. In a study by a CIGRE working group, most of the manufacturers that
were consulted do not actually count particles but establish a factory procedure that
simply calls for a certain number of passes of the oil through a filter. After installation at
the site, a similar procedure of oil filtering is recommended for these more sensitive
transformers. This precaution allows some contamination to be eliminated that could
originate from the coolers, the erection procedure or the oil itself. Only a few
manufacturers have set limits on the particle count acceptable before commissioning. In
those few cases, the applicable limit is 1,000 particles larger than 5 m per 100 ml oil
volume. A more realistic measurement of particles is somewhat easier at this stage
since the filtering creates a large oil circulation outside the transformer tank and allows
for the use of on-line particle counting with a continuous supply of homogenized oil.

13
CIGRE SC A2 (ex 12) WG 17, - Particles in Oil, Nov. 1999

122
3.2.3.5.2 Classification of contamination level

3.2.3.5.2.1 Bare electrodes

Most of the reported experiments are made with bare electrodes, using test cells as
specified by IEC 60156 or ASTM D-1816. However, some of the tests have been
carried out with plane electrodes or bushing shields of very large dimensions.
The presence of particles, whether conducting material or cellulose fibers, always
reduces the average breakdown voltage. The reduction factor varies widely and it
cannot be readily related to the oil volume under stress, voltage application method or
type of contaminant. The particle counting method is another possible cause of the
dispersion. It must be noted that these reductions in dielectric strength are applicable
only to the oil and cannot be applied to a system where the electrodes are covered with
solid insulation.
It is the CIGRE Working Group’s opinion that, because of its small volume, the IEC
electrode is not the best configuration for evaluating the effect of particles. In normal
transformer oil, the amount of particles per unit of volume is rather small and the
standard procedure for oil testing does not allow sufficient time for the particles to move
to the right place. Coaxial test cells have been suggested by France and tried by others.
This gives a continuous oil flow and therefore a larger oil volume is actually tested. The
coaxial test cell appears to be the best tool presently available to quantify the effect of
particles on the dielectric strength of insulating oil.
The effect of moisture is significant, especially in the presence of cellulose fibers. This
can be best illustrated with a set of results from Sinz [72]. The effect of moisture is
obvious, as is the better sensitivity of the coaxial test cell as compared to the IEC test
cell. It is therefore recommended that the water content be reported along with the
particle content especially when dealing with cellulose particles. The voltage application
method is also questioned. It is argued by some researchers that the step voltage
procedure allows the test cell to be under voltage for a longer time and increases the
probability of a particle moving closer to the area of maximum field stress.
It is also noted that from a practical point of view, the average breakdown voltage is not
as interesting as the minimum breakdown voltage. It is suspected that the particles
might increase the dispersion of the breakdown voltage and it is recommended that the
dispersion be reported along with the average values. It appears that the IEC test
method is not appropriate for showing the detrimental effect of the particles. There is a
need to develop a standardized method that would call for a large oil volume and a long
duration of voltage application.
3.2.3.5.2.2 Covered electrodes
Very little experimental data is available on the effect of particles with electrodes coated
with insulating material. For natural (factory) contamination, a reduction of 29% on the
average breakdown voltage was found.
Hydro-Québec, in collaboration with EHV Weidmann, has run some tests with plane
electrodes and a combination of pressboard sheets and pressboard spacers to
simulating the main insulation between high-voltage and low-voltage windings in a large

123
transformer. The introduction of aluminum powder in the insulating oil only slightly
reduced the average breakdown voltage (7%), but the reduction on the minimum value
was more significant (32%). It is possible that the breakdown mechanism involved here
is quite different from the one in section 3.2.3.5.2.1.
3.2.3.5.3 Contamination deposited on insulating surface
Some researchers have endeavored to quantify the impact of conducting particles when
deposited on insulating structures. Hydro-Québec, in collaboration with EHV Weidmann,
has investigated the effect of aluminum deposited on spacers in the main insulation of a
transformer. In this case, the aluminum was diluted in a solvent which was then applied
with a brush on the side of the pressboard spacer. Two pressboard pieces spaced 12-
mm apart were used to simulate a 12-mm oil duct; three pressboard pieces equally
spaced at 6-mm were used to simulate two, 6-mm oil ducts. The reduction in average
breakdown value was 24 % for the single duct and 14% for the double oil duct.
A similar test, without the pressboard barrier, was carried out on a larger scale by ABB
at the request of Hydro-Québec. Here again the aluminum powder was applied with a
brush. With this contamination, the average breakdown voltage was reduced from
400 kV to 280 kV, a reduction of 30 %.

Table 3-22: Typical contamination levels encountered on power transformer insulating oil

(ISO class) Maximum count per Contamination Typical occurrence

100 ml designation
5 m 15 m
Up to 8/5 250 32 None IEC cleanliness requirement for sample
bottle filled with a clean solvent
9/6 to 10/7 1000 130 Low Excellent oil cleanliness encountered
during factory acceptance test and
transformer commissioning
11/8 to 15/12 32000 4000 Normal Contamination level typical for
transformers in service
16/13 to 17/14 130000 16000 Marginal Contamination level found on a
significant number of transformers in
service
18/15 and above High Contamination level rare and usually
indicative of abnormal operating
conditions

Vincent [73] reported an experiment with a piece of pressboard contaminated with

carbon particles and subjected to divergent fields in a rod-plane configuration. In this
case, carbon particles were collected on the surface of the pressboard by subjecting it
to an AC electric field in an oil container that is heavily polluted with carbon. The electric
field which allowed the particles to be collected was perpendicular to the surface while
the test field was tangential. During the test with the 1-min step voltage application, it
was observed that the electric field near the tip of the rod had a cleansing effect on the

124
pressboard, progressively removing most of the deposited carbon. The breakdown
voltage of the contaminated sample was therefore not significantly lower than the clean
one.
3.2.3.5.4 Recommended corrective action
Corrective action for the reduction of particle content should be initiated only after
proper evaluation of the dielectric strength of the oil. For screening purposes, the IEC
60156 test procedure is adequate but if there is a discrepancy between the
contamination level and the dielectric strength of the oil, the dielectric test should be
repeated with a procedure capable of showing the detrimental effect of particles, if any.
For the purpose of these recommendations, a “marginal” dielectric performance is
defined as a reduction of 30% or more of the new oil characteristics. The recommended
action for EHV transformers in service is summarized in Table 3-23.
Table 3-23: Recommended action for contaminated oil

Contaminatio Dielectric Recommended action

n level strength
Good No further action.
Normal
Poor Identify type of particles.
Good Probably dirt or dry cellulose. Repeat
dielectric strength with a test procedure
appropriate for particles.
Marginal
Marginal Identify type of particles. Check moisture
content. Filtering may be considered.
Good Recheck particle count. Recheck
dielectric strength with a test procedure
appropriate for particles. Investigate
source of particles.
High
Marginal Filtration or replacement is strongly
recommended.

125
 3.2.4 WINDING RESISTANCE TEST
This test is a measure of the resistance of the conductors in the transformer winding.
The resistance measurement is corrected to either 75 °C or 85 °C, depending on the
average winding temperature rise of the transformer. The correction temperature is the
average winding rise plus 20 °C. If the temperature rise for the transformer is 55 °C, the
winding resistance is corrected to 75 °C, and if it is 65 °C, the resistance is corrected to
85 °C. The winding resistance will typically change if there are shorted turns, loose
connections on bushings, loose connections or high-contact resistance in tap changers
and broken winding strands. These conditions will typically lead to hotspots in the
winding or the affected areas and generate hot metal gases in the oil. The gases to look
for in a DGA in case of poor connections are ethylene, ethane, and to some extent,
methane. If the DGA suggests the possibility of any of the situations mentioned above,
a winding resistance test is in order.

Figure 3-14: Low Resistance Ohmmeter - Biddle Model 247001 (Courtesy of Megger)14

3.2.4.1 MEASUREMENT METHOD FOR WINDING RESISTANCE MEASUREMENT

Winding resistance measurements are performed with a low-resistance ohmmeter such
as shown in Figure 3-14. For a three phase wye-connected transformer, the resistance
is measured for each phase-to-neutral winding; if delta-connected, the resistance is
measured for each phase-to-phase winding. Note that for delta-connected transformers,
the measured resistance for each phase is composed of a parallel combination of the
winding under test and the series combination of the remaining windings. It is therefore
recommended to make three measurements for each phase-to-phase winding in order
to obtain the most accurate results. It is also recommended to allow the transformer to
sit de-energized until temperatures are equalized (difference between top and bottom
temperatures does not exceed 5 °C – ANSI/IEEE C57.12.90) before making resistance
measurements.

According to IEC 60076-1, in order to reduce measurement errors due to changes in

temperature, some precautions should be taken before the measurement is made.

For Dry type transformers, the transformer shall be at rest in a constant ambient
temperature for at least 3 hours.
For Oil immersed transformers, the transformer should be under oil and without
excitation for at least 3 hours. In addition, it is important to ensure that the average oil

14
From website: http://www.megger.com/us/.

126
temperature (average of the top and bottom oil temperatures) is approximately the
same as the winding temperature.

To avoid an inadmissible winding temperature rise during the measurement, it is also

recommended that the measuring current should be limited to no more than 10 percent
of the rated current of the winding.

In order to diagnose possible problems, the measured results are compared to the
factory values, values of other phases of the same transformer, or sister units, if
available. Before making such comparisons, the resistance has to be converted to a
common temperature base of 75 °C or 85 °C, depending on what is reported on the
transformer factory test sheet.

The corrected resistance is calculated as:

RM (CF CT )
RCT
CF Winding Temp( o C )

where:
RCT = Corrected resistance
CF = 234.5 for copper windings; 225 for aluminum windings (IEEE C57.12.90)
CF = 235 for copper windings; 225 for aluminum windings (IEC 60076-1)
CT = 75 for 55 C rise transformers; 85 for 65 C rise transformers
RM = Measured winding resistance

Consistency in measurements and record keeping are the keys to making the proper
analysis using this test. If the unit has a tap changer, it is important to compare
resistances for the same tap position. The contact resistance of other tap positions can
be investigated by moving taps and repeating the measurements.

A measurement is deemed acceptable and no further investigation is needed if the

individual phase readings are within 2% of the other phase readings for three phase
transformers or within 2% of the reported factory value for single phase transformers.
Changes greater than 2% may be due to loose connections, broken conductor strands,
short circuits, or bad tap changer contacts, or they can be caused by uncertainty in the
temperature correction. For very low resistance values, it is not uncommon for
measurements to be outside of the 2% limit even in a perfectly normal transformer. In
such cases the measurement tolerances of test equipment may not be sufficient to
resolve the acceptable 2% limit between measurements.

When readings are outside the 2% range, it is recommended to investigate further or to

consult the transformer manufacturer to determine acceptability of the results.

127
 3.2.5 TRANSFORMER TURNS RATIO TEST (TTR)
The function of a transformer is to transform power from one voltage level to another.
The ratio test ensures that the transformer windings have the proper turns to produce
the voltages required. The “turns ratio” is a measure of the RMS voltage applied to the
primary terminals to the RMS voltage measured at the secondary terminals:

Np Ep
r
Ns Es

Where:
r = voltage ratio
E = open-circuit voltage
N = number of turns
p = primary
s = secondary

The IEEE standard (IEEE Standard 62) states that when rated voltage is applied to one
winding of the transformer, all other rated voltages at no load shall be correct within one
half of one percent of the nameplate readings. It also states that all tap voltages shall be
correct to the nearest turn if the volts per turn exceed one half of one percent of the
desired voltage. The ratio test verifies that these conditions are met.

The IEC 60076-1 standard defines the permissible deviation of the actual to declared
ratio as follows:

Principal tapping for a specified first winding pair: the lesser of ± 0.5% of the declared
voltage ratio or 0.1 times the actual short-circuit impedance. Other taps on the first
winding pair and other winding pairs must be agreed upon, and must not be lower than
the smaller of the two values stated above.

Deviations in turns ratio readings indicate problems in one or both windings. In

particular, the TTR test is useful for identifying shorted turns or open circuits in the
windings, incorrect winding connections, and other internal faults or tap changer
defects. If possible, the ratio at each tap setting should be checked against the
nameplate ratio for each tap.

Measurements are typically made by applying a known low voltage across the high-
voltage winding (as a primary) so that the induced voltage on the secondary is lower,
thereby reducing hazards while performing the test. For a three phase delta/wye or
wye/delta transformer, a three phase equivalency test is performed, i.e. the test is
performed across corresponding single windings. The appropriate test configurations for
various connections for three phase two-winding transformers are shown in Table 3-24.
One of a variety of test sets used for performing these measurements is shown in
Figure 3-15.

128
 Figure 3-15: Three Phase TTR Test Set (Courtesy of Megger)15

The TTR test value should not be greater than 0.5 % or less than 0.5 % of the
calculated values. For a three phase three-winding transformer, the following
measurements will be made in a TTR assessment.

Table 3-24: TTR Measurement Configurations

Connection Apply Voltage Measure Voltage Calculate Voltage

Across Winding Across Winding Ratios
Delta-Delta H1-H2 X1-X2 VH1-H2/VX1-X2
H1-H3 X1-X3 VH1-H3/VX1-X3
H2-H3 X2-X3 VH2-H3/VX2-X3
Delta-Wye H1-H2 X0-X3 VH1-H2/VX0-X3
H1-H3 X0-X2 VH1-H3/VX0-X2
H2-H3 X0-X1 VH2-H3/VX0-X1
Wye-Wye H0-H1 X0-X1 VH0-H1/VX0-X1
H0-H2 X0-X2 VH0-H2/VX0-X2
H0-H3 X0-X3 VH0-H3/VX0-X3
Wye-Delta H0-H1 X1-X2 VH0-H1/VX1-X2
H0-H2 X1-X3 VH0-H2/VX1-X3
H0-H3 X2-X3 VH0-H3/VX2-X3

For a three-winding transformer, the ratios can be from the primary to both the
secondary and the tertiary windings and can be used in further diagnosing which
winding may have a problem. For example, in a wye/wye/wye configuration, Table 3-25
can be used to diagnose possible problems in the 0-1 phase of the transformer.

Table 3-25: Using TTR to Diagnose Winding Problems

Measure
Voltage
X0-X1 Y0-Y1 Possible Diagnosis
Apply
Voltage
H0-H1 Ratio Abnormal Ratio OK Problem in X0-X1 Winding
Ratio Abnormal Ratio Problem in H0-H1 Winding
H0-H2
Abnormal
Ratio OK Ratio Problem in Y0-Y1 Winding
H0-H3
Abnormal

15 ®
From Megger Website:
http://www.megger.com/us/products/ProductDetails.php?ID=233&Description=ttr.

129
Note that the TTR test can only indicate if one of the problems listed above is present in
the transformer. It cannot pinpoint the exact location of the fault. This must be
investigated via an internal inspection, which may involve un-tanking the transformer.

130
 3.2.6 INSULATION RESISTANCE
The insulation resistance test, also called Megger test, is used to determine the leakage
current resistance of the insulation. The resistance is a function of the moisture and
impurity content of the insulation as well as the insulation temperature. At a constant
voltage, the resistance also depends on the strength of the electric field across the
insulation and therefore is a function of the size and construction of the transformer.
Primarily, this measurement gives information about the condition of the insulation and
ensures that the leakage current is adequately small.
3.2.6.1 MEASUREMENT
Insulation resistance of a transformer is measured by means of a resistance meter
using a DC voltage. In measuring resistance, it is recommended to always be sure that
the tank and core are grounded. Each winding of the transformer is then short circuited
at the terminals. Resistance measurements are made between each winding and all
other windings grounded. Windings are never left floating during insulation resistance
measurements. When any winding is installed with a solid ground, the ground must be
removed in order to measure the insulation resistance of that winding to the other
windings grounded. If the ground cannot be removed, the insulation resistance of that
winding cannot be measured. It is treated as part of the grounded section of the circuit.
Insulation resistance is expressed in mega ohms (M ).

On a two winding transformer the following measurement configurations are used:

1. Measure from the high voltage winding to the low voltage winding and
ground [H-LG]
2. Measure from the low voltage winding to the high voltage winding and
ground [L-HG]
3. Measure from the high and low voltage winding to ground [HL-G]

This test is easily performed in the field. Many manufacturers require that this test be
performed prior to energizing a transformer, to preclude start up failure caused by entry
of moisture into the transformer during shipment or storage. The test can also detect
other ground circuits that may exist in the transformer that may have been caused by
shipping damage. The test checks the complete circuit – bushings, leads and coils.

The measurement duration is 1 minute. The resistance readings R15 and R60 are taken
15 and 60 seconds after connecting the voltage. In order to compare these readings
with future measurements, it is important to record on the test report, the temperature,
measuring voltage, the meter used, as well as the measured resistances. Since
insulation resistances may vary with applied voltage, any comparisons must be made
with measurements at the same voltage.

WARNINGS
The following precautions should be taken in performing the insulation
resistance test:

131
 The test should be discontinued immediately if the current begins to increase
without stabilizing
Under no conditions should the test be made while the transformer is under
vacuum
After the test has been completed all terminals should be grounded for a period
of time sufficient to allow any trapped charges to decay to a negligible value
3.2.6.2 INTERPRETATION
The IEC Standard 60076-1 and the IEEE Standard C57.12.90 provide no limits for
insulation resistances. However, the ratio R60:R15, also called the absorption ratio, is
normally in the range 1.3 – 3 in a dried transformer. The condition of the insulation can
also be determined by comparing the measured resistance at 1 minute, R60, to a
minimum value for the voltage class of the winding. This comparison is performed only
after all measurements are converted to their 20 °C equivalents using the coefficients in
Table 3-26. For example, if the measured value is 20 M at 12 °C, according the table
this measurement is equivalent to 11,8 (=20 x 0,59) M at 20 °C.

The minimum measured resistance corrected to 20 °C is given by the relationship16 :

CE
R60
kVA

Where:
kVA is the rated capacity of the winding under test,
C is a constant:
o 0.8 for oil-filled transformers at 20 °C, or
o 16.0 for dry, compound filled or untanked oil filled transformers
E is the voltage rating of the winding under test
R60 is the 1 minute reading of insulation resistance of winding to ground with
other windings grounded or between windings in M at 20°C

Table 3-26: Insulation Resistance Correction Factors

For Conversion of Test Temperature to 20 °C [74]

Temp (oC) Coefficient Temp (oC) Coefficient Temp (oC) Coefficient

0 0.25 24 1.33 41 4.20
5 0.36 25 1.40 42 4.50
10 0.50 26 1.50 43 4.80
11 0.54 27 1.60 44 5.10
12 0.59 28 1.74 45 5.60
13 0.62 29 1.85 46 5.95
14 0.66 30 1.98 47 6.20
15 0.70 31 2.10 48 6.80
16 0.76 32 2.30 49 7.20
17 0.82 33 2.45 50 7.85

16
M. Horning et. al., Transformer Maintenance Guide, pp. 108-109, 2001

132
 Temp (oC) Coefficient Temp (oC) Coefficient Temp (oC) Coefficient
18 0.86 34 2.60 55 11.20
19 0.96 35 2.80 60 15.85
20 1.00 36 3.00 65 22.40
21 1.08 37 3.20 70 31.75
22 1.15 38 3.40 75 44.70
23 1.25 39 3.70 80 63.50
24 1.33 40 3.95

3.2.6.3 POLARIZATION INDEX

Polarization index is the relationship between the measured resistance after 10 minutes
and that measured after 1 minute. Since the conduction processes are enhanced for an
insulation system that is contaminated with moisture or impurities, the leakage current
will increase at a greater rate than for a dry, clean insulation. Consequently, under the
same test configuration, the insulation resistance of a wet or contaminated insulation
system will settle faster and at a lower value than that for a dry insulation. The result is
that the polarization index for a wet insulation will be lower than that for a dry insulation
system. Since the polarization index is a ratio, it does not require conversion to a
common temperature base before comparisons can be made. It also does not require
for there to be previous measurements before an assessment of the insulation condition
can be made. The following guidelines are used to assess the condition of insulation
based on the polarization index.

Table 3-27: Polarization Index Interpretation Guide

Polarization Insulation
Index Condition
<1 Unsatisfactory
>2 Good

133
 3.2.7 INSULATION POWER FACTOR TESTS
Insulation power factor tests are performed on transformer insulation to determine the
condition of the capacitive insulation between the windings, between windings and core,
and between windings and the tank or other grounded components in the transformer.
There are three test modes essential to the evaluation of an insulation system:
Ungrounded Specimen Test (UST), Grounded Specimen Test (GST), and Grounded
Specimen Test with Guard (GST/g). These configurations allow various sections of the
insulation system to be tested separately.

Power factor test instruments typically have three leads: an output high-voltage lead for
energizing the test object, input measurement, and ground leads that measure current
through the insulation system. Internally, switches allow either input lead to be
connected to a current/wattmeter input or guard, depending on the testing configuration.

In the UST configuration, current flowing in the insulation between the high-voltage lead
and the measurement lead is measured by connecting the measurement lead to the
current/wattmeter input. The ground lead is connected to the guard, and therefore
currents that flow through the ground lead are not measured by the meter. In the GST
configuration, all currents flowing from the HV lead to ground are measured by the
meter. This is accomplished by internally connecting both the measurement and the
ground leads to the input of the current/wattmeter. In the GST/g configuration, the
measurement lead is connected to the guard, and the ground lead is connected to the
input to the current/wattmeter device. The only measured current is what is in the direct
path from the HV lead to ground.

The UST values can also be calculated from the difference between the measured GST
and the GST/g values. The reason for making all these measurements is to allow for the
evaluation of the various sections of insulation in the transformer. However, the most
important of these measurements is the UST test, since it measures across the major
insulation of the transformer. The power factor is calculated from the measured current
and watts loss recorded by the meter according to the following equation:

PF(%) = 10 x Loss(Watts)/Current(mA)

A system that is widely used by utilities in measuring power factor of insulation systems
is Doble Engineering’s M4000 Automated Insulation Analyzer shown in Figure 3-16.

Figure 3-16: Doble M4000 Automated Insulation Analyzer 17

17
From the Doble Engineering Website: www.doble.com

134
3.2.7.1 T WO-WINDING T RANSFORMER
In order to perform power factor tests on a three phase, two-winding transformer, it is
necessary to connect all high-voltage bushings together and all low-voltage and neutral
bushings together.

Figure 3-17: Schematic of Two-Winding Transformer Insulation Capacitance for Power Factor
Measurements 18

The capacitance between these two terminals and between each terminal and the
ground terminal, represented by the tank and core, are shown schematically in Figure
3-17. In Figure 3-17, the capacitances are defined as follows:

CH represents the insulation between the high-voltage winding conductor and the
grounded tank and core. The capacitance takes into account the HV bushings,
structural insulating members, the de-energized tap changer insulation, and the
insulating fluid.
CL represents the insulation between the low-voltage winding conductors and the
grounded tank and core. The capacitance takes into account the LV bushings,
the winding insulation, the structural insulating members, the LTC insulation, and
the insulating fluid.
CHL represents the insulation between the high- and low-voltage windings and
includes the winding insulation barriers and the insulating fluid.

18
The IEC equivalent nomenclature for the winding terminals is as follows: H1=1U; H2=1V, H3=1W;
X1=2U; X2=2V; X3=2W, X0=2N

135
3.2.7.1.1 Testing of Two-Winding Transformers
For a two-winding transformer, there are six different tests that are performed to assess
the insulation condition in the various parts of the transformer insulation. For each test,
high voltage is applied to one set of windings, and current from the other winding and
the ground terminal are fed into the measurement equipment. Table 3-28 shows which
measurement lead is applied to the transformer windings for each test configuration. It
also shows which measurement leads, if any, are guarded and ultimately the insulation
capacitance that is measured. Figure 3-18 - Figure 3-23 show the actual test setup for
the tests described in Table 3-28.

Table 3-28: Power Factor Measurement Setup for Two-Winding Transformers

Test Mode HV Winding LV Winding Tank/Core Measured

Capacitance
GST HV Lead Meas. Lead Gnd. Lead CH+CHL
GST/g HV Lead Meas. Lead Gnd. Lead CH
(on guard)
UST HV Lead Meas. Lead Gnd. Lead CHL
(on guard)
GST Meas. Lead HV Lead Gnd. Lead CL+CHL
GST/g Meas. Lead HV Lead Gnd. Lead CL
(on guard)
UST Meas. Lead HV Lead Gnd. Lead CHL
(on guard)

Figure 3-18: Power Factor Measurement of CHL + CH Insulation (GST)

136
Figure 3-19: Power Factor Measurement of CH Insulation (GST/g)

Figure 3-20: Power Factor Measurement of CHL Insulation (UST)

137
 Figure 3-21 : Power Factor Measurement of CHL + CL Insulation (GST)

Figure 3-22: Power Factor Measurement of CL Insulation (GST/g)

138
 Figure 3-23: Power Factor Measurement of CHL Insulation (UST)

3.2.7.2 T HREE-WINDING T RANSFORMER

The various insulation capacitances of a three-winding transformer are shown in Figure
3-24. The power factor of each of these insulation sections can be examined by the
measurement configurations defined in Table 3-29.

Figure 3-24: Schematic of Three-Winding Transformer Insulation Capacitance for Power Factor
Measurements

139
 Table 3-29: Power Factor Measurement Configuration for Three-Winding Transformers

Test Mode HV Winding LV Winding TV Winding Tank/Core Measured

Capacitance
GST/g HV Lead Meas. Lead Meas. Lead Gnd. Lead CH
(on guard) (on guard)
GST/g Meas. Lead HV Lead Meas. Lead Gnd. Lead CL
(on guard) (on guard)
GST/g Meas. Lead Meas. Lead HV Lead Gnd. Lead CT
(on guard) (on guard)
UST HV Lead Meas. Lead Gnd. Lead Gnd. Lead CHL
(on guard) (on guard)
UST HV Lead Gnd. Lead Meas. Lead Gnd. Lead CHT
(on guard) (on guard)
UST Gnd. Lead HV Lead Meas. Lead Gnd. Lead CLT
(on guard) (on guard)

3.2.7.3 TYPICAL I NSULATION POWER FACTOR VALUES

In a study conducted by Doble Engineering Company, the power factor for the high-
voltage winding to ground insulation for 760 transformers shows the distribution shown
in Figure 3-25. The corrected power factor for up to 95 percent of the transformers was
below 0.7 %.

Figure 3-25: High Voltage to Ground Insulation Power Factor for Representative Good Insulation
Systems

140
3.2.7.4 GENERAL GUIDELINES FOR ASSESSING POWER F ACTOR VALUES
In making an assessment of a power factor reading, it is advisable to compare the test
results to previous measurements. The rate of increase in power factor would show a
condition that has stabilized or is rapidly deteriorating. The following are general
guidelines provided by Doble Engineering in assessing power factor results for oil-filled
power transformers:

Table 3-30: Power Factor Diagnosis for Oil-Filled Power Transformers

Power Factor Reading Possible Insulation Condition

0.5% Good
>0.5% BUT 0.7% Deteriorated
>0.7% BUT 1.0% (& Increasing) Investigate
>1.0% Bad

For oil-filled distribution transformers, the power factor numbers in the table are
doubled. For power factor values that are classified as bad or investigate, other test
methods are necessary to positively identify the cause of the high power factor. Such
tests include dissolved gas-in-oil analysis, moisture-in-oil analysis, dielectric frequency
response analysis (DFR), frequency response analysis (FRA/SFRA), and power factor
tip-up test. Most of these tests are discussed in more detail in later sections. A
discussion of the power factor tip-up test follows.
3.2.7.5 POWER FACTOR T IP-UP T ESTS
The power factor tip-up test is performed by applying voltage in equal steps from zero to
the maximum allowed voltage. The test is performed on the section of insulation with
highest power factor reading. For each applied voltage, the current and watts loss
through the insulation is measured, and the power factor is calculated. If moisture or
other polar contaminants are the cause of the high power factor, the measured power
factor will be essentially the same for all applied voltages. If the power factor increases
with voltage, there is likely ionic contamination and/or carbonization of the oil or
windings for oil-filled transformers. For dry type transformers, the problem may be due
to ionic contaminants or the presence of voids in the winding insulation.

141
 3.2.8 CORE INSULATION RESISTANCE MEASUREMENT
Generally, the core laminations in a core form type transformer are insulated from
ground, and the core is deliberately grounded at a single point. Measurement of the
core insulation resistance allows for investigating accidental grounds which result in
circulating currents if there is more than one connection between the core and ground.
The dielectric withstand of the core-to-ground insulation is typically specified to be
above 2 kV AC. The intentional core ground connection is usually mounted under a
manhole at the top of the transformer or through the tank wall via a small low-voltage
bushing. Either design allows the ground to be easily disconnected and allows a
measurement of the resistance between core and ground. However, there are shell
form designs in which the core ground is inaccessible. In such cases this measurement
cannot be made.

Several factors can lead to an inadvertent ground connection to the core: the core-
ground insulation can deteriorate to a point where the insulation becomes resistive; the
core-ground insulation can become damaged during transportation of a transformer; or
the core-ground insulation can become damaged in a through-fault incident. If an
unintentional core ground is established as a result of any of the above conditions, there
will likely be circulating currents in the core. The result will be hotspots in the core and
surrounding metal structures. The presence of these hotspots can be detected using a
DGA screening. Key gases to look for are ethane, ethylene, and/or possibly methane.
Depending on the location of the hotspots, cellulose may be involved, and the gases
may include CO and CO2.
3.2.8.1 MEASUREMENT AND DIAGNOSIS OF INADVERTENT CORE GROUNDS
The gas signature attributable to hotspots due to inadvertent core grounds can also be
present if there is a poor connection at the bottom of a bushing or a bad tap changer
contact. Therefore, this test is only necessary if a winding resistance test shows that all
connections are good and if the tap changer contacts are assessed to be in good
condition.

The test is performed using a standard DC Megger® such as the one shown in Figure
3-26. The two test leads of the Megger test set are connected between the isolated
core-ground lead and the transformer ground. A DC voltage of no more than 1000 volts
is applied across the leads, and the resistance is measured. Depending on the resulting
resistance, Table 3-31 can be used to guide what action must be taken.

142
 Figure 3-26: DC Megger Test Set (Courtesy of Megger)19

Table 3-31: Diagnosing Inadvertent Core-Ground Problems

Measured Core Ground

Possible Interpretation Action
Resistance
New transformer. Good
1000 M NONE
coreground insulation.
Service aged transformer.
100 M Acceptable core ground NONE
insulation.
Deteriorating core ground Investigate cause of deterioration
10 to < 100 M
insulation. and mitigate.
Deteriorated insulation is
Investigate and correct before
1 to < 10 M possible cause of circulating
re-energization.
currents.
Check to make sure a limiting
resistor is not being used in the
Possible high-resistance ground
200-1000 Ohms core-ground circuit. If not, there is
between core and ground.
a possible high-resistance ground
that must be corrected.
Solid connection between core Investigate and correct before
< 10 Ohms
and ground. re-energization.

If the core-ground insulation is less than 10 M , the first step in investigating the
inadvertent ground connection is to switch to an ohmmeter and measure the resistance
between the core and ground. This should help establish whether there is a solid
ground connection or a high-resistance ground present. In either case, there are field
techniques available in eliminating the unintentional grounds (see IEEE Standard 62).

19
From AVO website: http://www.avomegger.com/.

143
 3.2.9 EXCITATION CURRENT TESTS

The excitation current test is one of the means of identifying problems associated with
the core or winding of the transformer. The test can possibly detect core problems
such as shorted core laminations and poor joints. Winding problems detected include
short circuited or open circuited turns, poor electrical connections, tap changer
problems, and other possible core and winding problems. The exciting current
consists of a magnetizing component and a loss component. The magnitude of the
magnetizing component is determined by the shape of the performance curve of the
core steel, its operating flux density, and the number of turns in the primary winding.
The loss component is determined by the losses in the core. Joint construction
severely affects the magnitude of the excitation current. Changes in the hysteresis
and eddy current characteristics due to handling the steel also affect the excitation
current.

To perform the test, voltage is applied to the primary windings one at a time with all
other windings left open. The excitation current of a transformer is the current which
the transformer draws when voltage is applied to its primary terminals with the
secondary terminal open. It is important to perform the excitation current tests before
any direct current (DC) tests. DC tests leave a residual magnetism in the core that
would distort an excitation current test. Before performing an excitation current test,
the following steps are necessary [75]:

Disconnect all loads and de-energize the transformer.

It is recommended that the test voltage be applied to the HV windings. Exercise

caution in the vicinity of all transformer terminals because voltage will be
induced in all windings during a test.

Winding terminals normally grounded in-service should be grounded during

tests, except for the particular winding energized for the test.

For routine tests, the load tap changer (LTC) should be set to neutral, then to
one step above neutral, then to one step below neutral, and then to full raise or
full lower. To ensure that the tap selector is functioning properly throughout the
entire range of selection, you may want to perform tests on all LTC positions.

Test voltages should not exceed the rated line-to-line voltage for delta-
connected windings or rated line-to-neutral voltage for wye-connected
windings. These tests are generally made at 2.5, 5, or 10 kV, as the capacity of
the test equipment permits.

Test voltages should be the same for each phase. Because of the nonlinear
behavior of exciting current, test voltages should be set accurately if results are
to be compared. If excitation tests have previously been performed, the same
test voltage should be used for the current test.

144
Excitation current tests performed on all tap positions of a transformer with a
reactance-type load tap changer can have the following patterns. The currents
measured on the even steps and neutral positions are similar to each other but
different from those measured on the odd steps. The currents measured on the odd
steps are similar to each other. The difference is attributed to how the reactor-
switching device is connected to the tap winding when the tap is on an even or odd
position. For the even numbered and neutral positions, the two contacts of the
reactor-switching device are on the same stationary contact. For odd numbered
positions, the switching contacts bridge adjacent stationary contacts [76].
3.2.9.1 MEASUREMENT SETUP

The excitation current test can be performed using any high-voltage source and a
precision amplifier. However, since both are present in a power factor test set, these
test sets are normally used to perform the excitation current test. The testing mode
for all measurements is set to UST (Ungrounded Specimen Test). See Figure 3-27,
Figure 3-28, and Figure 3-29 for the setup of the excitation current measurements for
various transformer configurations. Table 3-32 is a summary of the test connections
and the means for analyzing test results.

For single phase transformers, the test is performed with high-voltage windings
energized alternately from opposite ends and reading the excitation current in both
configurations. The two currents obtained should be the same. Currents recorded for
single phase transformers should be compared either with similar units or with data
obtained from previous tests on the same unit. If single phase excitation current tests
were included in the factory test specifications, comparing test data reveals changes
undergone between the factory and the field.

For three phase wye-connected transformers, the three measurements routinely

made are H1-H0, H2-H0, and H3-H0. The usual pattern of the exciting current values
is such that two of the measured currents are high and similar, and the remaining one
is lower. The lower value is usually associated with the winding wound on the middle
leg because the reluctance of the magnetic circuit associated with this winding is
lower than the other two windings. This should also be done on the individual phases
of three phase transformers if the unit is suspect, or if the initial exciting current
measurements are questionable.

For three phase delta-connected transformers, the three measurements routinely

made are H1-H2, H2-H3, and H3-H1. The usual pattern for these transformers is two
measured currents that are approximately equal and higher than the third measured
current. Again, the lower current value is ordinarily associated with the winding
wound on the middle leg [77]. With delta-connected transformers, the two higher-
valued currents are occasionally not equal. This can be attributed to the shunting
affect of the un-energized winding being parallel with the energized winding. Test
procedures are available to eliminate the shunting effect of the un-energized winding
[76].

145
 Table 3-32: Excitation Current Test Connection Using Power Factor Test Set20

Transformer Measured
Energized Measurement Floating Normal Current
Type and 21 Ground Lead Excitation
Lead Lead Terminals Pattern
Connection Current
H1 H2 X1 X2 IH1-H2
Single Phase IH1-H2 ~ I H2-H1
H2 H1 X1 X2 IH2-H1

Three Phase
H1 H0 H 2 H 3 ,X1 X2 X3 IH1-H0
Core Form
H2 H0 H 1 H 3 ,X1 X2 X3 IH2-H0 (IH1-H0 ~ IH3-H0 ) > I H2-H0
W ye-Connected
H3 H0 H 1 H 2 ,X1 X2 X3 IH3-H0
3-limb core

Three Phase
H1 H0 H 2 H 3 ,X1 X2 X3 IH1-H0
Shell Form
H2 H0 H 1 H 3 ,X1 X2 X3 IH2-H0 (IH1-H0 ~ IH3-H0 ) > I H2-H0
W ye-Connected
H3 H0 H 1 H 2 ,X1 X2 X3 IH3-H0
D core
Three Phase
IH1-H0 ~ IH2-H0 ~ I H3-H0
Core Form H1 H0 H 2 H 3 ,X1 X2 X3 IH1-H0
(The middle phase
W ye- H2 H0 H 1 H 3 ,X1 X2 X3 IH2-H0
may be slightly
Connected H3 H0 H 1 H 2 ,X1 X2 X3 IH3-H0
higher)
5-limb core
Three Phase
H1 H0 H 2 H 3 ,X1 X2 X3 IH1-H0
Shell Form
H2 H0 H 1 H 3 ,X1 X2 X3 IH2-H0 (IH1-H0 ~ IH3-H0 ) > I H2-H0
W ye-Connected
H3 H0 H 1 H 2 ,X1 X2 X3 IH3-H0
7-limb core

Three Phase H1 H2 H3 X1 X2 X3 IH1-H2

Delta- H2 H3 H1 X1 X2 X3 IH2-H3 (IH2-H3 ~ I H3-H1 ) < I H1-H2
Connected H3 H1 H2 X1 X2 X3 IH3-H1

Table 3-32 lists the forms of transformer construction, the associated magnetic core
configuration, and the usual pattern of core excitation current measurements. In old
designs with non-step lap cores, the quality of the joint gaps has a large effect on the
magnitude of the exciting current such that end phases can have significantly
different measured values of exciting current. The magnitude of the difference can
well be in the same range or even higher than the difference between the measured
exciting current of the middle and end phases. Therefore, the rules on the relative
magnitudes of the exciting current may not apply to these cores. In such cases, only
much greater differences need to be considered as an indication of a problem.

20
The IEC equivalent nomenclature for the winding terminals is as follows: H1=1U; H2=1V, H3=1W;
H0=1N; X1=2U; X2=2V; X3=2W, X0=2N
21
All measurements are performed with the test set in UST mode.
If the secondary winding is wye connected, the neutral (X0) should be connected to ground.

146
 Figure 3-27: Excitation Current Test Method for Single Phase Transformers

Figure 3-28: Excitation Current Test Method for Three Phase Wye-Connected Transformers

Figure 3-29: Excitation Current Test Method for Three Phase Delta-Connected Transformers

147
3.2.9.2 ANALYSIS OF EXCITATION CURRENT RESULTS

If the excitation current is less than 50 mA, the difference between the two higher
currents for a three phase transformer should be less than 10 %. If the excitation
current is greater than 50 mA, the difference should be less than 5 %. In general, if
there is an internal problem, these differences will be greater. When this happens,
other tests should also show abnormalities and an internal inspection should be
considered. If factory tests or prior tests exist, the results should be compared with
them to assess any deviations. High precision does not appear to be necessary in
excitation current tests. The serious faults that have been found have increased
excitation current magnitudes by greater than 10% over normal values [75].

148
 3.2.10 INFRARED THERMOGRAPHY ANALYSIS OF TRANSFORMERS AND ACCESSORIES
Thermography is a method of inspecting electrical and mechanical equipment by
obtaining heat distribution pictures. This inspection method is based on the fact that
most components in a system show an increase in temperature when malfunctioning
[78]. Any localized problems caused by a change in local resistance will consume more
power and generate heat. The local temperature of the resulting hotspot will be higher
than the surrounding temperatures or that of a reference point. By observing the heat
patterns in operational system components, infrared thermography is now used to
detect loose connections, unbalanced load and overload conditions, component
deterioration, and other potential problems [79].
3.2.10.1 T HE T HERMOGRAPHY PROCESS
The inspection tool used by thermographers is the thermal imager (infrared camera).
These are sophisticated devices that measure the natural emissions of infrared
radiation from a heated object and produce a thermal picture. Modern thermal imagers
are portable with easily operated controls (see Figure 3-30 for an example IR camera).
As physical contact with the system is not required, inspections can be made under full
operational conditions, resulting in no downtime.

Figure 3-30: Infrared Camera - FLIR Model ThermaCAM® P65 22

When an object is heated, it radiates electromagnetic energy. The amount of energy is

related to the object’s temperature. The thermal imager can determine the temperature
of the object without physical contact by measuring the emitted energy. The energy
from a heated object is radiated at different levels across the electromagnetic spectrum.
In most industrial applications, it is the energy radiated at infrared wavelengths which is
used to determine the object’s temperature. The thermal imager focuses the emitted
energy via an optical system onto a detector. The detector converts infrared energy into
an electrical voltage which is used to build the thermal picture in the operator’s
viewfinder on board the thermal imager after amplification and complex signal
processing.

22
FLIR website: http://www.flirthermography.com/cameras/camera/1016/.

149
3.2.10.2 CRITERIA FOR EVALUATING INFRARED MEASUREMENTS
When carrying out thermographic inspections, faults are often identified by comparing
heat patterns in similar components operating under similar loads. There is typically
software available with the infrared camera to analyze the temperature signature of the
object under test. A reference point is established on the object for normal temperature.
The temperature rise of all other points on the object is then evaluated in relation to the
reference point temperature. If there are hotspots on the object, the criticality of the
hotspots is evaluated in regards to the magnitude of deviation from the reference
temperature (temperature rise above reference). There are several guidelines for
diagnosing the criticality based on the temperature rises. For example, in performing
temperature-rise tests on transformers, it is recommended that the surface temperature
of the tank, as measured by an infrared camera, be no more than 20 °C higher than the
top oil temperature of the transformer [80].

Criteria established by NASA in evaluating electrical components at its facilities are

given in Table 3-33.

Table 3-33: Infrared Temperature Criteria 23

Temperature Above
Criticality Condition
Reference, Industry
Nominal possibility of
Nominal 0 to 10 oC permanent damage. Repair
next maintenance period.
Possibility of permanent
Intermediate 10 to 20 oC
damage. Repair soon.
Probability of permanent
damage to item and
Serious 20 to 40 oC
surrounding area. Repair
immediately.
Critical over 40 oC Failure imminent.

3.2.10.3 EXAMPLE USES OF INFRARED T HERMOGRAPHY DIAGNOSTICS ON POWER T RANSFORMERS [81]

This section provides a few examples24 of the use of infrared thermography to diagnose
problems in transformers and accessories.
3.2.10.3.1 Loose connection at bushing outlet terminal
When there is a loose connection at the terminal from the bushing to the bus work, it will
lead to overheating of the bushing top terminal when under load. The thermograph will
show the bushing terminal as hot, while the body of the porcelain will show normal
temperatures. Figure 3-31 shows a thermograph of a hot bushing terminal.

23
NASA RCM Specs.
24
Examples are used courtesy of FLIR Systems: www.flirthermography.com.

150
 Figure 3-31: Bushing Terminal Overheating Thermograph

3.2.10.3.2 Blocked oil flow in radiators or radiator shut off

In case of a malfunction that stops or restricts the flow of oil through a radiator, this will
show up on an infrared scan. The image will reveal dim areas where the oil flow is
restricted and brighter areas where normal oil flow is taking place.

Figure 3-32: Thermography of a Shut-Off Radiator Bank

3.2.10.3.3 LTC overheating

Under normal operating conditions and because of I2R and eddy current heating, the
main tank of a transformer will have a higher temperature than the LTC tank in which
there is essentially no heat generation under non-switching conditions. If hotspots
develop in the LTC compartment, this will increase the overall temperature of the LTC
compartment, which may become hotter than the main transformer tank. Such a
situation will be evident on an infrared scan, as shown in Figure 3-33.

151
 Figure 3-33: LTC Compartment Overheating Due to Possible Hotspots in LTC

3.2.10.3.4 Low oil level in transformer or bushing

If a transformer (or especially a bushing) has a low oil level, a thermograph will show a
dim image for the region without oil and a much brighter image in the areas with oil. An
example of this defect is shown in Figure 3-34.

Figure 3-34: Low Oil Level in Transformer

3.2.10.3.5 Moisture contamination of surge arrester

When the internal components of an arrester become contaminated with moisture due
to poor sealing or defects in the porcelain, the resistance of the internal components will
increase. Depending on the extent of the contamination, sections of the surge arrester
body will show localized overheating as compared to other arresters on the transformer.
In this case, the moist regions will show up as dim regions in the thermograph image
[82].

152
 3.2.11 BUSHINGS

3.2.11.1 ANSI & IEC – COMMON DIAGNOSTIC T OOLS

3.2.11.1.1 Oil leakage inspection

A visual inspection for leakage may be performed during normal station supervision.
3.2.11.1.2 Insulator inspection and cleaning
Under conditions of extreme pollution it may be necessary to clean the insulator
surface. The bushing MUST be offline before and during any cleaning operations.
3.2.11.1.2.1 Porcelain insulators

Clean the porcelain insulator with water-jet or wiping with a moist cloth. If necessary,
ethyl-alcohol or ethyl-acetatte may be used.
3.2.11.1.2.2 Silicon rubber insulators

Clean the porcelain insulator with water-jet or wiping with a moist cloth. If necessay,
ethyl-alcohol or ethyl-acetatte may be used. 1,1,1, -Thrichlorethane or Methylchloride
are not recommended due to their possibly harmful and environmentally detrimental
properties.
3.2.11.1.3 Thermovision

Hot spots on the bushing surface can be detected by using an Infrared (IR)-sensitive
camera (see Figure 3-35). At maximum rated current, the bushing outer terminal should
show a temperature of about 35-45 °C above the ambient air. Significantly higher
temperatures, especially at lower current loading, can be an indication of bad
connections.

Figure 3-35 : Measurement indicating poor current path between bushing inner and outer terminal

153
3.2.11.1.4 Oil sampling from bushing

Oil samples shall preferably be taken during dry weather conditions. If, due to some
urgent reason, the sample is taken under any other conditions, the following must be
observed:

- Clean the area around the sampling plug carefully.

- Protect the area around the sampling plug from rain.

The internal pressure of the bushing must not be altered before and after the sampling
as the bushing is supposed to work within a specified range. This requirement is
satisfied if the sample is taken when the mean temperature of the bushing is between
0°C and 30°C.

The time when the bushing is open shall be as short as possible. Flushing with dry air or
nitrogen is normally not necessary.

The oil removed from the bushing shall always be replaced by the same volume of new
transformer oil. The new oil shall comply with IEC 296, class II and shall be clean and
dry.

The gasket shall always be replaced when the bushing is re-sealed.

Sampling procedure for GOB, GOE and GOH

The sample is taken from a plug in the top of the bushing, preferably with a syringe with
a rubber hose connected.

The location for the sampling plug is shown in Figure 3-36. The dimension of the gasket
is given in Table 3-34. The material of the gasket shall be Nitrile rubber with a hardness
of 70 Shore.

154
 Figure 3-36 : Location of oil sampling plugs on some of the most common bushing types.

The tightening torque for the M8 sealing plug on GOB, GOE and GOH shall be 20 Nm.
The tightening torque for the M16 sealing plug on GOE shall be 50 Nm.

Table 3-34: Dimensions for gaskets.

Gasket d (mm) D (mm) T (mm)

M8 8 16 3
M16 14 35 4
5/8" 14 35 4

Sampling procedure for GOEK, GOM and other bushings with sampling valve on the
flange

Connect the end of the hose to a suitable nipple and connect the nipple to the valve on
the flange. The thread in the valve is (R 1/4") BSPT 1/4". Suck out the oil. Depending on
the temperature the pressure inside the bushing might be above or below atmospheric
pressure. After the sampling is finished the bushing shall not be energized within 12
hours.
Sampling procedure for GOA, GOC and GOG

On the GOA, GOC and GOG bushings, the oil samples are taken from the hole for the
oil level plug on the top housing as shown in Figure 3-36. If the bushing is vertically

155
mounted, the oil level is right at the plug level at 20°C. The sample is sucked out by a
syringe. If the oil temperature is slightly higher than 20°C the oil level will be above the
plug level. In such a case the hose on the syringe should be equipped with a nipple as
shown in Figure 3-37. The oil plug is removed and the hose with the nipple is attached
immediately.

If the temperature is below 20 °C, the oil level will be below the plug and the sample is
sucked out according to Figure 3-38. The tightening torque for the 5/8" sealing plug shall
be 50 Nm.

Figure 3-37 : Sampling on GOA at T>20 °C

Figure 3-38 : Sampling on GOA at T<20 °C

3.2.11.1.5 Dissolved Gas Analysis (DGA)

This method for diagnostics can only be used on oil filled bushings, for example, GOx
types. Normally, it is not recommended to take oil samples from bushings. The bushing
is sealed and tightness tested at the time of manufacturing. In order to take an oil
sample, the bushing has to be opened and this introduces a risk of improper re-sealing
after the sampling is finished.

However, when a problem is known, for example high power factor over C1, there might
be a need for oil sampling and gas analysis. The interpretation of the analysis is done
according to Technical Report IEC 61464. If questions remain, ABB can assist with the
evaluation.
3.2.11.1.6 Moisture analysis

There is a risk of improperly sealing a bushing if it is opened to take an oil sample.

However, when a problem is known, for example high power factor over C1, there might
be a need for oil sampling and moisture analysis.

156
It is sometimes difficult to get the proper moisture content in bushing oil. Compared to a
transformer, a bushing has a much higher ratio of paper to oil. This means that
regardless of the bushing manufacturing process, there will always be much more
moisture in the paper than in the oil. In paper the moisture content is measured in
percent, whereas in oil the moisture content is measured in parts per million (ppm).

Depending on the temperature of the bushing, the moisture will move from the paper to
the oil or from the oil to the paper. Due to this, a bushing will always show much higher
moisture content in the oil after a certain time of high temperature operation. To get a
proper value, the oil sample should be taken at least 48 hours after the entire bushing
has reached room temperature.

The bushing is delivered from ABB with maximum moisture content in the insulating oil
of 3 ppm. If considerably higher concentrations are measured, the sealing system is
likely damaged on the bushing.

If the moisture content is greater than 10 ppm, a tan measurement of the bushing C1
capacitance should be performed. If the moisture content is greater than 20 ppm, the
bushing should be taken out of service.
3.2.11.1.7 Dielectric Frequency Response Analysis (DFRA)
This method which is discussed elsewhere in greater detail in this handbook involves
measuring the capacitance and dielectric losses over a frequency spectrum rather than
at a fixed frequency. The status of the insulating material can be obtained from
analyzing the measured loss and capacitance spectra. This method may in the future
become the preferred method and an alternative to DGA for diagnosing bushing
problems. The main advantages are that the bushing does not need to be opened and
proper analysis can be performed regardless of the temperature of the bushing during
the measurement. The shape and frequency shift of the spectra are the main elements
used for diagnosis.
3.2.11.1.8 Partial Discharge measurements
Partial discharge measurement is primarily used as routine testing method by the
manufacturer. Partial discharge may indicate external corona or internal insulation
breakdown. If used for diagnostic on installed transformers it will show the sum of the
partial discharges in the bushing and transformer insulation. External discharges in
switchyards may be suppressed by use of external connected measuring coils. By use
of newly developed acoustic sensors, partial discharges may be located. This method
requires skilled personnel, who have knowledge of bushing and transformer design to
do the measurement.
3.2.11.1.9 De-polymerization analysis
De-polymerization analysis is a method of determining ageing of cellulose in OIP
bushings. The method requires that the bushing is taken apart and a paper sample is
taken from the condenser body.

157
3.2.11.2 DIAGNOSTICS TECHNIQUES FOR BUSHINGS COMPLYING WITH THE ANSI/IEEE STANDARDS

3.2.11.2.1 Condenser Bushing Power Factor Tests

Table 3-35 shows a listing of the possible power factor tests for bushing insulation. The
test connections for these tests are shown in Figure 3-39 – Figure 3-40.

Table 3-35: Power Factor Tests for Bushings

Test Center Potential/ Flange Measured Insulation Involved

Mode Conductor Power Factor Capacitance
Tap
UST HV Lead Meas. Lead Gnd. Lead C1 Main core insulation
GSTg Meas. HV Lead Gnd. Lead C2 Tap insulation core insulation
Lead between tapped layer and
(on guard) bushing ground sleeve, portion of
liquid or compound filler, portion
of watershed near flange

Figure 3-39: Bushing C1 Power Factor Measurement Setup (UST)

158
 Figure 3-40: Bushing C2 Power Factor Measurement Setup (GST/g)

In performing power factor tests on bushings, the following practice is recommended:

Short circuit the windings under test

Clean bushings to minimize the effects of surface leakage currents
Ground opposite windings
Remove test tap cover from bushing under test
Perform C1 test in UST mode
If necessary, perform overall test in GST mode
Perform C2 test in GST/g mode
Replace test cap cover

3.2.11.2.2 Factors Affecting C1 and C2 Capacitance and Power Factor Measurements

As mentioned above, the C1 and C2 capacitance of condenser bushings rated 115 kV
and above are strictly controlled by design and are mainly dependent upon the condition
of the oil-impregnated paper insulation. The power factor and capacitance test values
under normal circumstances are not affected much by external factors. However, under
conditions of contamination and high humidity, these measurements may be
significantly affected. In addition, capacitively coupled resistive paths to ground may
affect these measurements. These may include supporting structures, wooden crates
that are moist/wet, resistance between bushing mounting flange and the transformer
tank, stray effect from other objects, and external connections during testing. Although,
the IEEE Standard C57.19.01 specifies a limit 0.5 % for C1 power factor for oil
impregnated paper insulated bushings, ABB Type O Plus C, AB, and T condenser
bushings have C1 power factor values that are well below this limit.

159
Condenser bushings rated 69 kV and below as mentioned earlier, have the main C1
capacitance, which is strictly controlled by design. The capacitance and power factor
values behave the same behavior and characteristics as those for the 115kV and above
bushings. However, these bushings have an inherent C2 capacitance, which is
dependent upon a few outer layers of paper with adhesive, an oil gap between
the flange and the condenser core, and the tap insulator. Variations in adhesive
in the outer paper layers and other factors can result in power factor variations
in bushings of the same style number. In addition, the close proximity of the C1 layer
with the mounting flange results in greater fringing effect between the two parts. As a
result of this, the porcelains, oil, and air surrounding the bushing can affect the C2
power factor test values. In particular, high current Type T condenser bushings with a
short mounting flange and a long internal C1 layer/foil tend to exhibit higher power
factors because of greater coupling effect between the C1 layer/foil and the surrounding
materials. Depending upon the design, the C2 power factor of these bushings can range
from 0.1 % to 2 %. It is important to note that the IEEE Standard does not specify any
limit for C2 power factor.

For bushings rated 69 kV and below, the IEEE Standard only requires stamping
of C1 power factor and capacitance on the nameplate. As a result of frequent
requests from customers, ABB Inc. Alamo, TN started stamping the C2 power factor and
capacitance test values on bushing nameplates since December of 2002. With this
addition, the nameplates of all AB, O Plus C, and T condenser bushings are now
stamped with factory test values of C1 and C2 power factor and capacitance. However,
because of the reasons mentioned above, users may see a greater variation in C2
power factor and capacitance values in different bushings of the same design.

It is important to compare the initial test values before installation with the
nameplate values. To verify nameplate values (especially for Type T bushings), the
measurements should be made with the bushing mounted on a metallic test tank/stand
with the lower end porcelain immersed in dry good quality oil. There should be sufficient
clearance (at least 16 - 20 inch) from the bushing lower porcelain/terminal to the
grounded tank. For C2 measurement, the center conductor should be guarded and the
test tap voltage should not exceeding 1 kV.

Once the bushing has been installed in the apparatus, it should be retested to establish
a benchmark value. It is important to compare the subsequent field test values with the
initial benchmark value after installation.

Table 3-36 provides typical and questionable power factor values for bushings from
several manufacturers and of various types.

160
 Table 3-36: Typical Bushing Power Factors 25

Manufacturer Type Description Typical PF (%) Questionable PF (%) Comment

Type S, no form letter (through
General Electric A Through Porcelain 3 5
porcelain) redesigned as Type A
General Electric A High Current 1 2
Type S Form F, DF & EF were
Flexible cable,
General Electric B 5 12 redesigned as Type B, BD, and BE
compound-filled
respectively
Oil-filled upper portion,
General Electric D 1.0 2.0
sealed
General Electric F Oil-filled, sealed 0.7 1.5
Oil-filled upper portion,
General Electric L 1.5 3.0
sealed
General Electric LC Oil-filled upper portion 0.8 2.0
Oil-filled expansion
General Electric OF 0.8 2.0
chamber
Force C & CG, Rigid
General Electric S 1.5 6
Core Compound-filled
See special instructions for Type U
General Electric U
in section that follows.
Epoxy Resin Core, plastic
LAPP ERC 0.8 1.5
or oil-filled
Typical C2 power factors for older
Paper Resin Condenser PRC design range from 4-15% due
LAPP PRC, PRC-A 0.8 1.5
Core to injected compound during
manufacturing process
Class LK-
Ohio Brass 0.4 1.0
Type A
ODOF, Class Change of 22% from Manufactured prior to 1926 and
Ohio Brass 1.0-5.0
G, Class L Nameplate value after 1938
ODOF, Class Change of 16% from Manufactured between 1926 and
Ohio Brass 2.0-4.0
G, Class L Nameplate value 1938
Ohio Brass S, OS, FS 0.8 2.0
Westinghouse RJ Solid Porcelain 1.0 2.0
Westinghouse D Semi Condenser 1.5 3.0
Bushings on OCB and instrument
transformers 92 kV to 139 kV
Westinghouse 1.5 3.0
(except Type O, O-A1, OC, and
O+C)
Bushings on power and distribution
Westinghouse 1.0 2.0 transformers of all ratings (except
Type O, O-A1, OC, and O+C)
Modern
Condenser 0.25-0.5 0.5-1.0 (e.g. ABB Type A, O+C)
Bushings

25
Doble Testing Power Apparatus Bushings, 2004 International Conference of Doble Clients

161
3.2.11.2.3 Bushing Hot Collar Test
In cases where a bushing does not have a bushing tap, the C1 and C2 power factor
measurements described above cannot be performed. In such cases, a hot collar test is
performed. This test applies to compound-filled bushings, solid porcelain bushings, gas-
filled bushings, and oil-filled bushings that are not equipped with taps and for which the
bushing overall test cannot be performed. The hot collar test is also useful for various
other bushing checks:

To check bushing oil level on oil-filled bushings without either sight glasses or
liquid level gauges
For bushings with suspect or defective oil level gauges, to check bushing oil
level
As a supplement test when overall or tap tests indicate possible problem.

The test is performed by applying single or multiple collars to various sections of the
bushing. Figure 3-41 shows the setup for a single-collar test in UST mode. This
configuration measures a portion of the insulating watershed, sight glass, core
insulation in upper area, and liquid or compound filler in the upper area of the bushing.
Figure 3-42 shows a similar setup but in GST mode. In addition to the items measured
in the UST mode, this configuration also measures the surface leakage from the collar
to the LV lead and from the collar to the bushing flange. Because the test measures
smaller sections of material, very small dielectric losses and currents are recorded.
Consequently, small changes in either value have tremendous impact on the value of
the calculated power factor. It is therefore advisable to use the value of the measured
dielectric loss as the determining factor in assessing the results of the hot collar test.
The recommended acceptable limits for hot collar tests are 0.1 W at 10 kV and 0.006
W at 2.5 kV. Also, the dielectric loss for the same section in the same type of bushing
should be approximately equal. As a cautionary note, because relatively small currents
are being measured in this test, it is important to clean and dry the bushings before
performing this test. The following cleaners have been suggested by various utilities: dry
clean cloth, water and soap, ColoniteTM, and WindexTM with Ammonia. It is never
recommended to use evaporative solvents on bushings.

162
 Figure 3-41: Hot Collar UST Mode Power Factor Test

Figure 3-42: Hot Collar GST Mode Power Factor Test

A hot collar test can yield one of three results: watt losses in normal range, increased
watt losses, or decreased current. Increased values in watt losses ( 0.1 W) typically
indicate contamination in the insulation system. Decreased values in current (compared

163
to similar bushings) may indicate the presence of voids in the insulation or low liquid or
compound level in the bushing.
3.2.11.2.4 What to do when Bushing Power Factor Tests are Doubtful
The following steps are helpful in confirming or clarifying bad bushing power factor
results:

1. Re-check all connections, including ground lead and bushing flange ground
2. Make sure ground connection is good
3. Check test circuit used for the measurement
4. Check test set and test set leads
5. Visually inspect bushing sheds and oil
6. Clean and dry all surfaces
7. Compare and analyze results of similar bushings
8. Research the history of the bushing for flashover or line surge activity
9. Verify temperature correction factor was used for C1 and overall tests (note that
C2 power factors are not temperature corrected)
10. If still uncertain, contact the manufacturer

3.2.11.2.5 Special Case – Type “U” Bushings [83]

3.2.11.2.5.1 History
General Electric, a major player in the electrical world since the early 1900s, was
engaged in the development and manufacture of apparatus bushings since as early as
1920. In the quest to develop the best bushing in the world, GE created many different
types and styles of bushings such as Types A, F, L, LC, OF, T, and U for both
transformer and circuit breaker applications.

Let’s concentrate on the Type U bushing history and technology first. Type U bushings
were manufactured with voltage ratings from 15 kV through 800 kV. A Type U bushing
is a condenser design with oil-impregnated paper inside an oil-filled shell. The shell
consists of a cap, an upper porcelain weather casing, a metal mounting flange, a lower
porcelain, and a lower porcelain support. For sealing purposes, all parts are held
together under a centrally clamped spring tension method. The principle behind a
condenser bushing is to incorporate equal capacitance layers to provide equal voltage
steps, resulting in a uniform voltage gradient throughout the bushing body and over the
bushing surface.

The type of design and the materials within a condenser core may differ between
manufacturers, but the design intention is the same. The type of construction used in
some Type U designs was a herringbone pattern, surface-printed ink that formed the
capacitive layers. A plain Kraft paper was wound into the condenser between the active
ink-lined paper layers. For most of the production, both the lined paper and the plain
paper were .008 inches in thickness (see Figure 3-43).

164
 Figure 3-43: Surface-Printed Ink Condenser

In 1979, American Electric Power Service Corporation reported increasing power

factors in Type U bushings at the Doble Client Conference. Since 1979, the concern for
the Type U bushing rising power factor has increased dramatically due to documented
accounts of bushing failures.

Do you have Type U bushings on your system?

Most likely you do. From 1954 to 1986, the time period that GE was manufacturing Type
U bushings, GE was the leader with 65 to 70 percent of the US market. They were
supplying bushings to their own transformer manufacturing facilities and to other
transformer manufacturers, as well as supplying replacement bushings directly to end
users. In this timeframe, the Type U bushing was known as the best product on the
market, utilizing standardization of parts with a proven field record.

So, what is the cause(s) related to the increase in power factor in Type U
bushings?
Through Doble Client Conferences, utility feedback, insurance company reports,
General Electric documents, and our own investigations, ABB has accumulated data
and has the following concern for Type U bushings.

The condenser design with ink-lined paper with plain Kraft paper allowed a gap at the
ends of the active layers in the condenser core. A heavily loaded transformer will
generate heat internal to the bushing, subject the bushing to a higher immersion oil
temperature, and consequently increase internal temperature in the bushing. The
heated bushing oil expands and intensifies the pressure in the confined gas space,
which causes an increased quantity of gas to become dissolved in the oil. Cyclic
reduction in transformer load and/or reduction of ambient temperature allow cooling of
the bushing oil. As the oil cools, it contracts, reducing the pressure of its gas blanket. If
the pressure reduction occurs rapidly enough, the gas-saturated oil will develop a
tendency to evolve bubbles of gas. This evolution will first occur in the highest

165
electrically stress regions of the bushings, normally between the lined paper and the
plain paper layers of the bushing core. A critical combination of gas bubbles and
dielectric stress causes partial discharges to occur within the insulation system.

The long-term effect of the discharges is an increase in the dielectric losses in the
insulation system, resulting in an increased power factor.

Have you heard of migrating ink?

This is a process that could also be a contributing factor to Type U bushing rising power
factors. Although GE designed and specified the herringbone ink process, they did not
manufacture the paper, nor did they apply the Rescon conductive ink. The paper/ink
process was completed by outside contractors. Reports as early as 1979 show that
portions of the Rescon ink “herringbone pattern” had transferred from the printed paper
layers to the plain Kraft paper layers. Investigations have revealed where Rescon
printed paper made contact with the overlapping plain paper, evidence of corona action
or evidence of slight burning was found. (See Figure 3-44) Ink/particulates aggravated
GE’s manufacturing system. During the cutting of hued and plain Kraft paper while
winding the condensers, ink/paper particulates were generated, further complicating the
rising power factor phenomenon. By 1985, GE had made many internal quality
improvements to the design and processing of bushings. GE implemented an oil
flushing procedure for all bushings in order to reduce the particles that may have
originated with the bushing core insulation. Also, GE commissioned a new closed-loop
continuous filtration oil system intended to improve bushing oil quality.

Figure 3-44: Rescon Conductive Ink Transfers from the Printed Paper Layers (left) to the Plain
Kraft Paper Layers or Conductor (right), Resulting in Corona Action and Slight Burning (circled)

166
What kV ratings of Type U bushings used herringbone ink processing?
The herringbone ink process was used in Type U bushings in the voltage range 15-345
kV. However, some Type U bushings in this voltage range have metal foil designed
condensers. Most bushings 345 kV and above have foil designed condensers, but many
have herringbone lined paper.

Should you be concerned only with Type U condenser bushings rated 15-345 kV?
Type U bushings were manufactured using a flex seal design. The flex seal is a copper
diaphragm located in the top cap of bushings 161 kV and above. The flex seal (see
Figure 3-45) was designed to allow for the expansion/contraction or movement of parts
during thermal cycling of the bushing.

Figure 3-45: Flex Seal Design

The flex seal diaphragm in many cases, depending on catalog number and application,
carries the current from the main conductor to the cap cover to the upper terminal
connection. As the diaphragm experienced movement, acting as an accordion, the
diaphragm could experience mechanical stresses, which would crack and result in a
leak. Since the diaphragm is internal to the bushing, and is placed above normal oil
level, where could the bushing leak?

During processing of the oil in the transformer, the oil could be evacuated from the
bushing by vacuum if the bushing was inclined, or the bushing could become filled with
oil during the transformer vacuum/fill process. If the bushing is full of oil (with no
expansion space) and if the bushing is applied at higher temperatures, the oil will
expand and compromise the gasketing system.

The flex seal system is connected to the main conductor with a swell seal gasket and a
seal nut. This connection is also under oil and under spring tension of the bushing. The

167
upper connection at the cover relies totally on the cover bolt tightness to adequately
carry the current from the flex seal through the cover to the customer terminal
connection. If the cover bolts have become loose over time, hotspots will develop, which
will compromise the cover gaskets. This situation is best revealed in the field by utilizing
thermal scans with infrared apparatus.

Hotspots such as this can lead to catastrophic failure if not resolved immediately. GE
recognized that the flex seal design could be improved upon, so they introduced the slip
seal design in 1976 (see Figure 3-46). The slip seal design totally eliminates the flex
seal but still allows the bushing to expand and contract during thermal cycling.

Figure 3-46: Slip Seal Design

What about top terminal overheating issues?

Many Type U bushings were designed and manufactured to have the ability to change
top terminals in the field. For instance, if a customer damaged the external threading of
a top terminal, they could replace the top terminal without removing the bushing from
the transformer. Also, draw lead bushings have a removable top terminal to allow
disconnection from the transformer winding lead without requiring entry to or removal of
oil from the transformer.

Type U bushings, if designed to have removable top terminals, require routine

maintenance to ensure top terminal tightness. If the top terminal becomes loose, a
hotspot may occur. Overheating of the top terminal may deteriorate the bushing’s
gasketing system, which could compromise the integrity of the insulating system and
possibly result in failure. Slip seal bushings, 161 kV and above, rated 1,600 amperes
and above, are perfect candidates for top terminal overheating if adequate maintenance
is not performed.

168
How do you know if your Type U bushings have herringbone ink condensers or
foil condensers, flex seal systems, slip seal designs, or removable top terminals?

Contact ABB! ABB Alamo has the documentation for all GE bushings. We have all of
the original design, test, and manufacturing data for Type U bushings. If you have the
catalog number and the group number from the nameplate of your bushings, ABB can
help identify the type of bushing design to evaluate your critical needs, such as bushing
maintenance, repair, refurbishment, or replacement.

Can a Type U bushing be refurbished?

Depending on the age, voltage class, current rating, design, and the condition of the
existing bushing, Type U bushings may be refurbished. Certain Type U bushings are
excellent candidates for refurbishment. If the bushing external parts are in good
condition and the concern centers on the herringbone ink condenser or flex seal
system, it is very economical to refurbish Type U bushings rated 115 kV and above or
bushings below 69 kV that have a high current rating (such as 4,000 amps and above).
The key to refurbishing Type U bushings is access to the original design documents and
having trained, experienced people. All bushings refurbished by ABB will be updated
with the latest ABB design enhancements and will carry a new nameplate and warranty.

Were Type U bushings manufactured and supplied to the field with oil
contaminated with PCB?

Yes! We cannot determine the content of PCB in a bushing by the serial number,
catalog number or the group number off of the nameplate. The only way to determine
the PCB level is to have the oil tested. We can give some guidelines. Bushings
manufactured by GE Pittsfield from 1954 to 1973 can have PCB levels that range from
50 to 500 ppm. From 1973 to 1980 we have test reports reporting levels from 2 to 50
ppm PCB. From 1981 to 1986 the levels are normally non-detectable or less than 1 ppm
PCB.

What criteria should be used to evaluate bushings on your system?

If you have bushings with herringbone-lined ink paper condensers, GE’s

recommendations, “Criteria for Concern,” for Type U bushings in 1979 were:

If the capacitance has increased by 10 % or more from nameplate, remove the

bushing from service.
If the P.F. is below 1.5 %, there is no cause for concern.
If the P.F. exceeds 1.5 %, but is less than 3 %, the bushing is in the region for
concern.
If the capacitance change is below 5 % of nameplate value, there is little risk of
failure.
If the P.F. exceeds 3 %, remove the bushing from service.

169
In 1985, Doble Company published recommended limits for Type U bushings.

A power factor of 1.0 % is questionable, rather than 1.5 %.

Today, ABB has approximately a 65 % market share of new bushings sold into the US
and is the leading supplier of replacement bushings for Type U bushings to the Utility
and Industrials in the United States. ABB has the following recommendations:

If the bushing power factor doubles original nameplate value, the bushing is
questionable and should be replaced or refurbished.
If the capacitance increases to 110 % of the original installation value, the
bushing is questionable and should be replaced or refurbished.

How can ABB make these recommendations, and on what basis can these
statements be made?

Being the sounding board for 170 major utilities and many industrials across the US, we
have seen the electrical industry increase awareness of Type U bushings due to high
power factors and failures of Type U bushings.

At the same time, we have noticed maintenance periods have been extended beyond
recommended levels. In today’s competitive marketplace, companies have downsized
maintenance programs and extended periodic maintenance from 1 year intervals up to
3 years and as high as 5 years or more.

Through field surveys and field experience, we have noted that if a Type U bushing is
exhibiting a rise in power factor, the rise accelerates very quickly once the action has
started. Therefore, many utilities know that if they are on a 3- or 4-year maintenance
interval and a bushing exhibits a rising power factor, the bushing will not perform for the
next 3- or 4-year period without failure.

The normal practice is to remove the bushings from the transformer immediately. Once
the corona (partial discharge) activity has started, the remaining service life of the
bushing can be very short, and it could fail catastrophically.
3.2.11.2.5.2 Recommendation

If possible, measure power factor and capacitance on a yearly basis.

If power factor is on the rise, replace or refurbish bushings. If you have flex seal
design bushings, thermal scan the units for hotspots, check for low or high oil
levels, and complete power factor and capacitance testing on a yearly basis. If
bushings exhibit any of the above-mentioned scenarios, the bushing should be
replaced or refurbished.

170
 If you have bushings with removable top terminals, proper maintenance must be
applied on a yearly basis either by thermal scan or manual inspection methods.
For manual inspection of top terminals, check to see if the terminal can be
loosened first. If the terminal cannot he removed, the terminal may have seen
overheating and/or corrosion build-up and should be removed from service.
If the terminal can be removed, inspect the top terminal gasket and look to see if
there are signs of corrosion. If the terminal gasket appears to be brittle or have a
permanent set, replace the gasket. When replacing the gasket, be sure to
lubricate the gasket with petroleum jelly to prevent twisting of the gasket as the
terminal is tightened. Tighten the top terminal to the correct torque values with
the proper tools or fixtures.

Top Terminal Size Torque

Inch – Threads ft lbs (N m)
1.125-12 35 (48)
1.500-12 100 (136)

If bushing top terminals show signs of corrosion or the top terminal cannot be
removed, we recommend replacement or refurbishment of the bushing. Top
terminal overheating can compromise the bushing gasketing system or create
loss of life of the bushing insulating system. This could result in a catastrophic
failure if the proper action is not taken.

Bottom connected bushings 161 kV and above rated 1,600 amp and above can be
refurbished to the new ABB Unified top terminal design per Figure 3-47. The ABB
Unified top terminal design eliminates top terminal maintenance and overheating,
corrosion, or deteriorating gasketing systems.

Figure 3-47: Unified Top Terminal

171
 Who can rebuild or refurbish Type U bushings to be like new?

Some major utilities have tried to rebuild their own bushings, a few small business
service shops have tried, and other bushing manufacturers have also tried to rebuild
Type U bushings. Most rebuilds by people other than ABB rely on guesswork or reverse
engineering to determine the makeup and design of the original bushing.

GE went through many design changes through the years. GE designed and
manufactured over 5,000 different catalog or styles of bushings, and within each catalog
or style there are an average of 7 design and manufacturing changes. That means there
are over 35,000 different Type U bushing designs in the field today. The key to
rebuilding Type U bushings is to have all the documentation, such as the drawings,
design changes, manufacturing processes, and test data. ABB has this design and
original manufacturing information as well as design engineers and technicians
experienced with GE technology. Table 3-37 shows typical design information for Type
U bushings. ABB will not rebuild bushings without the original design information. If
applicable or economical for the customer, ABB rebuilds Type U bushings to the latest
technology.
Table 3-37: Typical Type U Bushing Design Information

Bushing Current Herringbone Foil Removable Flex Slip Economical to Refurbish

kV Rating Ink Condenser Condenser Top Seal Seal
Design Terminal Design Design
15-69 400 yes yes
400/1,200 yes yes
2,000- yes
3500
4,000 yes
115-138 800 yes yes
800/1,200 yes
1,800 yes yes
161-230 800 yes yes See See
Note 2 Note 3
800/1,200 yes See See yes
Note 2 Note 3
1,600 yes yes See See yes
Note 2 Note 3
345 800 See Note1 See Note 1 yes See See
Note 2 Note 3
800/1,200 See Note 1 See Note 1 yes See See yes
Note 2 Note 3
1,600 See Note 1 See Note 1 yes See See yes
Note 2 Note 3
550 800 yes yes See See yes
Note 2 Note 3
800/1,200 yes yes See See yes
Note 2 Note 3
1,600 yes yes See See yes
Note 2 Note 3
800 800 yes yes See See yes
Note 2 Note 3
800/1,200 yes yes See See yes
Note 2 Note 3
1,600 yes yes See See yes
Note 2 Note 3
Note 1: To verify herringbone ink or foil design condensers, the bushing catalog # and group # from the nameplate must be supplied.
Note 2: To verify if bushing utilizes flex seal design, the bushing catalog # arid group # from the bushing nameplate must be supplied.
Note 3: To verify If bushing utilizes slip seal design, the bushing catalog # and group # from the bushing nameplate must be supplied.

172
Are there other reasons why a customer should refurbish bushings?

Depending on state and government regulations, the economical benefit in

refurbishment can vary. If a customer buys new product, what happens to the old
product? Most likely, the customer must scrap or dispose of porcelain materials, metals
materials, and bushing oil on top of dealing with the PCB issues. Sometimes the
disposal fees are very expensive. Do you know the regulations and laws of disposal in
your state? They are changing daily. Be careful. There are organizations that provide
services to decontaminate PCB laden bushings. The decontaminated parts can be used
to rebuild a bushing that carries the full warranties of a new bushing. Refurbishing
bushings could be an economic and viable solution to your problems.
3.2.11.2.6 Type “T” Bushings

Is the Type “T” bushing a predecessor to the Type “U” bushing manufactured by
General Electric?

Type “T” bushings were designed and manufactured by General Electric for low-voltage,
high-current, low-corona, transformer applications. GE supplied low-voltage, high-
current, stud type or bulk type bushings (Type “A” bushings) for many years, and then
the market demanded a bushing with low corona values. GE’s answer to the market
demand was the ultimate low-corona condenser bushing technology, the Type “T.” GE
manufactured Type “T” bushings from 1971 to 1985. Type “T” bushings range from 25
kV to 34.5 kV and current ratings 600 ampere draw lead to 18,000 ampere bottom
connected. These bushings were designed for low-voltage applications; therefore, GE
designed bushings for horizontal and vertical applications. To achieve maximum low
corona values, not obtainable by bulk type bushings, GE incorporated a condenser into
the design.

Why is there a concern with Type “T” bushings?

Type “T” bushings are basically designed and manufactured in the same manner as
Type “U” bushings. Outside shell and mechanical parts are very similar. What about the
condenser core process? The condenser design and process is the same as the Type
“U” using herringbone ink lined printed paper.

Should you be concerned about all Type “T” bushings?

No. Some Type “T” bushings are designed for high-temperature (125 °C) applications.
Units designed for high-temperature applications used Nomex winding paper with foil
inserts for gradients. The ink process could not be applied to the Nomex winding paper.

173
Is the concern for Type “T” bushings as valid as the concern for Type “U”
bushings even though they are a low-voltage bushing?

Yes! Even more so. The normal application of these bushings is on the low-voltage side
of a transformer with higher current ratings, higher temperatures, and sometimes they
are applied in bus ducts. When these bushings are subjected to thermal cycling, gas
bubbles trapped in high-stress areas of the lined ink printed paper condenser can create
partial discharge leading to a high power factor or failure of the bushing.

How do you know if you have herringbone ink lined paper or foil gradients in your
Type “T” bushings?

Contact ABB. If you know the General Electric catalog number and the group number
from the nameplate of the bushing, ABB can research the General Electric drawings in
our archives and verify the type of design. If you wish to discuss applications, such as
high temperature, ABB can also verify if the units are suitable for 105 °C or 125 °C
applications. Many transformer manufacturers, utilities, and contractors tend to misapply
bushings in high-temperature applications assuming that higher current rated bushings
can be applied at higher temperature ratings. Overload conditions described in IEEE
Standard C57.19.100 section 4 are normally abused more with Type “T” and bulk-type
bushings than other types of bushings. The updated ABB “Criteria for Concern”
(power factor and capacitance values) and recommended maintenance applies to Type
“T” bushings as well as Type “U” bushings.

Can you buy new bushings to replace Type “T” bushings or can Type “T”
bushings be refurbished?

Yes & yes! ABB offers direct replacement bushings for Type “T” bushings. ABB
manufactures Type “T” bushings today with the same dimensional and electrical
characteristics as the General Electric bushings for ease of installation, proper fit, and
application, but ABB has incorporated into today’s Type “T” the advanced technology
and superior condenser design of the ABB Type O Plus C bushing.

Although Type “T” bushings are low voltage, they are typically high current, and the
economics of refurbishment is well worth the effort. Normally, a refurbished bushing is
approximately 65 % of the cost of a new bushing. Please be aware that GE went
through many gasketing system design changes in the early stages of the Type “T”
design. ABB utilizes the original GE design data and drawings to update bushings to the
best design and latest technology when refurbishing bushings to “as new” condition.
3.2.11.3 DIAGNOSTICS AND CONDITIONING ON ABB BUSHINGS COMPLYING WITH THE IEC STANDARD

In general, bushings delivered from ABB shall be considered maintenance free.

However, inspection and field service experience will in some cases lead to the need for
diagnostics on bushings. In the following section, a review of different methods and
interpretations is given.

174
WARNING
Make sure that the transformer is de-energized and out of service before any
work is performed on the bushing.
3.2.11.3.1 Capacitance and tan measurement

Prior to taking a condenser bushing into service, and on suspected faults, the
capacitance and dissipation factor should be measured and compared with the values
given on the rating plate or in the routine test report. In connection with these tests, the
electrical connection between transformer tank and bushing flange shall also be
checked, for instance with a buzzer.
3.2.11.3.2 Temperature correction
The measured dissipation factor value shall be temperature corrected according to the
correction factors given in Table 3-38. GOx stands for all oil-impregnated paper
condenser bushings (OIP) and GSx stands for resin-impregnated paper condenser
bushings (RIP). For all bushings it shall be assumed that the bushing has the same
temperature as the top oil of the transformer. The test should be performed at a
temperature as high as possible. Correction shall be made to 20°C. The corrected
dissipation factor (tan ) shall be compared with the value on the rating plate or in the
test report.

Table 3-38 : Correction factors for tan

Range (°C) Correction to 20°C OIP Correction to 20 °C RIP

0-2 0.80 0.76
3-7 0.85 0.81
8-12 0.90 0.87
13-17 0.95 0.93
18-22 1.00 1.00
23-27 1.05 1.07
28-32 1.10 1.14
33-37 1.15 1.21
38-42 1.20 1.27
43-47 1.25 1.33
48-52 1.30 1.37
53-57 1.34 1.41
58-62 1.35 1.43
63-67 1.35 1.43
68-72 1.30 1.42
73-77 1.25 1.39
78-82 1.20 1.35
83-87 1.10 1.29

Interpretation of the measurement, OIP and RIP bushings

0-25% increase: The value is recorded and no further measures are taken.

25-40% increase: The measuring circuit is checked regarding leakage and external
interferences. External interference can come from nearby current carrying equipment
and bus bars. If the difference remains, the problem may be due to moisture. The

175
gaskets of the oil level plugs need to be replaced according to the product information
for the bushing. The measured value is recorded, and the bushing can be put back into
service.

40-75% increase: Perform the measures discussed for 25-40% increase and repeat
the measurement within one month.
More than 75% increase: The bushing shall be taken out of service. However, if the
dissipation factor is less than 0.4%, the bushing may be restored to service even if the
increase in percentage from the initial value is greater than 75%.

Capacitance: The measured capacitance, C1 shall be compared with the value given
on the rating plate of the bushing or with the 10 kV routine test report. If the
measurement is more than 3% from the nameplate value, there could be a partial
puncture of the insulation. An extremely low value C1 value (disruption) may be due to
transport damage and the bushing must not be returned to service. In either case,
please contact ABB. The C2 capacitance is influenced by the way the bushing is
mounted onto the transformer and should not be used for diagnostics.

Comments on dissipation factor of OIP bushings: The dissipation factor is a critical

property in oil filled condenser bushings and is mainly determined by the moisture level
in the paper and the amount of contamination in the insulation system. The power factor
is also very much dependent on the temperature; the principal behavior is shown in
Figure 3-48 for different temperatures and moisture levels.

Figure 3-48: Tan as a function of temperature and moisture level in OIP bushings.

It is clearly visible that the measurements at elevated temperature are more sensitive.
At 20 °C, moisture levels between 0.1% and 1% show approximately the same

176
dissipation factor. At elevated temperature (90°C) they differ by a factor of 5 or more.
For proper diagnostics, the important property is the dissipation factor at elevated
temperature and not the dissipation factor at 20 °C.

Comments on dissipation factor RIP bushings: Before a RIP bushing is put in

service on a transformer, it is possible for its tan value to deviate from the nameplate
value. This deviation is most probably due to moisture penetration into the surface layer
of the RIP. For example, this can happen if the bushing is stored without its protective
sealed bag. This allows air with high humidity to penetrate the outer surface layer of the
bushing. Normally, the tan value will decrease to its initial nameplate value if the
bushing is stored indoors, in a controlled humidity environment for a week. If the
transformer is energized with the bushing in service, the tan value will decrease to its
nameplate value within a couple of hours.

Comments on power factor measurements between the test tap and the mounting
flange on OIP or RIP bushings: There are several reasons to not use this value as a
diagnostic tool.

Primarily this dissipation factor is specified to be less than 5% according to IEC

60137. This means that unless specified, no attention is paid to decreasing this
dissipation factor value in the same manner as for the dissipation factor over the
main insulation.
The test tap is connected to the outermost earthed layer on the condenser body.
The solid layer outside the earthed layer contains an adhesive together with
cellulose to make the condenser body more stable. The dissipation factor of the
combination of cellulose and adhesive is much harder to control than that of only
cellulose insulation. It is for this reason that the dissipation factor of this insulation
section is not used for diagnostic purposes. Moreover, the adhesive material affects
the dissipation factor differently for different bushings.
It should be pointed out that under operational conditions, the outer layer of the
bushing insulation is earthed. Consequently, the insulation between the outer layer
and the mounting flange is not subjected to an electrical stress and therefore do not
cause any dielectric heat losses.
It is likely that if the bushing is placed in contaminated areas, contaminants on the
outside of the test tap affect the results. Moisture around the test tap also affects the
measurement.

It should be pointed out that if the test voltage (500V if the test-tap insulation is
delivery tested with 2kV and 2.5 to 5 kV if the test tap is delivery tested with 20kV) is
exceeded, partial discharges may occur in the region of the test tap. This will affect
the measurement.

Taking all the variations mentioned above into account, the dissipation factor of the test
tap insulation can vary between 0.4-3.0 %.

177
 3.2.12 MEASUREMENTS FOR ASSESSING THE CONDITION OF OLTCS/LTCS26

3.2.12.1 NUMBER OF O PERATIONS

It is common to measure the number of operations of the LTC. From the number of
operations, it is possible to estimate the level of deterioration of the device based on
experience. This measure is typically a function of the LTC manufacturer and type.
3.2.12.2 RESISTANCE OF THE ELECTRICAL CONNECTIONS
It is known that the initial contact resistance has a very strong influence on the
estimated useful life of the contact. If the connection resistance of the contacts is
known, it is possible to calculate an estimate of the remaining life of the contacts. This
is done with help of a mathematical ageing model that depends on such quantities as
the current load, the connection design, ambient temperature, and others. The contact
resistance can be measured with a micro-ohmmeter and the transformer in a de-
energized state.
3.2.12.3 T EMPERATURE
This measurement is based on the fact that under normal operating conditions, the
main tank of a transformer, because of the I 2R and eddy current heating, will have a
higher temperature than the LTCcompartment where there is essentially no heat
generation under non-switching conditions. Under steady state conditions, the
temperature difference between the two tanks will follow a known pattern. As the LTC
switch contacts age and wear, their resistance increases and hotspots develop under
normal loading conditions. The hotspots will increase the overall temperature of the
LTC tank, and the difference between it and the main tank temperature will begin to
deviate from the known pattern. The onset of severe contact wear can therefore be
estimated by using the temperature difference between the main tank and the LTC.
Most of the systems available on the market use magnetic clamp temperature sensors
and computer software to measure and track the temperature difference.
3.2.12.4 MOTOR CURRENT
Under normal operating conditions, the motor that drives the LTC gears and switching
contacts have a distinctive signature. Any significant deviations from this signature
may signal problems (gear or contact wear, binding, etc.) in the LTC mechanism. For
LTCs in which the switching mechanism is controlled by a spring, deviations of the
motor current from the normal signature can be used to diagnose looseness in the
tensioning of the spring.
3.2.12.5 ACOUSTIC SIGNAL
During the switching of the LTC, an acoustic signal is generated [84]. This signal can
be measured using a piezoelectric sensor. If there is a change in the gears or the
switching contacts, the acoustic signature will be different from the normal case. To
perform this diagnosis, the measured acoustic signals are compared with a certain
number of past signals using a sophisticated software program. Such a system can

26
See section 1.7.1 for comments on OLTC (IEC designation) or LTC (IEEE designation)

178
be applied online and can generate warning or alarm signals if certain values are
exceeded.
3.2.12.6 RELAY T IMING
The switching of the LTC is controlled by relays. The switching conditions of the
relays can be used to assess the switching process. This information may be
recorded and stored in a data Historian database. Any deviation of the relay timing
from normal signatures is an indication of possible faulty conditions in the LTC.
3.2.12.7 GAS-IN-OIL ANALYSIS
When an LTC operates, arcing occurs and the expected fault gases, acetylene and
hydrogen, are produced. The presence of certain levels of these gases in LTC oil is
therefore normal. However, as the contact wears or coking develops on the contacts,
the duration of arcs is increased during LTC operation. The generation of acetylene and
hydrogen increase accordingly. The rate of generation of these gases can therefore be
used as a means of determining the condition of the LTC contacts. Coking and
misalignment of contacts are the most common problems that occur in LTCs. The
coking process leads to exponentially increased heating or “thermal runaway” and
carbon buildup. It is now well established that the coking problem results in the
production of the ''hot metal gases'' (methane, ethane, and especially ethylene) [85].
The concentration of these gases depends on a number of variables including the
number of tap changer operations, breathing type, manufacturer, model type, etc. For
example, free-breathing LTCs rapidly lose gases to the environment, while sealed LTCs
retain much of the gases produced.
3.2.12.7.1 Items Specific to the European Practice

3.2.12.7.1.1 Scope

This section deals with the use of Dissolved Gas Analysis as condition indicator in
conventional On-Load Tap-Changers (OLTCs) that use mineral oil insulation. By
conventional OLTCs, we refer to OLTCs in which arc quenching takes place in the
mineral oil. The information in this section is therefore not valid for electronic OLTCs;
OLTCs with vacuum interrupters and OLTCs that use insulating liquids other than
mineral oil. It is mainly valid for OLTCs that have dehydrating breathers as interface
against to the atmosphere.
3.2.12.7.1.2 History
Historically, DGA of oil in OLTCs have been considered worthless because of the large
amount of gases normally generated by the arcs. This opinion has however, been
reconsidered in recent years and the understanding today is that quite a lot of
information can be gained from DGA of oils in OLTCs.
3.2.12.7.1.3 Faults in OLTCs possible to indicate by DGA

There are three basic faults in OLTCs that can be detected by DGA:
Discharges and arcings
Thermal faults
Ageing of cellulose insulation

179
Discharges and unwanted arcs cannot be detected because these are produced
during normal operation of the OLTCs. Ageing of cellulose is of no importance in
OLTCs manufactured by ABB since cellulose insulation is not used in the designs.
Thermal faults are possible to detect. During normal switching, the arcs generate
acetylene and hydrogen. In addition, the three gases indicating thermal faults,
methane, ethane and ethane are also generated. The temperature in the center of the
arc is several thousands centigrade and the molecules are totally degraded. Upon
recombination, the resulting gases are mainly hydrogen and acetylene. However,
there is a temperature gradient from the plasma channel in the center of the arc to the
surrounding oil. This gradient is high enough to produce all the thermal fault gases but
in certain ratios to the hydrogen and acetylene produced.
This relation between the gases is fairly constant as long as the gases are generated
by the arcs alone. If there is another source of thermal fault gases, such as an
overheated contact, the relation will change and a fault can be detected in an early
stage before any severe failures occur.
3.2.12.7.1.4 The Stenestam ratio

This ratio, named after its inventor, is the following:

(CH4+ C2H4+ C2H6)/C2H2
This is the relation between the thermal fault gases and the arc gas acetylene.
Hydrogen is not used since it is not reliable; it is highly volatility and has a low
solubility in the oil.
3.2.12.7.1.5 Important principals for interpretation of DGAs in OLTC

There are some important things to bear in mind before an interpretation is made
Never try to interpret DGAs where the gas amounts are very low. For a useful
ratio, the amount of acetylene should be at least 500 ppm.
A single sample does not give reliable information. The most reliable
information is gained when samples are taken within certain intervals to
produce a trend. In cases where the ratio is in the gray zone (between normal
and faulty), it is always recommended to take new samples with a certain
intervals in order to produce a trend.
Sampling and storage of samples are important for getting a correct result.
3.2.12.7.1.6 Interpreting the Stenestam ratio

These are the limits used to interpret the Stenestam ratio:

< 0.5 – No overheating present
0.5 to 5 – Take new samples. The higher the value, the shorter the interval
5 - An overheating has occurred. The unit should be taken out of service and
be repaired as soon as possible. Contact the manufacturer for advice!
In the range 0.5 to 5, the recommended intervals for new samples are as follows:
0.5 to 1 - 3 to 6 months
1 to 3 - 1 to 3 months
3 to 5 - within one month.

180
 3.2.12.7.1.7 Typical gas concentrations
The gas concentrations themselves do not give any useful information since the
concentrations are dependent on a large number of factors including load, number of
operations, piping, breathing system, temperature variations, oil volume, type of
connection and type of OLTC. However, to give an idea of what the levels are in some
typical cases, the following examples can be given in Table 3-39 (all values in ppm v/v):

Table 3-39: Typical Gas Concentrations in OLTCs

Low current/few “Normal Industrial

Gas operations operation” service

Hydrogen 1000-5000 1000-15000 <35000

Acetylene 500-5000 2000-30000 30000-150000
Methane >300 300-2000 <20000
Ethane <100 <500 <30000
Ethylene 50-300 300-5000 <70000
Propene <100 100-1000 <15000 1)
Propane <10 10-200
Carbon monoxide <700 <700 <700
Carbon dioxide 500-3500 500-3500 1000-3500
Oxygen 15000-35000 10000-35000 1000-35000
Nitrogen 40000-70000 40000-70000 40000-70000

1) The sum of propene and propane

As can be seen, the levels can vary very much. In case there are questions concerning
the gas concentrations, you are advised to consult the manufacturer. Typically, in order
for the manufacturer to provide proper advice about a unit, it is important for the user to
supply the following information:

OLTC type designation (from the rating plate on the motor drive mechanism)
Serial No. of the OLTC (from the rating plate on the motor drive mechanism)
No. of operations since last oil replacement
Rated load and average load since last oil replacement
Type of breather on the conservator
Daily load variations
Daily oil temperature variations
Faults, if any, that have occurred since last oil replacement

181
3.2.12.7.2 Important to bear in mind
The interpretation of DGA in OLTC oils is an evolving field of knowledge. The
information in this document is the best of our knowledge today. In the future, it is
expected that new information will complete and perhaps revise the content in this
document.
3.2.12.7.3 North-American Practice
When an LTC operates, arcing occurs and the expected fault gases, acetylene and
hydrogen, are produced. The presence of certain levels of these gases in LTC oil is
therefore normal. However, as the contact wears or coking develops on the contacts,
the duration of arcs is increased during LTC operation. The generation of acetylene and
hydrogen increase accordingly. The rate of generation of these gases can therefore be
used as a means of determining the condition of the LTC contacts. Coking and
misalignment of contacts are the most common problems that occur in LTCs. The
coking process leads to exponentially increased heating or “thermal runaway” and
carbon buildup. It is now well established that the coking problem results in the
production of the ''hot metal gases'' (methane, ethane, and especially ethylene) [86].
The concentration of these gases depends on a number of variables including the
number of tap changer operations, breathing type, manufacturer, model type, etc. For
example, free-breathing LTCs rapidly lose gases to the environment, while sealed LTCs
retain much of the gases produced.

Several users have established threshold gas concentration values for various models
and configurations of LTCs. It is advisable that users develop their own set of threshold
values. However, in the absence of such data, there are generic threshold values that
can be used. Table 3-40 is a summary of normal gas levels from an IEC study of
several normal and failed units. Note, however, that this study gave no difference to the
breathing configuration or type of interrupting mechanism.

Table 3-40: Normal Gas Formation as a Function of Operational Count [87]

Number of Gas Concentration (ppm)

Operations Hydrogen Methane Acetylene Ethylene Ethane
500 6,870 1,028 5,500 900 79
3,600 12,125 5,386 35,420 6,400 400
49,000 14,320 10,740 53,670 35,839 3,944

A similar table developed by Weidmann-ACTI laboratories provides normal gas

concentration levels for various LTC configurations (see Table 3-41).

182
 Table 3-41: 90% Values of Gas Concentration for Different Breathing Configurations [88]

Gas Concentration (ppm)

LTC Configuration
Hydrogen Methane Acetylene Ethylene Ethane
Free Breather 1,418 379 2,733 851 107
Free Breather
467 141 2,811 473 130
w/Desiccant
Sealed 1,781 572 3,744 1,337 178
Vacuum 72 61 42 49 79

Once any of the gases in a unit rises above the “normal” threshold limits, there are
several gas ratios that can be used to diagnose the specific problem. Analysis
performed by Jakob et al. [88] suggests that regardless of operational count, the ratios
of arcing gases to heating gases would remain the same in a problem-free LTC. As the
contact surface changes and becomes more resistive, there is a correspondingly
greater increase in the generation of heat gases than arcing gases. There is reason for
great concern if the concentration of heat gases is more than that of arcing gases in an
LTC. The two most useful ratios for diagnosing overheating problems in LTCs are the
Ethylene/Acetylene and the Ethylene/Ethane ratios. The calculated ratios for LTCs
under “normal” operation based on the data in Table 3-41 are given in Table 3-42.
.
Table 3-42: Diagnostic Gas Ratios for LTCs

Ethylene/Acetylene Ethylene/Ethane
LTC Configuration
Ratio Ratio
Free Breather 0.31 7.95
Free Breather
0.17 3.64
w/Desiccant
Sealed 0.36 7.51
Vacuum 1.17 0.62

3.2.12.8 MOISTURE
The insulation capability of the oil decreases with the water content. The water
content can be measured by a moisture sensor in an online monitoring system. Also,
the moisture content and oil breakdown values can be measured at the same time as
the gas-in-oil analysis.

183
3.3 ADVANCED DIAGNOSTIC TOOLS
3.3.1 ASSESSMENT OF MECHANICAL PROPERTIES - FREQUENCY RESPONSE ANALYSIS
(FRA)

This chapter is based on material from references [89], [90], [91], as well as own
experience within ABB.
3.3.1.1 INTRODUCTION
3.3.1.1.1 Purpose of FRA measurements
Frequency response analysis (FRA) is an advanced, non-destructive diagnostic method
used to detect mechanical movements or damages in the active parts of a transformer
(windings and core), by comparison with reference data from the same unit or from
other, similar units.
The goal of FRA is usually to test if physical displacements in the active part of the
transformer have occurred with age or after a particular event (e.g. refurbishment,
repair, accident, transportation, through fault, quality check, etc.). FRA measurements
provide indications of damage to the transformer, which can be investigated further
using other techniques or by an internal inspection.
A wide range of fault types is expected to induce characteristic changes in the FRA
signature, for example:

axial winding collapse

clamping failure
winding buckling
turn short circuit
broken core grounding
circulating currents in core
broken delta connection
bad short circuiting contacts

3.3.1.1.2 When should FRA measurements be performed?

It is most appropriate to perform FRA test after some incident or condition that has the
potential of causing mechanical movement or electrical damage to the transformer
assembly.
There are a number of good reasons for performing FRA measurements, depending on
the circumstances:

For prevention
Baseline (or “fingerprint”) measurement for future reference, for new transformers
usually on the test floor directly after manufacturing
As part of a routine diagnostic protocol, to check for changes during service time
After installation or relocation, to check for transformer integrity

184
For detection of damages
After unusual disturbance during shipment, to check for damages in the active
part
After failures in operation compromising the transformer condition (e.g. through-
fault event, close lightning strike)
As a failure diagnostic tool during short circuit testing of a transformer.

For sister reference

To obtain “good” reference data to compare with a sister unit in trouble
3.3.1.2 STANDARDS
FRA measurement procedures have recently been under study by IEEE and CIGRE
committees. The conclusions of Cigré Working Group A2.26 on FRA have been
published in spring 2008 [91], The IEEE guide is under preparation and a standard draft
of the IEEE WG PC57.149 on FRA [92] will probably be published in the near future.
Since 2005, China has an official FRA standard [93] (to our knowledge, as the first and
so far only country in the world). An IEC standard is in preparation [94] and will be
published in the near future (IEC 60076-18 Ed1).
3.3.1.3 GENERAL DESCRIPTION OF THE FRA METHOD
3.3.1.3.1 Principle of the measurement
FRA consists of measuring the electric response (expressed by transfer functions in the
frequency domain) of transformer windings over a wide range of frequencies and
comparing the results of these measurements with a reference set. A transfer function is
obtained as the ratio between the injected signal in a selected access of the transformer
and the signal received from another access of the transformer (see Section 3.3.1.5.5
for more information about the tested accesses in transformers). Usually the signal is
injected between a transformer bushing and ground, and the response is measured at
another bushing to ground (see Figure 3-49).

There are basically two distinct ways of injecting the wide range of frequencies required:
either via an impulse into the winding (impulse response method) or via a frequency
sweep using a sinusoidal signal (frequency response method). The impulse response
method has evolved from an earlier test method known as low voltage impulse
measurement, or LVI. Due to conversion of the time response into the frequency
domain using the technique of Fast Fourier Transform (FFT) it is in principle equivalent
to the sweep frequency method. Both methods are currently used within the industry
[95]. The impulse response method offers the advantage of a somewhat shorter
measurement time than the frequency response method. The frequency response
method on the other hand offers the following advantages over the impulse response
method:

Better signal to noise ratio

Equal, or near equal, accuracy across the whole measurement range
Wide range of injected frequencies

185
 Iin Iout

Zs
Device
Under Test
Us Uin Zp,in Zp,out Uout
(Transformer)

Figure 3-49: Principle of FRA measurements

Terminology in Figure 3-49:

Us input signal injected to the reference point
Zs internal impedance of the signal generator
Zp,in internal impedance of the measurement probe at the reference point
Uin measured voltage at the reference point, called input or reference voltage
Iin current injected to the transformer at the reference point
Zp,out internal impedance of the receiver probe at the measurement point, often
called FRA measurement impedance (50 for most systems)
Uout measured voltage at the measurement point, called output or response
voltage
Iout current at the measurement point

FRA measurements are usually performed in a frequency range from 10-20 Hz up to 1

or 2 MHz. For large power transformers, the interesting frequency range from the point
of view of winding integrity is typically from about 10 kHz to 500 kHz.

3.3.1.3.2 Practical set-up

A typical measurement set-up is shown in Figure 3-50. A voltage signal is injected via a
signal coax cable at some transformer terminal, and measured at the same injection
point with a separate measurement coaxial cable. The transferred signal at some other
terminal (usually the other end of the same winding) is measured with a third coaxial
cable. It is advisable to choose the coaxial cables all of the same length, in order to
compensate for phase lag and damping. All cable shields are connected to ground in
the shortest possible way, at both the transformer terminals and the measurement
device.

186
 Reference
Signal
Source Signal Cable Cable
Phase
Bushing

Uin
Neutral Response
Bushing Signal Cable
Uout
Source signal (swept Source and
frequency or impulse) Reference Cable Response Cable
Shield Grounds Shield Ground

Transformer
under test

FRA Device

Figure 3-50: Typical practical connection between FRA device and transformer

The input impedance of the measurement device is usually chosen to equal the wave
impedance of the cables (e.g. 50 ), in order to minimize signal distortion due to
reflections at the cable ends.
3.3.1.4 COMMERCIAL EQUIPMENT
In recent years, a number of dedicated FRA measurement devices have been
developed and introduced into the market. Although FRA measurements can be
performed with a general-purpose network analyzer, specialized modern equipment
makes the measurements easier and faster.

From the extended list of FRA measurement devices available today in the market, in
ABB Service groups three of them are the most recognized (Figure 3-51): FRAX from
Megger, FRAnalyzer from Omicron and M5300 from Doble.
3.3.1.5 DETAILED MEASUREMENT PROCEDURE
IMPORTANT: All tests should be performed by qualified test personnel who are familiar
with the test equipment. They should be capable of basic interpretation of the test
results, allowing them to distinguish between valid and invalid results (see Section
3.3.1.7 below), or at least have access to support for this purpose.
The following test procedures should be followed when performing the FRA test.

187
3.3.1.5.1 Test preparation
Firstly, the test should be done in a safe and controlled manner irrespective of
test location (Local Safety Recommendations). During field measurements, all
accesses at the high and low voltage sides of the transformer should be
grounded.
The transformer under test should be disconnected from any power systems or
supplies. Auxiliary accessories should be turned-off or not move during the test
(e.g. tap changers, pumps, fans, etc.).
The transformer tank will be used as the test reference ground, so insure you
have a good electrical connection to the tank.
The transformer should be preferably assembled, with bushings mounted, and
oil-filled.
Core-ground bushings should be connected to the reference ground (= the
transformer tank). They normally do not form a part of the test process.
Any deviation from the above (like un-mounted bushings, no oil, core
ungrounded, etc.) will most likely affect the test results and must be clearly
recorded in the test protocol.
It is strongly recommended to take pictures:
o of the transformer (overall),
o of its nameplate,
o of typical measurement connection used,
o of short circuit connections

Product Picture

FRAX-101
produced by
Megger

FRAnalyzer
produced by
Omicron

M5300
produced by
Doble

Figure 3-51: Some commercially available FRA instrumentation used within ABB

3.3.1.5.2 Tap changer position

Transformers with On-Load Tap Changer (OLTC) are suggested to be
measured in the tap combination that places all sections of the windings in the

188
 circuit. FRA test at other tap positions are allowed in particular circumstances
(e.g. to check for mechanical problems in OLTC).
Transformers with De-Energized Tap Changer (DETC) are recommended to be
measured in the working position.
On field measurements, it is necessary to perform FRA measurements using the
same OLTC position as defined in the reference measurements in order to make
a comparison possible. When no reference exists, proceed as recommended in
the two previous points.

3.3.1.5.3 Treatment of un-tested terminals

For terminals not under test, including neutrals, disconnect all terminals not
involved in the measurements and leave them in open circuit.
When windings not under test are required to be short circuited, the short-circuit
wires should be as short as possible using strong wire/braid.

3.3.1.5.4 Test leads:

Three test leads are recommended: Excitation source, reference (= in) signal and
response (= out) signal. They should ideally have the same lengths.
Test coaxial cable shields involved in the measurements must be grounded at
both ends (at the base of the test bushing flanges and at the BNC connection to
the test set).
The shielded test leads should terminate close to the bushing terminals or
flanges, and the area of the loop formed by signal wire, ground connection of the
shield, and bushing should be minimized.
The test-set ground should be directly connected to the reference ground (= the
tank) or follow manufacturer’s recommendations.

3.3.1.5.5 Test Set-up

The internal connection between windings (determining the “vector group”) of a
transformer defines the number and the type of FRA measurements to be performed.
We distinguish between four major groups of transformers according to their internal
connection:
Single phase transformers (normal or auto transformers)
Single phase transformers with tertiary (normal or auto transformers)
Three phase transformers (normal or auto transformers)
Three phase transformers with tertiary (normal or auto transformers)

Additionally, in the case of three phase transformers, the connection between windings
can be Wye (Y), with or without a neutral, or Delta (D). For special zig-zag or extended-
delta transformers, the list of test needs to be adapted following general principles
introduced here under.

In general, any FRA test on a transformer belongs to one of the following two kinds:
1. Self-winding tests measure the transfer function between the two ends of the
same winding (i.e., the signal is injected at one end of the winding and the
response measured at the other end of the same winding). If the transformer is

189
 connected in a Y, this is between a winding terminal and the neutral, for all three
phases of the high- or low-voltage windings. This kind of test is more common
and will be discussed in more detail below.
2. Inter-winding tests measure the transfer function between high-voltage and low-
voltage windings on the same phase of the transformer, for all three phases. This
kind of test will not be discussed further here.

For transformers designed and installed under the ANSI/IEEE standards, the following
self-winding test configurations are most common. The tests can be performed with the
un-tested terminals open or short-circuited. Open-circuit tests are performed on a
winding with all other connections floating and disconnected. (NOTE: The only
exception to this is where a delta winding has one corner completed external to the
tank). Short circuit tests are performed on a HV winding by shorting together the LV
connections, but without grounding them. The neutral is NOT included in the shorting
process.

Single Phase – Double Wound

Open Circuit Tests Short Circuit Test
All other terminals floating Terminals shorted together and floating
HV Winding LV Winding X1-X2 shorted H1-H2 shorted
Test 1 Test 2 Test 3 Test 4
Single Phase H1-H2 X1-X2 H1-H2 X1-X2

Single Phase – Auto Transformer

Open Circuit Tests Short Circuit Test
All other terminals floating X1-N shorted together and floating
HV Winding LV Winding
Test 1 Test 2 Test 3
Single Phase H1-X1 X1-N H1-X1

Two-Winding Transformers
Open Circuit Tests
All other terminals floating
HV Windings LV Windings
Test 1 Test 2 Test 3 Test 4 Test 5 Test 6
Delta-Wye H1-H3 H2-H1 H3-H2 X1-X0 X2-X0 X3-X0
Wye-Delta H1-H0 H2-H0 H3-H0 X1-X3 X2-X1 X3-X2
Delta-Delta H1-H3 H2-H1 H3-H2 X1-X3 X2-X1 X3-X2
Wye-Wye H1-H0 H2-H0 H3-H0 X1-X0 X2-X0 X3-X0

Short Circuit Tests

Terminals shorted and floating (neutral not included in the short)
HV Windings LV Windings
(X1-X2-X3 shorted) (H1-H2-H3 shorted)
Test 7 Test 8 Test 9 Test 10 Test 11 Test 12
Delta-Wye H1-H3 H2-H1 H3-H2 X1-X0 X2-X0 X3-X0
Wye-Delta H1-H0 H2-H0 H3-H0 X1-X3 X2-X1 X3-X2
Delta-Delta H1-H3 H2-H1 H3-H2 X1-X3 X2-X1 X3-X2
Wye-Wye H1-H0 H2-H0 H3-H0 X1-X0 X2-X0 X3-X0

A three phase auto transformer may have a single common neutral (H0X0) or three
separable neutrals (N1, N2, N3). A tertiary winding may be present; it is tested the same
way in either version of neutral bushing arrangement.

190
 Three Phase Auto Transformer – Common Neutral – Main Windings
Open Circuit Tests Short Circuit Tests
All other terminals floating X1-X2-X3 shorted together
and floating (H0X0 not
included)
HV Windings LV Windings HV Windings
Test 1 Test 2 Test 3 Test 4 Test 5 Test 6 Test 7 Test 8 Test 9
Wye-Wye H1-X1 H2-X2 H3-X3 X1-H0X0 X2-H0X0 X3-H0X0 H1-X1 H2-X2 H3-X3

Three Phase Auto Transformer – Neutrals Separable – Main Windings

Open Circuit Tests Short Circuit Tests
All other terminals floating X1-N1 shorted together
X2-N2 shorted together
X3-N3 shorted together
HV Windings LV Windings (shorted terminals remain
floating)
Test 1 Test 2 Test 3 Test 4 Test 5 Test 6 Test 7 Test 8 Test 9
Wye-Wye H1-X1 H2-X2 H3-X3 X1-N1 X2-N2 X3-N3 H1-X1 H2-X2 H3-X3

If the tertiary winding is brought out as three separate bushings (the corners of the
tertiary delta), then three separate tests may be performed. If only one corner of the
delta is brought out, as is the case for external completion of the delta winding, then
only one test may be performed.

Additional tests to be performed in the case of an auto-transformer with tertiary winding

are described in the table below:

Three phase Auto transformer – Tertiary Winding

Open Circuit Tests Short Circuit Tests
All other terminals floating Y1-Y2-Y3 shorted together
and floating
Test 10 Test 11 Test 12 Test 13 Test 14 Test 15
Single Corner Ya-Yb Not applicable
Full Delta Y1-Y3 Y2-Y1 Y3-Y2 X1-N X2-N X3-N

In the case of three winding transformers, the normal two winding transformer is
extended to take into account the tertiary winding as a winding coupled with the
secondary winding.

Three-Winding Transformer – (Wye-Delta-Delta)

Wye- Open Circuit Tests – all other terminals floating
Delta- HV Windings LV Windings LV Windings
Delta Test 1 Test 2 Test 3 Test 4 Test 5 Test 6 Test 7 Test 8 Test 9
H1-H0 H2-H0 H3-H0 X1-X3 X2-X1 X3-X2 Y1-Y3 Y2-Y1 Y3-Y2
Short Circuit Tests
X1-X2-X3 all shorted together Y1-Y2-Y3 all shorted together and Y1-Y2-Y3 all shorted together
and floating; all other terminals floating; all other terminals floating and floating; all other terminals
floating floating
Test 10 Test 11 Test 12 Test 13 Test 14 Test 15 Test 16 Test 17 Test 18
H1-H0 H2-H0 H3-H0 H1-H0 H2-H0 H3-H0 X1-X3 X2-X1 X3-X2

191
 Three-winding Transformer – (Delta-Wye-Wye)
Delta- Open Circuit Tests – all other terminals floating
Wye- HV Windings LV Windings LV Windings
Wye Test 1 Test 2 Test 3 Test 4 Test 5 Test 6 Test 7 Test 8 Test 9
H1-H3 H2-H1 H3-H2 X1-X0 X2-X0 X3-X0 Y1-Y0 Y2-Y0 Y3-Y0
Short Circuit Tests
X1-X2-X3 all shorted together Y1-Y2-Y3 all shorted together Y1-Y2-Y3 all shorted together
and floating; all other and floating; all other and floating; all other terminals
terminals floating terminals floating floating
Test 10 Test 11 Test 12 Test 13 Test 14 Test 15 Test 16 Test 17 Test 18
H1-H3 H2-H1 H3-H2 H1-H3 H2-H1 H3-H2 X1-X0 X2-X0 X3-X0

3.3.1.6 REPORTING OF FRA MEASUREMENTS

An attempt to standardize FRA reporting, including a proposal for a specific XML data
format, is currently made by an IEC working group [94].

A complete protocol of FRA measurement should contain:

3.3.1.6.1 General information:

Date and time where the measurements are performed,
Place where the measurements are performed,
Location of the transformer (e.g., inside or outside),
Ambient temperature
Customer name and person of contact,
Reason for test (routine, incident, new, warranty, etc)

3.3.1.6.2 Transformer information:

Complete nameplate information (preferably, a clear photograph of it)
Photographs of the test object as measured, showing the positions of the
bushings and connections
Reference number (factory code, serial number)
Rated voltage levels (HV and LV)
Rated power
Vector/Coupling group
Year of manufacture
Type of construction (e.g., core form or shell form), number of legs (3- or 5-leg),
winding type, etc.
Temperature of test object (e.g., top oil temperature),
On-Load Tap Changer (OLTC) position
De-Energized Tap Changer (DETC) position
Active part oil-filled (yes/no)
Bushings mounted (yes/no/test bushings)
Core demagnetisation performed (yes/no)

3.3.1.6.3 Description of each measurement:

Terminal on which the input voltage was applied
Terminal on which the response signal was measured

192
 Terminals not under test which were left floating
Terminals not under test which were short-circuited and/or grounded
Name of the file containing the measured data (magnitude and phase)

3.3.1.6.4 Instrumentation:
Name of the FRA equipment
Frequency bandwidth
Injected voltage level
Number of points per decade
Type of internal calculation, if performed (e.g., ratio, impedance, admittance, etc.)
Internal impedances of the voltage source, reference probe and response probe
(if deviating from the “standard” 50 )

3.3.1.6.5 Cabling:
Length and connection of coaxial cables
Length and connection of not-shielded cables

3.3.1.7 BASIC INTERPRETATION AND ON-SITE QUALITY CHECK

When performing an on-site FRA measurement campaign on a power transformer it is
essential to insure good quality and reproducibility of the data before leaving the site.
The reason is that it will usually be difficult, costly, or perhaps even impossible to return
later and repeat a faulty measurement on the same unit.

Below we provide some examples of “normal” measurement results for self-winding

tests on different transformer types. For all plots, the frequency scale was chosen
logarithmic so that the low frequency features are easily visible.

The basic idea is that the measurement engineer should have a clear idea of what kind
of behavior to expect, in order to minimize the risk of leaving the site with erroneous
measurement data.

193
3.3.1.7.1 Some “normal” FRA spectra

LV short-circuit

HV short-circuit

LV open-circuit

HV open-circuit

Figure 3-52: The four different self-winding tests on the middle limb of a two-winding transformer
(65/8.5 kV, 5.3 MVA, YNd11). The correlation of resonance peaks/dips in the FRA magnitude
with zero crossings in the FRA phase is indicated.

194
 U phase
W phase

V phase

Figure 3-53: Phase-to-phase comparison (measured phase-to-neutral) for open-circuit

measurements on the HV winding (same 3-phase transformer as in Figure ). Black curve:
middle limb. Blue and red curves: side limbs. Note the characteristic double-dip structure of
the side limbs.

195
 LV windings measured phase-to-neutral

HV windings measured phase-to-phase

(since the neutral is not accessible)

Figure 3-54: 3-phase transformer Yy, without neutral terminal at HV side (open circuit
measurements). Note the peculiar shape of the HV phase-to-phase response at frequencies
around 1 kHz which is also common for Delta-connected HV windings.

Figure 3-55: Single-phase transformer, end-to-end open circuit measurements. Black curve:
LV winding. Green curve: HV winding.

196
3.3.1.7.2 Meaning of different frequency ranges in an FRA spectrum

“global” resonances of the

transformer as a whole

short-circuit “local” resonances

inductance seen within a winding
short-circuit
inductance seen from LV side
from HV side

open-circuit
inductance seen
from LV side

open-circuit
inductance seen
from HV side

measurement
connections start to
influence

Figure 3-56: Information contained in different frequency ranges

(A) When only the current FRA measurement data are available:
3.3.1.7.3 Comparison between open- and short-circuit measurements
Typically, open- and short-circuit measurements deviate strongly at low frequencies due
to the magnetic core influence, up to somewhere between 10 and 100 kHz depending
on the size of the unit (the larger the transformer, the lower is this limit). Above that,
open- and short-circuit responses usually are quite similar. This holds for both high- and
low-voltage winding measurements. (See for instance Figure 3-52).

3.3.1.7.4 Comparison between high- and low-voltage windings

Open-circuit measurements on high- and low-voltage windings often display very similar
features at low frequencies. Sometimes the curves are, in the “negative-dB” range,
almost copies of each other which are vertically shifted with respect to each other by a
factor corresponding to the turn ratio squared (see e.g.Figure 3-52).

3.3.1.7.5 Comparison between phases

In a three-phase transformer (or in single-phase transformers with accessible
intermediate neutral point) the responses of the individual phases are usually very
similar to each other because of their identical winding design. However, there are three
typical differences between phases which often occur:

197
 1. Open-circuit measurements at low frequencies (up to somewhere between 1 and
10 kHz, depending on the size of the unit) display differences due to the
magnetic core influence. Typically the responses of the two side phase windings
are quite similar to each other (some difference between them may be
attributable to remanent magnetization) and have a double minimum of the FRA
response, whereas the middle phase is more different and displays a single
minimum (Figure 3-53).
2. Both open-and short-circuit measurements display differences at high
frequencies (MHz range) due to a phase asymmetry of the leads structure and
the tap changer.
3. In the case of transformers which have grounded, delta-connected tertiary
windings, it is common to observe a substantial deviation between the LV phases
around 20 kHz because of the phase asymmetry introduced by the ground
connection. Therefore, if the ground connection of the tertiary is accessible form
the outside, it should be removed while leaving the delta connection intact.

One important consequence of the above is that at least short-circuit measurements at

low frequencies (below some kHz) should always be very similar for the three phases
(or between sister units).

(B) When further data are available

3.3.1.7.6 Comparison with historical data
If earlier measurements on the same unit are available, it is strongly recommended to
compare them with the new ones directly on site (taking into account any actions
performed on the transformer between measurements, see section 3.3.1.7.8). Any
disagreement may point to measurement problems, which then can be addressed
immediately.
When comparing new measurements with previously performed ones, be aware (and
keep a note of) whether those measurements have been performed with the same
instrument and with the same cables. It is also very important to use the same
“measurement direction” (phase-to-neutral or neutral-to-phase) as in the previous
measurement, since these usually produce slightly different results.

3.3.1.7.7 Comparison with twin or sister units

Same comment as above for historical data: a comparison with available measurements
on identical or similar units may help to quickly detect measurement problems. As
already mentioned above in 3.3.1.7.5, it may also be useful to decide which level of
agreement is to expect in a phase-to-phase comparison.

3.3.1.7.8 History of the unit

When performing FRA measurements on site it is useful to know the reason for
performing the measurement. If there are previous events which may have affected the
integrity of the unit (like transport incidents, network short-circuit events, etc.) it helps to
have in mind some expectation beforehand about how these may affect the FRA
signature.

198
3.3.1.7.9 Other diagnostic data
In the same spirit, it is useful to collect other known diagnostic information beforehand
and try to form an opinion whether FRA deviations will be expected. Examples are
anomalies in turn ratio, short circuit tests, winding capacitance measurements, DGA or
DFR deviations, etc.

3.3.1.8 EXAMPLES OF PROBLEMS DIAGNOSED USING FRA

3.3.1.8.1 Axial Winding Collapse

Axial winding collapse is likely to have the following characteristics:

Produced within a transformer winding due to excessive axial forces during a
fault
Windings shift relative to each other
Gassing may result
Transformer integrity is compromised
Failure likely to be catastrophic if transformer continues in service

Figure 3-57 shows an example of axial winding collapse in a transformer.

Figure 3-57: Transformer with axial collapse of winding.

Figure 3-58 shows a typical signature of a winding that has experienced axial collapse.

199
 Right shift of resonance frequencies in
the damaged winding

Figure 3-58: FRA signatures of winding before (blue) and after (red) axial collapse. The damaged
phase shows a right shift of resonance frequencies, indicating a decrease of stray
capacitances between discs.

3.3.1.8.2 Hoop Buckling

Hoop buckling is produced within a transformer winding due to excessive compressive
forces during a fault. It would typically have the following characteristics:
The winding loses shape and gains a “bump” when seen on end.
It results in a bent–but not broken–winding.
Gassing may result.
The transformer is likely to be able to continue service.
Transformer integrity is compromised.

Figure 3-59 shows an example of hoop buckling failure in a transformer winding.

200
 Phase 1

Phase 2

Phase 3
Buckled X1 winding

Figure 3-59: Buckling of inner winding (X1) on Phase 1.

Figure 3-60 shows typical FRA signatures of windings that have suffered hoop buckling.

Figure 3-60: FRA Signatures of transformer with buckling of inner winding on Phase 1.

201
3.3.1.8.3 Shorted Turns
Shorted turns in transformers are produced by turn-to-turn faults and may have the
following characteristics:
Adjacent turns lose paper and braze/weld together.
They result in a solid loop around the core.

Figure 3-61 shows typical physical evidence of shorted turns in a transformer winding.

Figure 3-61: Shorted winding turns

Figure 3-62 shows traces from the same transformer before and after a fault that
resulted in shorted winding turns.

202
 –20

Magnitude [dB]

–40

–60

–80

0.1 1 10 100 1000

Frequency [kHz]

0 After the fault incident:

Phase A has clear short-circuit behavior
(reduction of the inductance).
Magnitude [dB] The other two phases have normal open circuit
measurement behavior.
–20

–40

–60

–80
0.1 1 10 100 1000
Frequency [kHz]

Figure 3-62: FRA traces for diagnosis of shorted turns.

203
 3.3.2 ASSESSMENT OF THERMAL PROPERTIES

3.3.2.1 DEGREE OF POLYMERIZATION (DP) [96]

Cellulose materials, such as paper and pressboard, form the major part of power
transformer insulation. The chemical structure of the cellulose polymer is shown in
Figure 3-63. The molecular formula may be written as (C6H10O5)n, where C6H10O5 is the
monomer unit and n is the number of monomers in the polymer chain. n is also known
as the degree of polymerization (DP or DPv).

The complete fine structure of the cellulose fiber consists of fibrils, micro-fibrils and
chains. The polymer chain is the ultimate fine structure of the cellulose fiber. The DP is
measured by viscosity measurements according to ASTM method D4243 after
dissolving the fibers in cupriethylene diamine solvent. Only the chains remain intact
when the fiber is dissolved. The average of several viscosity measurements gives the
DP value (the approximate number of chains remaining in the solution). Solutions of the
cellulose chains in the solvent increase the viscosity, depending on the size (molecular
weight) of the chains.

Figure 3-63: The Cellulose Molecule

Unprocessed virgin Kraft paper has DP in the range of 1,000-1,400; after drying and oil
impregnation the DP drops to approximately 900-1,100. As the insulation ages, the DP
drops to lower values with time. When it reaches a value of 200, the tensile strength
usually drops to about 20% of its original value and is generally considered at the end of
life. However, if the transformer does not experience short circuit forces and other types
of vibration, it may continue to function until the cellulose becomes carbonized and
brittle, at DP values of 100-150. Unprocessed wood fibers may have DPs as high as
1,500.
3.3.2.1.1 DP versus Life Plots

[97] provides an expression for the expected life of cellulose insulation to that follows
the equation:

204
 1 1
DPEnd DPStart 13350
Expected Life( yrs ) exp
A 24 365 T 273

Where: DPEnd and DPStart are respectively, the DP value of the paper at the end and
start of the life of the insulation; T is the temperature of winding in contact with the
insulation in degrees Celsius, and A is a factor that depends on the type of insulation
and the conditions of operation the insulation is under. The value of the A factor is
influenced by moisture in the insulation, acidity of the oil, and oxygen concentration in
the oil. Recent research performed at SINTEF Energy Research in Norway [98] has
shown that the values of A, which affects the ageing rate of the insulation, is quite
different for thermally upgraded insulation (Insuldur) than for non-upgraded Kraft paper.
A sample of the values they report for various conditions is replicated in Table 3-43.
These values assume activation energy of 111 kJ/mol.

Table 3-43: Estimated Ageing Factors, A

Insulation Conditions Kraft Paper Insuldur

Dry & Clean (2.0 ± 0.5) × 108 (6.7 ± 1.4) × 107
Acidic Oil (2.4 ± 0.7) × 108 (1.1 ± 0.6) × 108
Oxygenated Oil (8.3 ± 2.8) × 108 (3.5 ± 1.4) × 108
1% Water Content (6.2 ± 2.9) × 108 (1.1 ± 0.5) × 108
3-4% Water Content (21.0 ± 7.8) × 108 (2.6 ± 1.7) × 108

If it is assumed that the DP of transformer insulation is 1,000 at the start of life and 200
at the end of life, the expected life under the conditions specified in Table 3-43 can be
calculated using the life expectancy equation above. Figure 3-64 shows the life
expectancies of Kraft paper, and Figure 3-65 shows the life expectancies of thermally
upgraded paper using the nominal values of the A factors in Table 3-43. The life
expectancy curves for nonupgraded and thermally upgraded insulation in power
transformer as given in the IEEE C57.91-1995 loading guide are shown on the
respective charts for comparison. If we consider upgraded insulation life at 110oC hot
spot temperature, the Lundgaard formula gives a life of 9.4 years for dry insulation while
the IEEE curve gives a life of 20.5 years. This may be attributed to the difference in the
amplitude and exponent constants in the Arrhenius equations. For example, the
activation energy used in the Lundgaard equation is 13,350, while that used in the IEEE
equation is 15,000.

205
 10000.0
Dry & Clean (Kraft)
Acidic Oil (Kraft)

1000.0 1% Water Content (Kraft)

3-4% Water Content (Kraft)
Life Expectancy (years)

IEEE C57.91-1995 Curve for 55C Insulation

100.0

10.0

1.0

0.1
50 60 70 80 90 100 110 120 130 140 150
o
Temperature [ C]

Figure 3-64: Life Expectancy of Kraft Paper Insulated Transformers

100000.0
Dry & Clean (Insuldur)

Acidic Oil (Insuldur)

10000.0
1% Water Content (Insuldur)

3-4% Water Content (Insuldur)

Life Expectancy (years)

1000.0
IEEE C57.91-1995 Curve for 65C Insulation

100.0

10.0

1.0

0.1
50 60 70 80 90 100 110 120 130 140 150
o
Temperature [ C]

Figure 3-65: Life Expectancy of Insuldur Insulated Transformers

206
3.3.2.1.2 Latest Research Findings on DP Analysis

The following is a summary of the latest findings on DP analysis for power transformer
diagnostics:

Most researches and transformer insulation experts suggest DP at the end of life
in the range 100-200, with most favoring a value of 200 [99].
Pahlavanpour [100] presents results that show that the DP of Kraft paper starts
decreasing at 120°C. The rate of decrease of DP increases rapidly with
increasing temperature and reaches end of life at 180°C.
Hill [101] found a nonlinear relationship between tensile strength and DP. They
found the tensile strength of Kraft paper decreases slowly with decreasing DP
until the DP reaches a critical value of 500. At this point, the decrease in tensile
strength is more rapid with decreasing DP.
Lundgaard [98] found that the ageing of Insuldur paper is slower by a factor of
about three and is less sensitive to moisture; the activation energy for the ageing
of Insuldur paper is the same as that of Kraft paper; the ageing of Insuldur does
not produce as much furfural (furans) as Kraft paper; the effect of water on
ageing of transformers is more dramatic than oxygen; ageing of oil also
increases the acidity of the oil; Insuldur paper produces more acids than Kraft
paper.
Moser [102] reports that an increase of 0.5 % water content in an ageing
transformer will reduce the value of DP by one-half.

If we have actual DP measurement or an estimated value of the DP at any given time, it

can be substituted for the value of DPStart in the above equation to estimate the
remaining life based on the prevailing insulation conditions. The difficulty, however, is
that in order to get a sample of paper, the transformer must be opened. Moreover, the
areas of greatest deterioration of cellulose material in a transformer (the hottest spot), is
usually not easily accessible for obtaining paper samples. For this reason, other
methods such as the furanic analysis discussed in the next section, are used to
estimate the DP of the transformer insulation.
3.3.2.2 FURANIC COMPOUND ANALYSIS
This method of transformer condition analysis involves the measurement of trace
furanic compounds in oil in transformers and relating the furanic content to the degree
of polymerization of the paper insulation and to the remaining life of the insulation.
3.3.2.2.1 Origin of Furanic Compounds
The furanic compounds are five-member heterocyclic ring compounds that are
produced when cellulose breaks down, due to thermal stress. Cellulose degradation
produces other products also, such as carbon oxides and water, which are the ultimate
degradation products. The furanic compounds are intermediate degradation products,
which are liquids, and remain in the oil in trace quantities. At least six furanic
compounds have been detected in transformer oil in units: furoic acid, 5-hydroxymethyl-

207
2-furfural (HMF), furfuryl alcohol, furfural, 2-acetylfuran, and 5-methyl-2- furfural,
according to Burton and others who published the original research on furanic analysis
in 1984 [103] (see Figure 3-66).

Figure 3-66: Furanic Compounds from Paper Degradation

3.3.2.2.2 Detection of Furanic Compounds

The most common analytical technique is HPLC (High Performance Liquid
Chromatography), which involves extraction of the furanic compounds from oil, followed
by injection of a small quantity of the extract into special column in the HPLC
equipment. The column separates the components, and a chromatogram of peaks is
produced that shows the individual compounds and its concentration. Details may be
found in IEC Standard 61198 [104] and other publications. Other techniques such as
GC-MS [105] and colorimetry [106] are described in the literature.

3.3.2.2.3 Correlation Curves of Furanic Content with DP

Furan levels in transformers are typically less than 0.1 ppm (some laboratories report
furan content in parts per billion (ppb) – 1,000 ppb is equivalent to 1 ppm) and can
remain at this level throughout the life of the transformer. However, in many older units
levels of up to 1 ppm, and in some cases 10 ppm, have been measured [107]. In a
study [108] of over 5,000 European transformers, a significant number were found to
have furan content higher than 1 ppm. Several researchers have reported correlation
curves between the furan content in oil and the corresponding average DP of the
cellulose insulation. Although none of these equations are exact, they allow one to
estimate the DP of the insulation from the more easily obtained furan content. Four
equations proposed by Chendong [109], DePablo [110], Pahlavanpour [111], and
Shkolnik [112] are given below:

1.51 Log10 (F)

Chendong : DP
0.0035
7100
DePablo : DP
8.88 F
800
Pahlavanpour : DP
(0.186 F) 1
1.17 Log10 (F)
Shko ln ik : DP
0.00288

208
Where: DP is the estimated DP value and F is the 2-furfural (or furan) content in ppm. A
graphical representation of these equations is given in Figure 3-67.

1000
Estimated DP Value

Chendong - Kraft/Upgraded
DePablo - Kraft
Pahlavanpour - Kraft
Shkolnik - Upgraded

100
0.01 0.1 1 10 100
Furfural Content (ppm or mg/liter of oil)

Figure 3-67: Correlation of Furan Content with DP

The latest research on this topic as mentioned in section 3.3.2.1.2 suggests that the
ageing of thermally upgraded Insuldur paper does not produce as much furans as Kraft
paper. Other findings [113, 114] are that the most significant production of furans occurs
below a DP value of 400. It is known that the Shkolnik curve was derived from data
collected on transformers with predominantly thermally upgraded insulation and is in
agreement with recent research findings. Since the Pahlavanpour and DePablo curves
were derived from European transformers, it is highly likely to be for predominantly Kraft
insulated transformers and the recent research discussion bears this out. The
Pahlavanpour curve is derived from the same data as the DePablo curve, but assumes
a more realistic ageing pattern for transformers: 20% of the winding paper and the inner
paper layers degrade twice as fast as the rest of the paper insulation. The
Chendongcurve is suspected to be derived from transformers with mixtures of Kraft and
thermally upgraded insulation [115].

Use the curve in Figure 3-67 that is closest to the insulation type for the transformer
under consideration to estimate the DP value. Once an estimate of DP has been
obtained, the equation and conditions outlined in section 3.3.2.1.1 can be used to
estimate the remaining life of the insulation.
3.3.2.2.4 Issues to Consider in Using Furan Analysis
Below is a list of some of the issues that should be considered in applying this analysis
to transformers:

209
 1. Long-term stability of furanic compounds in oil: If these compounds are not
stable in oil for long periods, we should at least know the rate of degradation. If
we know the rate of generation from lab studies, we could perhaps adjust for
the losses and estimate the absolute furanic content.
2. Distribution of furanic compounds between paper and oil: The furanic content
of oil is related to the furanic content of the paper insulation. The latter is
dependent on the paper to oil ratio and the temperature. The distribution
between paper and oil should be known over a wide range of ratios and
temperatures (just as moisture distribution between paper and oil).
3. When the oil in a transformer is changed (as in a reprocessing operation),
most of the furanic compounds are lost. This is similar to dissolved gases
being lost during oil change. It may, therefore, be necessary to maintain
accurate records of prior analysis.
4. Correlation curves for different types of paper and pressboard materials would
be needed to make meaningful correlation to DP from furanic estimations.
5. Thermally upgraded paper and non-upgraded paper give significantly different
results because of the chemicals used in thermally upgraded papers. Make
sure you are using the proper correlation curves for your transformers.

It is not advisable to de-energize a transformer based on furanic analysis alone. This

test just provides an indication of the health of the paper. Furanic analysis is
recommended by many experts to give an indication of remaining life when the CO2/CO
ratio is less than 3 or greater than 10.

210
 3.3.3 DIELECTRIC FREQUENCY RESPONSE AS A TOOL FOR TROUBLESHOOTING
INSULATION POWER FACTOR PROBLEMS

3.3.3.1 INTRODUCTION
The insulation power factor (or dissipation factor, tan 27) measured at power frequency
is the most commonly used electrical test in North America and elsewhere for routine
evaluation of the insulation system of large power transformers and bushings. Typically,
defects in the insulation show up as either high or “unusual” power factor values [116].
While the test provides an important benefit to identify when a problem exists in the
transformer insulation, it is often difficult to determine the exact cause of the high or
unusual power factor reading. Often, the utility owners of the transformers will opt to just
process the transformer to remove moisture in an attempt to correct the power factor
results. This expensive process often works if the power factor problem is attributable to
moisture. However there are many instances when the underlying cause of the high
power factor remains unknown.
3.3.3.2 DIELECTRIC FREQUENCY RESPONSE AND X-Y MODEL
The dielectric frequency response test (DFR) has been recently developed as a
diagnostic tool for transformer insulation system testing. The DFR measurement is
similar to the power factor or tan measurement, except that it is a series of power
factor measurements at multiple frequencies. The advantage of doing the measurement
at multiple frequencies is that it provides much more information which makes it
possible to distinguish properties of both the cellulose and oil insulation separately.
Since the effect of moisture and ionic contamination on the dielectric properties of the
insulation system is more pronounced at low frequencies, the preferred measurement
frequency range is 1,000-0.001 Hz. The dielectric properties evaluated are the real and
imaginary capacitances (or permittivities) and the dissipation factor. Figure 3-68 shows
an example of a dielectric frequency measurement showing the complex permittivity
with the 50 Hz values marked.

The DFR measurement is similar to the power factor or tan measurement, except that
it is a series of power factor measurements at multiple frequencies. The advantage of
doing the measurement at multiple frequencies is that it provides much more
information so that the dielectric parameters of the insulation may be determined. Since
the effect of moisture and ionic contamination on the dielectric properties of the
insulation system is more pronounced at low frequencies, the preferred measurement
frequency range is 1,000-0.001 Hz. The dielectric properties evaluated are the real and
imaginary capacitances (or permittivities) and the tangent delta (dissipation factor),
which is essentially equivalent to the power factor for the range of values under
consideration.

27
power factor (PF) is essentially equivalent to dissipation factor (DF) for the range of values under
considerations in transformer diagnostics. i.e. PF DF 1 DF 2

211
 Real part = Capacitance

Imaginary part = Loss

50 Hz

Figure 3-68: Example of dielectric frequency response measurement showing the complex
permittivity.

The DFR measurement is performed with an instrument capable of measuring the

complex impedance of the transformer winding insulation. One such device is shown in
Figure 3-69. A variable frequency voltage is applied to the object, and the complex
calculations are performed by digitizing and processing voltage and current signals. The
normal complement of measurements is possible, including winding capacitance to
ground with and without guard and the ungrounded specimen test. See section 3.2.7 for
coupling details.

Figure 3-69: IDA200 Dielectric Response Measurement System (Courtesy of General Electric)

Knowledge of the transformer designs is helpful in identifying the test setup and
connections. This information is also needed later for the analysis of the DFR results to
determine the dielectric properties of the insulation, including moisture content.

212
The analysis of the DFR measurement involves modeling the transformer insulation
materials and structure as measured [117]. Obviously the insulation structure is quite
different for different types or designs of transformers. The detailed information required
for this model is obtained from manufacturing drawings of the transformer insulation
system. For modeling purposes the complex geometry is represented by a simplified
equivalent structure, the X-Y model, Figure 3-70. The relative barrier content, X,
represents the solid insulation in series with the oil in the main insulation. The relative
spacer coverage, Y, represents the relative amount of the circumference that is covered
by the spacers. Typical values for X and Y are 10 – 40 %. In the absence of design
data, representative values for these parameters can be suggested by rough estimates,
typical values are for example X=Y=20 %.

Spacer y
Barrier

Oil 1-y

x 1-x

Figure 3-70: Typical Layout of Transformer Insulation Duct

In the modeling the oil is characterized by its DC conductivity and power frequency
permittivity. The dielectric properties of the cellulose are characterized by DC
conductivity, power frequency permittivity, and dielectric response function. ABB has
compiled a database of such dielectric properties for oil-impregnated cellulose at
various moisture contents and temperatures. This database is merged into an analysis
tool developed by ABB that models the insulation geometry and the insulating materials
(oil, paper, pressboard, etc.) of the transformer.

In the modeling procedure, the design information and test data as described above are
input into the algorithm that calculates the response of the composite system. The
moisture in cellulose and oil conductivity values could then be optimized for a best fit of
the calculated response curve to the measured DFR data. In particular the influence of
the oil and the solid insulation on the dielectric response could be separated from each
other. Figure 3-71 shows an example of the ABB analysis tool. The red markers are
measured complex capacitance of a new dry transformer; the green markers measured
complex capacitance of an old wet transformer; and the solid line is a model fitted to the
old wet transformer measurement. The model provides an estimate of the moisture
content of the cellulose and the oil conductivity.

ABB has used these tools for the past several years for analysis of transformers both in
the factory and in the field. The experience gained from the tests and analyses have

213
shown the potential of the DFR test for identifying not only moisture problems, but also
other defects in the transformer insulation structure.

Model

Old Transf.

New Transformer

Figure 3-71: ABB analysis tool for DFR measurements on transformers. Measurements of New dry
transformer (red markers), Old wet transformer (green markers) and Model (lines) fitted to old
transformer estimating moisture content of cellulose and oil conductivity.

3.3.3.3 CAUSES OF HIGH POWER F ACTOR IN T RANSFORMER INSULATION

The traditional view is to ascribe high power factor measurements to either moisture in
the insulation or ionic contamination of the oil and/or cellulose insulation. However,
there are several conditions in a transformer that can lead to high or unusual power
factor measurement results. Some of the causes that have been diagnosed by ABB
experts are listed below:

Moisture in the cellulose insulation

High oil conductivity due to ageing or overheating of the oil
Chemical contamination of cellulose insulation
Carbon tracking in cellulose
High resistance in the magnetic core steel circuit

In the sections that follow a description of the use of DFR measurement to diagnose
each of these conditions is presented.
3.3.3.3.1 Comparison of DFR to Power Factor Measurement
In our analysis so far, it has become evident that several factors contribute to the proper
diagnosis of the status of transformer insulation. Some of these factors have either
significant or no effects on the power factor measurement at power frequency.

214
3.3.3.3.2 Influence of Oil Conductivity and Moisture on PF and DFR
The conductivity of mineral oil in a transformer can be affected by several factors such
as moisture, temperature, contamination, etc. The conductivity is directly related to the
dissipation factor of the oil and thus to the dielectric losses of the transformer. The
dielectric properties of cellulose insulation are also affected by moisture in the
insulation[118]. As a result, one cannot differentiate between the influence of moisture
and oil conductivity for the power factor measurement. This is demonstrated by the
following example. Figure 3-72 shows a typical 60 Hz power factor measurement result
for a single phase shell form transformer. Along with the power factor measurement is a
plot of the calculated DFR for the transformer assuming two different scenarios: one
with low moisture and aged oil and one with higher moisture and new oil.

As can be seen in Figure 3-72, it is not really possible to predict the moisture in the
insulation or the oil characteristic from the power factor test. In fact, it is entirely possible
to misinterpret a power factor reading as indicating high moisture when in fact it may
really be due to bad oil conductivity.

The DFR test on the other hand gives information to predict both the oil conductivity and
the moisture content in the cellulose insulation. Figure 3-73 shows the same information
as Figure 3-72, only the DFR test result is added. As can be seen from this data, the
actual transformer characteristic is 0.7 % moisture and an oil conductivity that is only
slightly aged.

1.000

Aged Oil, 0.5%

Moisture

Good Oil 1.3%

0.100
Moisture
Tan D

PF =. 00324

0.010

0.001 1 1 8 3 5

.001 .01 .1 1 10 100 1000

Frequency, Hz

Figure 3-72: Comparison of Power Factor Measurement to Two Different DFR Possibilities

215
 1.000

Aged Oil, 0.5%

Moisture
Good Oil 1.3%
Moisture
0.100
PF =. 00324

Tan D
Measured DR

0.010

0.001 1 1 8 3 5

.001 .01 .1 1 10 100 1000

Frequency, Hz

Figure 3-73: Comparison of Power Factor and DFR Measurement to Two Different DFR
Possibilities

3.3.3.4 DIELECTRIC FREQUENCY RESPONSE SIGNATURE AND I DENTIFICATION T ECHNIQUES

A further enhancement has been made to the DFR test by ABB experts to make it into a
defect identification tool. The method processes the measured DFR data and converts it
into a normalized signature function that is sensitive to changes in the normal dielectric
loss frequency spectrum. For the sake of this discussion, this normalized function is
referred to as the Dielectric Frequency Response Signature (DFRS). To use the DFRS
as a means for identifying defects in transformer insulation, it is necessary to
understand and model the internal and external insulation structures of the transformer.
Since different types of defects affect different portions of the insulation structure, it is
important to know the different possible test configurations and how to interpret the
corresponding results. Figure 3-74 shows a simplified model diagram for a two-winding
transformer that shows some of the possible insulation paths where defects may occur.

HV XV

CORE

HV Bushing XV Bushing

Figure 3-74: Simplified Insulation Circuit Diagram for a Typical Two-Winding Transformer

216
The first step in the identification process is to identify the particular insulation section
that is responsible for the high or unusual power factor result. This is accomplished by
making various measurements of the circuit including node-to-ground and node-to-node
with other nodes either grounded or guarded (these correspond to the UST and GST
tests of the traditional power factor measurements). In many transformers it is possible
to guard out the core or the bushings in order to sectionalize and isolate the various
insulation paths. The DFR test is then performed on this isolated part of the insulation
system and the resulting response curve is analyzed by the calculation method
described earlier. From this analysis, a DFRS is generated that is used in the defect
identification process.

The identification of the defect or the cause of high power factor is accomplished by
comparing the DFRS to other DFRS functions from transformers with known defects or
to results from laboratory tests. This comparison involves both the section of the
insulation structure tested and the shape of the DFRS. The insulation section under
evaluation depends on the type of test performed (for example, a test between the HV
winding and ground with the LV winding guarded would address one insulation section).
This is important since the shape of a DFRS varies depending on the particular
insulation section involved. The comparison of the shape of the DFRS involves a
comparison in magnitude and frequency of two DFRS functions, one from the unknown
specimen, and the other from the library of DFRS functions for known defects or causes
of high power factor. By matching the insulation section and the shape of the DFRS
functions, the potential defects or the causes of the high power factor can be identified.

3.3.3.4.1 Identification of high Core-Grounding Resistance Problems

A cause of high power factor in transformers is a high resistance in the core grounding
circuit, which may be caused by a higher than normal resistance between the individual
core laminations or between the core and the grounding strap. In the past, it has been
difficult to distinguish between high power factor readings caused by this higher core
grounding resistance and moisture in the insulation. As a result, a number of
transformers have undergone needless drying processes due to misdiagnosis. The
DFRS technique has proven to be helpful in identifying when a transformer has this
particular condition. Figure 3-75 shows a plot of the DFRS function for a normal and a
high core grounding resistance.

217
 Normal Core Ground

High Resistance Core

Ground

Figure 3-75: DFRS Functions for a Normal and a High Value of Core Grounding Resistance

This data came from measurements on a core form transformer when a resistor was
added in series with the core grounding strap. As can be seen from Figure 3-75, the
effect of a higher than normal core grounding resistance manifests itself as an increase
in DFRS at higher frequencies, but with no effect at lower frequencies. For the case
shown in Figure 3-75, the effect is also clearly detectable at power frequency. This
shows that a high core grounding resistance affects the power factor results at power
frequencies.

Figure 3-76 and Figure 3-77 show results where the DFRS method was used to
diagnose high core grounding resistance in cases where the transformer power factor
results were abnormally high. During a standard power factor test in the field,
Transformer 1 (shown in Figure 3-76) showed normal results, and Transformer 2
showed elevated power factor results above 0.5%.

The DFR test was performed to aid in the determination of the cause of the higher
power factor. The diagnostic interpretation of the DFRS functions for the two shell form
transformers is that the core-to-ground resistance is higher than normal for Transformer
2. This high core-to-ground resistance condition was later verified and corrected.

Shell Form Transformer 1

Shell Form Transformer 2

Figure 3-76: DFRS Functions for two Shell Form Transformers

218
Figure 3-77 shows the results of a test on a core-form transformer that exhibited high
(above 0.5 %) power factor results. The dielectric frequency response test was used to
isolate the section of the insulation that was causing the high power factor. The DFRS
function was then used to identify the cause. The insulation section was identified as the
low voltage to ground insulation and the analysis of the DFRS comparison identified the
cause as a high core-to-ground resistance. Upon inspection of the transformer, the high
core-to-ground resistance was traced to an auxiliary transformer used in a load tap
changer.

Figure 3-77 also shows the same DFRS function for the transformer after the auxiliary
transformer was modified to reduce the core-to-ground resistance. The DFRS function
measured after the repair is similar to the shape for a normal transformer core ground
as shown in Figure 3-75.

XV to Ground

XV to Ground after Repair

Figure 3-77: DFRS Function caused by a High Core to Ground Resistance in Auxiliary Transformer

Moisture in a transformer is concentrated in the cellulose insulation due to the moisture

equilibrium properties of the oil/paper system. Moisture in the cellulose causes an
increase in the dielectric loss of the insulation, which is reflected in higher power factor
test values. The DFR test is a valuable diagnostic tool for not only identifying if moisture
is the reason for a high power factor test result, but also for quantifying the actual
amount of moisture in the insulation. Figure 3-78 shows the effect of moisture on the
shape of the DFRS function. As can be seen in this example, the effect of moisture is
most pronounced at lower frequencies, which is the characteristic used in identifying
high moisture in transformers.

Normal Moisture

High Moisture

Figure 3-78: DFRS Function Showing the Effect of High Moisture

219
3.3.3.4.2 Identification of Paper Contamination Problems
Paper contamination in transformers is a rare occurrence. Nonetheless, it is of great
advantage to be able to recognize and differentiate between contamination and other
causes of high power factor. As with other types of transformer defects, this requires
experience from testing insulation systems, which happen to have this problem. With
limited experience in actual cases, it is not known if a single DFRS shape can be used
to identify all types of contamination. It is possible that different types of contamination
in different insulation sections will have different DFRS profiles.

Figure 3-79 shows a DFRS for a transformer where the high power factor was caused
by a chemical contamination of the windings. Although the profile is similar to that of the
high resistance core ground, the insulation section exhibiting the defect was different. In
addition, the frequency of the peak was lower than any of the core ground resistance
cases. This case opens an avenue for further study of contamination recognition using
the DFRS method.

Figure 3-79: DFRS Function Caused by Chemical Contamination of the Windings

3.3.3.4.3 Low Temperature Effect on Insulation Power Factor

This case involves a new 500 kV shell form auto transformer that was installed and
tested in the field. The test results showed more than a doubling of the power factor
between the factory values and the levels measured in the field after installation. The
DFRS method was employed to determine the cause of this increase in power factor.
Table 3-44 shows the comparison of power factor measured in the factory and in the
field. The DFR measurements were done both in the factory and in the field, and the
results are shown in Figure 3-80. The main difference in the conditions between the
factory and the field was the temperature of the transformer (33.5 °C in the factory and
9 °C in the field) and the actual oil in the transformer was different between the factory
and the field. The results of the DFRS comparison showed that both curves were
normal, and there was no degradation in the condition of the insulation system. The only
plausible explanation for the difference in the power factor results is that the
temperature correction factors used do not correctly represent the actual change in
power factor, especially when the temperature is below 20 °C.

220
 Table 3-44: Comparison of Factory and Field Power Factor Results (Corrected to 20°C)

Test Oil PF, % PF, % PF, %

Condition Temp. HV-gnd YV-gnd HV-YV
Factory 38 0.22 0.28 0.26
Field 9 0.35 0.39 0.41

Case 3 HV-XV DFRS

Measurement
Factory DFRS Data

Figure 3-80: DFRS Functions for Both Factory and Field Measurements on a Shell Form
Transformer

To further investigate the difference between the temperature correction factors used in
the industry, and the actual variation of power factor with temperature for this
transformer, a curve was developed using the DFR software. Figure 3-81 shows a
comparison of the calculated temperature correction to the commonly used correction
factors. From this curve, we can easily see why the power factor test results reported
were so different between the factory with an oil temperature of 33.5 °C and the field
with an oil temperature of 9 °C. It should be noted that this shell form transformer has
an unusually high ratio of paper-to-oil in the high-to-low space where the measurement
was made. It is not known at this point if this contributed to the exaggerated difference
between the actual variation in power factor vs. temperature and the Doble or ANSI
correction factor curves.

221
 2.4
2.2 ANSI Factors
2 Doble 230kV up
1.8 Case 3 Transformer
1.6

TCF
1.4
1.2
1
0.8
0.6
-5 0 5 10 15 20 25 30 35 40 45 50 55 60
Temperature,°C
Figure 3-81: Comparison of Calculated Temperature Correction Factor for Shell Form Transformer
to Commonly Used Correction Factors in the Industry

3.3.3.5 SUMMARY
This section demonstrates the usefulness of the dielectric frequency response signature
(DFRS) for identifying various causes of high insulation power factor, such as moisture
in cellulose, high resistance in the magnetic core steel circuit, and chemical
contamination. The method is suitable for troubleshooting power factor problems in the
factory or the field. In addition, it is important to emphasize that the dielectric response
test is used to estimate the dielectric properties of transformer insulation, such as the
average volume moisture [%] in solid insulation and the power factor of the oil. The DFR
method has been found to be more accurate than existing industrial methods, for
example, dew point test and moisture/oil equilibrium method. These methods
predominantly estimate the surface moisture of the insulation.

Using the derived DFRS method to diagnose the cause of high power factor will focus
maintenance resources in the proper direction and prevent costly drying or over-drying
of transformer insulation when moisture is not the issue.

It is clear that there could be problems in applying the Doble or ANSI temperature
correction factors for power factor for all transformers. As demonstrated in the one case
described above, the use of one of these factors can lead to erroneous conclusions
regarding the insulation condition. It is ABB’s opinion to not use any correction factor
unless a reliable calculated factor is used that is based on the actual properties of the
transformers insulation system. Further investigation is needed to determine a better
method for correcting power factor results that is applicable to different transformers.

222
 3.3.4 ASSESSMENT OF ELECTRICAL PROPERTIES - PARTIAL DISCHARGE
MEASUREMENTS [119]

3.3.4.1 PURPOSE OF MEASUREMENT

A partial discharge measurement (PD-measurement) is a non-destructive tool used to
establish the condition of a transformer insulation system. A PD-measurement makes it
possible to detect and localize areas within the transformer that are exposed to elevated
dielectric stresses. Partial discharge measurements are explicitly specified in standards
or in customer specifications. They are to be carried out in conjunction with dielectric
tests in high-voltage laboratories using AC-voltage in the power frequency range. For
onsite PD measurements (for example on repaired transformers) other types of PD-free
excitation may be required. Partial discharge measurement should generally be the last
dielectric test conducted on the transformer.

A partial discharge is a partial voltage breakdown within a series of insulating elements

between two electrodes of different potential, (capacitances C2’ and C3’, see Figure
3-82). During a typical PD measurement, the magnitude of the detectable value of
partial discharge activity is recorded as a function of the applied voltage. A partial
discharge can also be interpreted as the rapid movement of an electric charge from one
position to another.

Figure 3-82: Schematic of Part the Transformer Insulation

223
Where:
BU = bushing
HV = high voltage
NT = neutral terminal
C1, 2, 3 = active part of transformer (including oil)
C1 = weak region
’ ’
C2 and C3 = test object capacitance

If the two line terminals are connected together via an external capacitor Ck, the charge
movements within the series-connected insulation links (capacitances C2’ and C3’) will
also be reflected in the charge of the external capacitor, Ck. The charge movements can
be detected as circulating current impulses i(t) in the parallel-connected capacitors Ck
and the test object.
3.3.4.2 ELECTRICAL PD MEASUREMENT ON T RANSFORMERS
Circulating PD current impulses generated by an external PD source (in the test circuit)
or by an internal PD source (in the insulating system of the transformer) can only be
measured at the transformer bushings. Bushing capacitance C1 acts as the coupling
capacitor Ck, which is connected in parallel with capacitance Ct (test object = total
capacitance of the transformer insulating system). For power transformers, a measuring
impedance Zm, is generally connected between the bushing measuring tap and ground,
i.e. in parallel with bushing C2 capacitance (see Figure 3-83). For bushings without a
capacitive tap an external coupling capacitance Ck must be connected in parallel with
the bushing.

According to IEC, which is the preferred method for these tests, PD measurements are
conducted by measuring the “apparent charge,” q. In this context, the apparent charge
is obtained by integrating the PD current impulse using a “wideband” or “narrowband”
filter. The PD measuring system is connected via a coaxial cable to the measuring
impedance Zm (see Figure 3-83). The apparent charge q, measured in Picocoulombs
(pC), corresponds to the charge transferred during the V voltage drop compensation
process at one of the parallel-connected capacitances Ct (transformer insulation) and
bushing capacitance C1 or coupling capacitance Ck. This voltage drop V may be
caused either in the test object (internal partial discharge in the bushing or in the
transformer insulating system) or in the test circuit (external partial discharge). If PD
activity is detected during the test, the PD source must be investigated.

224
 Figure 3-83: PD Calibration and Measurement Setup for Transformers; Bushings with Capacitive
Taps

The magnitude of measurable apparent charge q in pC must be defined by the

calibrating procedure for each test circuit.
3.3.4.2.1 Calibration
Calibration of the PD test circuit is performed using a battery-operated calibrator. A
calibrator consists of a square-wave generator with adjustable amplitude Vo connected
in series with a small capacitor C0, which should be less than 10 % of C1. For PD
measurements on transformers, the calibrator is connected across the bushing, or
across the coupling capacitor connected in parallel with the bushing (see Figure 3-83).
Calibration must be performed separately for each bushing. Under the assumption that
C0 << Ck, the injected impulse from the square-wave generator corresponds to the
charge q, which is set to predefined values (100 pC, 1,000 pC, etc.) by the adjusting the
amplitude Vo. IEC 60270 recommends that the rise time of the injected impulse should
be less or equal to 60 ns, amplitude V0 between 2 V and 50 V, selectable polarity, and a
repetition rate of 100 Hz. Based on this, the injected charge is calculated as:

q0 = V0 × C0

The measuring circuit, consisting of the capacitance Ct of the test object, coupling
capacitor Ck, measuring impedance Zm, coaxial cable, and measuring system, is now
calibrated. During the PD test the measuring system values are read directly in pC. This
pC reading is only valid for the specific calibrated bushing.

225
3.3.4.2.2 PD measuring procedure
According to the IEC Standards [120], PD measurements shall be carried out in
conjunction with induced voltage test. Any wide bandpass filter or narrow bandpass filter
can be used as a PD measuring system. The first PD measurement should be made at
a low test voltage level (approx. 10 % rated voltage). This value serves as a reference
for the background noise level in measurement system. According to the lEC
Standards, the background noise level must be lower than half of the required pC value
of apparent charge for the specific transformer. PD activity must be checked at all HV
bushings on the transformer. The best way to accomplish this is to apply a multi-
channel PD measuring system capable of detecting PD activity at all bushings
simultaneously. A description of such a system (ICMsystem) used by some ABB HV
laboratories is given below.

The PD test is considered successful if no continuous PD activity greater than the

specified apparent charge amplitude in pC is detected at any bushing and if there is no
rising trend in the apparent charge amplitude during the long duration test. The
recommended acceptable values of apparent charge given in the IEC standards are:

300 pC at 130 % rated voltage

500 pC at 150 % rated voltage
The level of continuous PD activity does not exceed 100 pC at 110 % rated
voltage.
3.3.4.2.3 An Advanced PD system
The ICMsystem (manufactured by Power Diagnostix Systems GmbH, Germany) is a
specially designed modern PD system to meet the requirements of partial discharge
measurements on power transformers. The ICMsystem uses wide-band filters for both
digital data acquisition and further data processing of conventionally detected PD
signals. True parallel acquisition of PD impulse currents on multiple channels is
achieved by using multiple individual amplifiers (wide-band filters) connected to multiple
measuring impedances at the bushings via eight preamplifiers. The PD activity is
detected simultaneously on all multiple channels and processed in a controller unit. PD
readings can be weighted according to IEC in pC or according to IEEE [121] in V. An
example of the ICMsystem is shown in Figure 3-84. In addition to PD signal detection,
the ICMsystem offers independent channels for voltage measurements via a separate
tap at each measuring impedance, Zm.

226
 Figure 3-84: Advanced PD System, ICMsystem [122]

Besides conventional PD signal detection of the apparent charge values, the

ICMsystem is capable of performing a statistical analysis of the detected PD activity (i.e.
phase-resolving partial discharge analysis PD pattern) at each specific channel. A
PRPDA analysis produces a two- or three-dimensional PD pattern (phase angle,
discharge magnitude, and number of events are obtained, see Figure 3-85). For two-
dimensional PD patterns the third dimension (number of counts per channel) is
indicated by the color code. The PD pattern reflects the sum of all PD impulses
collected during a specific measuring time (for example, a preset time of 60 seconds
corresponds to 3,600 cycles for a 60 Hz test voltage power frequency). The PD pattern
provides a fingerprint of the partial discharge activity of a specific defect in the test
object.

A phase-resolved PD system presents great advantages over the conventional PD

system in its ability to perform a statistical analysis of the detected PD current impulses
(PD pattern). A conventional PD system delivers phase-resolved information about the
PD activity for only one cycle. A phase-resolving partial discharge analysis delivers
information for several cycles of the signal, which is important for identification of the
type of PD source for the following reasons:

PD patterns identify a specific type of PD source (image of the physical process, see
Table 3-45 and Table 3-46).
PD patterns are not influenced by the signal transfer function of the extended
insulating system (statistical behavior does not change).
PD patterns can be used to distinguish between superimposed PD defects on the
basis of different statistical behavior.

227
 Figure 3-85: Registration of PD Impulses in Advanced PD System (Statistical Analysis of PD
Impulses)

3.3.4.3 PROCEDURE FOR I NVESTIGATION OF PD SOURCES

If PD activity that exceeds the acceptable requirements is detected, the type of the PD
source and its location (external or internal in the insulating system) must be
investigated. The procedure for investigating the PD source should generally be
adapted to the behavior of the registered PD activity.

The first step is to exclude all possible external PD sources. Typical external PD
sources are:

Low-Voltage Power Supply: Noise such as thyrister pulses or harmonics from the low-
voltage power supply may especially influence a sensitive PD system, which is directly
connected to the power line. If these are present, a low-pass filter or insulating
transformer should be used. Due to the filtering effect of the step-up transformer of the
high-voltage source and of the HV filter in the PD system connection to the test object,
the noise from the low-voltage power line is usually sufficiently suppressed. If there is a
noise problem, a second step-up transformer may be used as an additional filter or a
PD system with a narrow-band filter could be used (f0> 1 MHz).
High-Voltage Source: An HV source must generally be PD free. If there is a problem,
the coupling capacitor can be connected directly to the source (without the test object)
to easily check the HV source.
High-Voltage Filter: In difficult cases (for example, a station with a lot of
electromagnetic interference), a PD-free, low-pass HV filter must sometimes be used.
Connections in the Test Circuit and Electrodes: All bushing tops (even grounded
bushings) and sharp metallic parts on top of the transformer (especially close to the
bushings) should be shielded. All connections should be PD free, i.e. with sufficient
radius). In addition, the bushing surface must be free of conducting particles. All
measuring impedances Zm, must have a good connection to ground. If there is a
problem, an ultrasonic detector (corona gun) may be used to detect an external PD

228
source. The PD type can be determined from the statistical analysis of the PD signals
(typical PD-pattern, see Table 3-45).
Coupling Capacitor: The coupling capacitor must be PD free. If there is a problem, the
coupling capacitor must be measured separately.
Conductive Objects Close to the Transformer under Test: Ungrounded conductive
objects close to the transformer under test could become charged to a high potential
due to the electric field. If the breakdown field value is reached, a pulse-like discharge
may occur. These PD impulses may be coupled to the PD test circuit and detected at
the measuring impedances and produce very high apparent charge amplitude. This PD
source can be recognized by comparing it with typical PD patterns, by visual
observation of the surroundings or by using an ultrasonic detector.

Table 3-45: Typical PD Sources in the Transformer Insulating System

Once all external sources have been eliminated and there is still partial discharge
activity, the process begins of identifying the type and possible location. In reality, the
five typical PD patterns appear in many variations. Because of the charging and
discharging effect at the PD site, there is continuous change of both the surrounding

229
area of the PD source and the PD source itself. Consequently, there are only a few PD
patterns that exhibit constant behavior during the test.

During an analysis, the basic PD pattern characteristics that should be analyzed are:

Phase position of the PD signals

Symmetry of the PD signals during the positive and negative sine wave
Number of PD signals per cycle
Reproducibility of the PD pattern

Interpretation and screening of the correct type of PD pattern from the real PD pattern
results requires experience and a strong interpolation capability. If PD defects are
superimposed, a comparison with the typical types of PD patterns and finding the
correct type of PD pattern becomes much more difficult. An overview of the typical PD
sources in the transformer insulating system together with their typical PD pattern and
their typical behavior during the test is presented in Table 3-46 . If there is a clear
indication of internal PD activity in the transformer insulating system, localization of the
PD source must follow. Localization of PD sources is more effectively done using
ultrasonic techniques as described in the next section.

230
 Table 3-46: Typical PD Defect Patterns in Power Transformer Insulation

Typical PD Defects in the Transformer Insulation Conducting Material = PD Pattern Type 1

231
 Typical PD Defects in the Transformer Insulation Conducting Material = PD Pattern Type 2

232
 Typical Defects in Transformer Insulation Bubbles = PD Pattern Types 3 and 4

3.3.4.4 ACOUSTICAL PARTIAL DISCHARGE MEASUREMENT ON T RANSFORMERS

3.3.4.4.1 Acoustic PD Wave Characterization

An acoustic PD signal is a mechanical vibration characterized by its frequency, f. A PD
source acts as a point source of acoustic waves. The intensity of the emitted acoustic
waves is proportional to the energy released during the discharge and may be
expressed as:

W1 =q1 V1

Where:
W1 = locally released energy
q1 = local charge
V1 = local voltage drop

Acoustic wave propagation occurs only if the wavelength of the generated wave is
small compared with the length of the propagation path. In a specific medium the
wavelength is given by:

233
 =v:f

Where:
= wavelength
= sound velocity in a specific medium (1400 m/s in oil)
f = frequency of mechanical vibration (acoustic wave)

Oil by itself is a perfect medium for the propagation of acoustic waves, with no
attenuation or dispersion occurring. However, in transformers, acoustic wave
propagation is heavily influenced by the complicated structure of the insulating system
(winding barriers, core, and tank walls). The amplitude is affected due to signal
attenuation and the signal shape is affected due to absorption and dispersion by the
different media in the path of the acoustic signal emitted by the PD source.

Besides the absorption and dispersion phenomena, the multitude of wave types
complicates the analysis of acoustic PD signals detected in a multi-material structure.
Two types of waves must generally be considered for the analysis of acoustic signals:

Transversal waves, attenuation dependent on wall thickness

Longitudinal waves, higher velocity than transversal waves (approximately by a
factor of two)

The velocities of sound for the two wave types for different materials found in
transformers are given in Table 3-47.

Table 3-47: Example of Sound Velocities

Material Velocity Density

[m/s] [kg/m3]
Oil 1,400 950
Pressboard 2,000 1,250
(parallel to fiber)
Pressboard 3,500 1,250
(perpendicular to fiber)
Steel, plate 3,200 7,900
(transversal wave)
Steel, plate 5,200 7,900
(longitudinal wave)

The two waves take different propagation paths from the source to the walls of the tank.
An acoustic sensor positioned at a defined location on the tank wall detects both directly
propagated waves and wall-propagated waves. As shown in Table 3-47, the two wave
types have different propagation velocities.

In order to localize the PD sources by analyzing the time difference between the electric
and acoustic PD signals, there must be a possibility to distinguish between directly

234
propagated and wall-propagated waves. This information is theoretically hidden in the
wave front of the acoustic signal that is detected at the sensor on the tank wall. Special
processing software is used to analyze these signals.

3.3.4.4.2 Acoustic Partial Discharge Localization

A minimum of three acoustic sensors (piezoelectric transducers, such as the one shown
in Figure 3-86) and at least a four-channel digital oscilloscope or signal recorder are
required to localize PD sources using the time difference between electric and acoustic
signals. An advanced detection system previously manufactured by ABB uses three
transducers arranged geometrically as shown in Figure 3-87. The three transducers are
placed close to one another in an equilateral triangle with L = 0.15 m so that the
propagation path of the acoustic waves from the PD source is nearly the same. The
system defines its own x, y, and z coordinate system. The normalized coordinates (x, y,
z) give the direction to the source and depend only on time differences in the detection
system. It is assumed that the distance between the transducers L is considerably less
than the distance of the acoustic sensors to the PD source.

Figure 3-86: R15 Piezoelectric Sensor (Courtesy of Physical Acoustics Inc.)

235
 Figure 3-87: Geometric Layout of ABB Three Transducer (TTD) Acoustic Detection System

A typical result obtained by the acoustic-system is shown in Figure 3-88. Due to the
different paths of electric and acoustic PD signals through the insulating system of the
transformer, there is no simple relationship between the amplitude of apparent charge
and the amplitude of acoustic waves. Knowledge of the transformer insulating system
and experience in the analysis of the results are needed to localize PD sources by
analyzing acoustic PD signals.

Figure 3-88: Typical TTD System Results

236
4 FAULT ANALYSIS28
4.1 GUIDANCE FOR PERFORMING FAILURE ANALYSIS
4.1.1 INTRODUCTION

Since the introduction of AC technology as the backbone of electricity supply systems

more than 100 years ago, the transformer has been an indispensable link in the chain of
power transmission from the power generation to the consumer. Its operational
reliability is a crucial factor for the dependability of supply of the entire system. The
failure of a transformer always results in an interruption in the energy supply of
consumers with considerable consequential costs. These costs affect not only the
supplier in the form of power loss and expenditure on repair work or replacement.
Consumers are also affected in terms of interruptions in work processes, production of
waste, loss of data, failure of safety equipment, possibly involving risk to life and limb,
e.g. in road traffic situations or in vital hospital functions.

In terms of the investment costs of the grid, transformers constitute a significant factor;
after all, they account for more than 30 % of the capital expenditure on a transmission
substation.

The failure statistics of the utilities and network operators present a high level of
operational dependability and reliability for power transformers. A typical example here
is provided by the failure and availability statistics from the Confederation of German
Network Operators (VDN) for the year 2004 (see Figure 4-1) [123]. For the medium-
voltage range (up to 36 kV), these statistics show a failure rate of 0.15 %; in the high-
voltage range (up to 220 kV) of 0.2 – 0.4 %, and in the extra-high-voltage range (380
kV) of 0.5 to 2 %. These figures, however, apply only to the German network; they
cannot be transferred to other national grids with different operating conditions. They
can, however, be regarded as typical for industrialized countries. In newly industrializing
and developing countries, higher figures could be expected.

28
Written by Dr.-Ing. Reinhart Baehr, Viernheim/Germany

237
 Figure 4-1 - Number of transformer failures with and without supply interruption (reference: 100
transformers in MV, HV and EHV)

In the past, highly qualified experts were available at both the manufacturers and the
utilities, which thanks to their extensive experience were able to swiftly and
knowledgeably assess failure histories, damage extent and measures for restoring
operational capability, repair or replacement of the defective units, and to take the
appropriate decisions. Now that in-house engineering capacities have been downsized
by reasons of rigorous economy measures in the course of privatization and
deregulation, this specialized competence is increasingly at risk of being lost. The
replacement of engineering knowledge and long years of experience by electronically
based expert systems (much advocated as an alternative) is not very successful, due to
the complexity of many fault scenarios, and can easily lead to incorrect decisions with
serious consequences.

Power transformers are designed according to individual specifications. The mean

number of identical units is in the range of only two. Even if the main data and design
principles are specified, there are major manufacturer-specific differences in the details.
This applies to the dimensions, the choice and quality of the materials concerned, and
the quality of the craftsmanship, as well as to the local electrical, thermal and
mechanical operating stresses. This makes it more difficult to draw up general rules,
criteria and algorithms for condition assessment and fault analysis by expert systems. In
addition, even in the case of transformers of identical design, their life history and their
individual service conditions will be different. It depends on the load, the age, the
previous stress from system disturbances, overloads, overvoltages and climatic
conditions. These influencing variables will be differently linked for each individual
transformer. Even highly specialized logic systems, e.g. fuzzy logic, suitable to evaluate
even vague data such as “fluctuating load, high switching frequency” etc., are not
suitable to produce optimum decisions for an individual fault case.

Irrespective of the methodology, the tools and the skills of the persons charged with the
failure investigation, success will depend primarily on the availability and quality of the

238
relevant data for the damage concerned. In the past, the lack of systematically collected
individual life data often impeded fast and accurate fault analysis and correct decision-
taking if a transformer failed during operation. Therefore it is advisable to maintain a
“logbook” right from the beginning for each transformer. This should contain a minimum
of characteristic design data and reference values (fingerprints) for important criteria,
describing the initial condition of the transformer. These data shall also include oil
reports, gas-in-oil analyses and PD-, DFR- and FRA- patterns before commissioning.
This logbook should be updated during the entire lifetime of the unit with operating data
and information concerning all unusual events in the system, such as short-circuits,
overvoltage events, overloads in the grid and maintenance work, oil analyses, gas
analyses, inspections and repair jobs, relocations, transports, measurements, etc.

Common definitions and a common recording system for these data using electronic IT
are recommended. The preparation of international recommendations and standards is
on the way (CIGRE, IEC) [124]. They will support comparability, and can be used to
provide improved predictions for operational reliability and residual lifetime.

4.1.2 FAILURE DEFINITION

The English language makes a distinction between “failure” and “fault”, with “failure”
denoting an incident that results in a disturbance of operation, and “fault” indicating the
faulty state of the equipment after a failure has happened. According to series IEC
60050, failure of a piece of equipment is defined as the loss of its ability to perform its
specified function. This definition is not very precise, it assumes both the total loss of
the transformer, e.g. due to a serious internal flashover, as well as the failure of a
peripheral auxiliary unit, such as a pump, which results only in temporarily interrupted or
restricted operation.

For the remarks below, “fault” is interpreted as a state that interrupts or endangers
trouble-free operation.

4.1.3 CLASSIFICATION OF FAILURES

Two different types must be differentiated here:

Type 1: gradually developing failure

Type 2: suddenly occurring failure

Type 1 is characterized by a steady rise in one or more characteristic in service

parameters beyond the limit value that guarantees reliable operation, e.g. by

continuous increase of decomposition gases dissolved in the oil

excessive rise in oil temperature
local discolorations of the tank paintwork
accelerated aging of the oil
remarkable change in the sound and/or vibration behavior

239
If the monitoring systems are signaling an incipient failure, but do not automatically
trigger a shutdown, failure analysis serves primarily to assess the hazard potential if the
transformer is kept in operation and to initiate scheduled precautionary measures. In
such cases diagnostic procedures are prioritized in the analysis. This includes using
monitoring systems for online checking, in order to gain time to make the requisite
preparations for healing the defect.

Type 2 covers all cases of unplanned interruptions in operation caused by the tripping of
a protective device, e.g. by

Buchholz relay
differential protection relay
overcurrent protection relay
overload protection
pressure-relief relay

In these cases, fault investigation and analysis serve to restore serviceability, and will
most times lead to major interventions in the defective unit, which may ultimately result
in repair and replacement. The measures to be initiated may be extensive, including not
only analysis of the failure cause but also the decision whether the necessary repair can
be performed on site and what preconditions must be met for this purpose. Likewise,
whether the transformer must be replaced immediately in order to assure reliability of
supply, and what spare units are available. In the worst case, a decision will need to be
made on whether to repair the defective transformer in the manufacturer’s facility or in
an appropriately equipped workshop, or to scrap it.

4.1.4 GENERAL INFORMATION ON MALFUNCTIONS AND FAILURES

Generally accessible information on the nature and frequency of transformer failures in
operation are unfortunately scarce and incomplete. Not every transformer owner (power
utilities, network and system operators) is prepared to publish fault histories and failure
rates for various reasons. There are no binding standardized definitions for which
events are characterized as failures. Some companies hesitate to publish failure figures
which might damage their image. Failure statistics permit conclusions to be drawn on
corporate strategies and business philosophies that should not really be disclosed to
competitors.

Generally the owners of transformers do in fact keep internal failure statistics. The
actual data and figures from different operators, however, are often not comparable.
This is due not only to the reasons outlined above, but also to the technical status, the
differing grid conditions, procurement practices, maintenance and operating
philosophies, climatic conditions and in some cases political conditions.

The consequence of breakdowns of the supply system must be evaluated very

differently. They depend on the technical status and social structure of the supply
territories concerned. The industrialized countries depend to a far greater extent on a

240
reliable power supply than do countries and societies with a comparatively low technical
status in terms of energy utilization.

4.1.5 SYSTEMATIC FAILURE ANALYSIS

The investigation of transformer malfunctions and failures is in many aspects
comparable to the detective work required for solving crimes. Both tasks necessitate the
same sequence of steps and activities as listed in Table 4-1:

Table 4-1 - Comparison between crime investigation and transformer trouble-shooting

Activity Crime Transformer failure

Collecting data and on the persons involved on the equipment involved

background information
Investigation of the circumstances of the failure-related
circumstances, operating
data
Diagnosis Autopsy of the victim Diagnostic measurements,
inspection and dismantling of
the transformer
Evaluation of all available the guilty party the cause of the failure and
information and analytical the action required
results for assessment of

For analyzing and clarifying transformer malfunctions and faults, a systematic

procedural approach is recommended, using the following steps and actions:

Collecting information on the equipment concerned

Reconstructing the failure history and the operating conditions at the time of the
malfunction/failure inception
Assessment of the status and extent of the damage on site and performing
measurements and diagnostic routines, if necessary including inspection of the
core-and-coil assembly
After clarifying the cause of the malfunction, decision on whether to keep the
transformer in operation or to perform additional investigations
Dismantling the transformer in a suitable workshop to clarify the cause of the
fault
Analyzing all data, measurements and observations for definitive fault
clarification and a decision on repair, reworking or redesign.
The first four steps are common to the two failure categories mentioned above. Further
investigations will depend on the decision to continue operation or to perform a repair
job. Possibly, the investigations might be identical in both cases. Performing special
measurements and tests may assist in clarifying the failure cause and help in estimating
the repair work required. The same diagnostic measures can, however, also be used to
assess the condition of the transformer as well as to assure its further serviceability. In

241
every case, the cause of the failure has to be clarified, in order to make the right
decisions.

The sections below deal with the individual steps involved, and their significance for
systematic trouble-shooting.
4.1.5.1 COLLECTING INFORMATION ON THE UNIT CONCERNED
Every fault analysis is based on data and information on the defective unit and its
operating environment, which - if interpreted correctly - will form the basis for all further
decisions. The more complete this information, the easier it will be to reconstruct the
failure history, to determine the fault location and extent, and to make correct decisions
on further action. The most important information and data are listed below, and their
importance for fault analysis and their impact on the decisions is described.

Data for identifying the object involved

Unambiguous identification requires the following data:

Type
Rating
Voltage
Ratio
Connection scheme
Frequency
Serial number
Year of manufacture and location
Manufacturer

Rating data
Type designation, rating definition and overload conditions will reveal the type the unit
(1-phase/3-phase) and whether it is being used as a common transformer in the grid or
as special transformer. This will enable initial conclusions to be drawn on how the unit
has been operated. Special transformers, for example, are often subjected to
intermittent stresses and high operating temperatures. Generator step-up transformers
are mostly exposed to higher thermal stresses than transmission transformers, which
often are loaded less than 50 %, depending on the operating philosophy involved.

Year of Manufacture and Location

The year of manufacture is of important for drawing conclusions about the technical
state of the art and the materials used in manufacturing the transformer. It is also an
initial rough indicator for the aging status of the insulation system. The location may also
be a significant information, since it can provide information of the specific network
configuration and the local geographical (altitude, accessibility) and climatic conditions
on site.

242
Connection
Connection scheme, winding configuration and ratio are important data for localizing the
fault in the transformer, and also for planning and conducting measurements within the
context of fault diagnostics. These data provide references to the dielectric stresses
inside the transformer.

Test reports
The results of factory tests and information on faults during manufacture and testing can
provide indications of prospective causes of faults, and therefore are of importance for
fault analysis. They are also useful for comparison with measurements obtained during
on-site failure investigation.

Maintenance reports
Reports on routine maintenance work, oil analyses, oil changes, gas-in-oil analyses,
etc. provide valuable hints to the transformer aging status and to any irregularities in
operation. Gas analyses before and after the failure inception, in particular, often permit
conclusions on the nature and cause of the fault concerned. Their interpretation is dealt
with elsewhere in this book.

Manufacturer
The name of the manufacturer can supply pointers to the type of design (core type /
shell type) and possibly to the technical competence and quality standard of the
manufacturer.

4.1.5.2 DATA AND INFORMATION AT THE TIME OF FAULT INCEPTION

Systematic failure analysis begins with collecting all available information on the failure
event, service conditions and the responses of the installed protective and monitoring
devices.

Time of failure occurrence

The time of failure occurrence is of particularly major importance, because it allows the
vital temporal correlation between the failure event and concurrent disturbances in the
network at the same time, for example, short-circuits, overvoltage events, overload and
dynamic loads, switching operations, defects in other pieces of equipment, protective
trips, triggering of surge arresters, plus environmental and weather conditions
(thunderstorms, sudden changes in air pressure, seismic disturbances).

Protective devices
Records of protective messages and trips are indispensable for failure assessment.
They can provide information on the temporal progress of the failure, on the fault’s
location, and on the size of the stresses occurring. They must contain the technical data
of the protective units, their settings and their response characteristics.

The first priority is to check the functions of the Buchholz protective relays and
overpressure protective devices (sudden-pressure relays, on-load tap-changer
protection). If the Buchholz relay is tripped, care must be taken to ascertain whether this

243
represents merely a warning to indicate that air or gas has gradually collected inside the
relay (signaling an irregularity in the transformer) or an incipient fault. In this case, the
analysis for flammability of the collected gas will decide what further action to take. A
Buchholz trip, by contrast, triggered by a sudden increase of the oil pressure inside the
transformer tank, almost always indicates an internal fault. Rare cases have, however,
been experienced in which suddenly released trapped air bubbles or winding
movements triggered by external short-circuits or seismic events have caused
protective trips without the transformer having suffered any internal damage.

Overcurrent and differential protective devices indicate faults inside the transformer
where windings or winding sections are affected and the ampere-turns equilibrium is
disturbed compared to the normal service condition. Information concerning response
times and response thresholds are useful to estimate the level of the stresses occurring.

Surge arresters
Surge arresters likewise supply important information for fault investigations. Surge
counters, if fitted, provide information on the number of overvoltage stresses exceeding
the protection level, and on pre-stress of the individual phases. Since surge arresters
exhibit only a limited capacity to extinguish lightning or switching surges, the damage of
arresters permits conclusions on the severity/duration of the stress involved or on
repeated surges.

Temperature indicators
Data concerning the load on the unit before the failure plus the measured values of the
oil and winding temperatures provide information on the thermal status in the defective
unit and the efficiency of the cooling system. This data is very useful for the
investigation of failures caused by thermal problems.

On-load tap-changers
Information on on-load tap-changer position and operations provide important hints on
the fault occurrence and on the fault localization. Fault statistics show that a substantial
proportion of all transformer failures are attributable to faults in the on-load tap-changer
or malfunctions of the changer or of the drive mechanism. Therefore the question of
whether on-load tap-changer was operated at the time of failure occurrence is of major
importance.

Fault-recorder
Abnormal events in the network are frequent causes of failures. For this reason,
information on short-circuits, overvoltage events like lightning strikes, switching
operations and overloads rank among the most important data. This information should
be as comprehensive as possible, and provide data on the time, location, duration of the
event and amplitudes of current and voltage in the phases involved. Mal-operations, like
asynchronous switching or switching in phase opposition, may cause transformer
failures. The most valuable information is provided by fault-recorder data. Careful
analysis of these recordings is able to reconstruct the chronological history of the
failure, its duration and the location/extent of damage of the unit.

244
Verbal information
Another important source of information are logbooks and manual records of the station
personnel on special events and observations. Interviewing employees can reveal
information on accompanying circumstances and on events in the past that were not
recorded in measurements or written files because their importance for the ongoing fault
was overlooked or not properly realized.

Further information relevant to faults includes:

earlier failure events

relocations and transports; if available: acceleration values
most recent work on the network or most recent maintenance work performed at
the transformer before the failure event
previous repair jobs
faults at comparable transformers from the same manufacturer

Gas-in-oil analyses results

Gas-in-oil analyses after and in comparison with corresponding analyses before fault
inception rank among the most important data available for assessing the damage and
deciding on further action. The composition and ratio of the characteristic decomposition
gases are the most important indicators for the nature and severity of an internal fault
and are thus indispensable for fault diagnostics.

In the case that the transformer has not yet failed, but there is a suspicion of an incipient
fault, the assessment of decomposition gases in the oil is actually the crucial key to all
further decisions.

(In Case Study 3 – Buchholz trip. Most probably the evaluation of the gas content would
have avoided the subsequent total damage).

Because of its paramount importance as a diagnostic tool, gas-in-oil analysis is dealt

with more comprehensively in Section 4.1.6.2.1.

Information on the oil type, ageing status and on oil-related work like refilling,
purification and degassing, drying processes, vacuum treatment, etc. are of additional
importance, e.g. when the gas-in-oil analysis indicates partial discharges inside the
transformer.

Conducting and interpreting gas analyses has for many years been a standard tool for
condition monitoring and fault analysis, and has been adequately described in
numerous publications. The recommended procedures and equipment for sampling,
laboratory analysis and evaluation of the results are specified in the relevant standards,
e.g. IEC 60567, IEC 50599 and IEEE C57.104.

245
Evaluation of all the above mentioned information will normally allow drawing reliable
conclusions on the nature of the fault and in many cases also estimate the extent and
location.

4.1.5.3 DECIDING ON CONTINUED OPERATION OR ADDITIONAL INVESTIGATIONS

A transformer malfunction is indicated either by the automatic activation of protective
devices that cause a trip of the unit or by a warning that does not lead immediately to a
breaker operation. In both cases, the question arises as to whether it is permissible to
switch on the transformer again and to continue operation or if the conditions
categorically prohibit continued operation. In order to take an appropriate and correct
decision, further detailed information is required.

The first step is to check which protection systems have responded and in what
sequence. In many cases, it is possible to conclude from the answer to this question
whether an internal fault is involved. The protection and monitoring systems of the
transformer itself are limited to the Buchholz relay, the temperature monitoring device
for the oil and for the windings, the on-load tap-changer protection (if an on-load tap-
changer is fitted), and in some cases explosion protection devices. In addition the
transformer is protected by the following external protection equipment: differential
relay, overcurrent protection relay and overvoltage protection/surge arresters designed
for direct protection of the transformer.

For an initial evaluation, the most important step is to control the Buchholz relay.
Irrespective of whether a warning or a trip has been indicated, the gas that has collected
in the Buchholz relay should be examined immediately for flammability and its
composition. The flammability can be very easily checked by flaring the gas from the
Buchholz relay. This will provide a first indication if there is any internal damage. If there
is only plain air found in the relay and if there are no other indicators for any damage to
the transformer, the unit can be approved for further operation. In the case of
transformers with open breathing systems, it is known that sudden fluctuations in air
pressure due to atmospheric disturbances, e.g. sudden drops in temperature before
thunderstorms, may lead to the release of air dissolved in the oil, thus simulating a
defect by triggering a Buchholz warning. If, for example, under comparable weather
conditions other transformers in the vicinity are known to exhibit Buchholz alarms, and
no flammable gases are detected, it can be concluded with a high degree of probability
that there is no transformer fault involved. In any case, to be on the safe side, gases
from the Buchholz relay should always be analyzed in the laboratory whenever there is
a Buchholz alarm.

In the event of a Buchholz trip, and if flammability is detected, both the Buchholz gas
and the gases dissolved in the oil must always be analyzed. The results of these gas
analyses provide unambiguous indications whether any internal damage is developing
or has already occurred, and what kind of damage it is. The evaluation and assessment
criteria are known, and laid out in the relevant standards and are also referred to in a
large number of publications, e.g. [125],[126],[127].

246
The tripping function of the Buchholz relay is released by a sudden oil flush in the
Buchholz pipe. Sudden movements of the winding due to short-circuit or seismic events
may be the cause of a trip. It is therefore essential to check whether events of this kind
have been recorded at the same time. If there is no internal electrical fault involved, the
gas analysis will not indicate any damage. The question then arises whether windings
have suffered permanent deformation that signifies a future risk of damage. To answer
this question further investigations and measurements are necessary, which are
described in the chapter on measuring and diagnostic procedures.

In any case the decision to reenergize a transformer after a trip requires careful
consideration of all available information based on profound transformer knowledge and
experience.

Even if the decision is to continue operating the transformer the condition of the unit
should be carefully documented. This includes:

instant of the failure message including date and time

instant load
measured temperature values for oil and where appropriate for windings
tap-changer position
all other measurement and diagnostic data available.

If a gradually developing fault is suspected or has been detected, but an immediate

shutdown is not absolutely necessary or for reasons of supply dependability would be
irresponsible, this data can be very useful for properly planning further action. In any
case, a transformer inspection should be performed on site, to detect any damage and
changes visible from outside, such as deformations, oil spill, discolorations, etc., which
can indicate faults and their causes.
4.1.5.4 ASSESSMENT OF THE EXTENT OF DAMAGE ON SITE
If after a fault the decision is to not re-energize the transformer, but to initiate
investigations, a distinction must be made between two different situations:

In the first case the transformer has a fault, but it is still serviceable,
though perhaps under restricted operation. Reasons for keeping it in
operation may be energy bottlenecks, lack of spares or the requisite lead
time for a transformer replacement and/or a repair job. In this case, the
owner has to estimate the risk of a total failure. This requires the need to
know the nature of the fault involved. Therefore the decision to opt for
monitored continued operation has to be preceded by an accurate check
of the serviceability of the transformer, including all the routine
measurements.

247
 There are situations where transformers regularly showed gas evolution
indicating an unambiguously developing internal fault, caused by thermal
overloading or partial discharges. In such cases it is essential to monitor
the gas generation rate by frequent oil sampling or via a gas monitor.

Likewise, following severe or frequent short-circuits, mechanical problems

may be indicated by a change in the audible noise and vibration behavior
or by alterations in short-circuit reactance. Then it is useful to conduct FRA
(Frequency Response Analysis) measurements to detect mechanical
changes and to initiate precautionary measures for a critical case. In the
past decade, modern methods have been developed for measuring and
locating partial discharges, also for use on site. They constitute an
indispensable tool for diagnosing incipient faults and tracking their
development (see also Case Study 1, chap. 4.1.8.1).

In the second case, the transformer is no longer operable by reason of an

obvious damage. Then further investigations on site will clarify whether a
repair job can be carried out on site and what measures must be initiated
to restore serviceability as soon as possible.

Irrespective of the action to be initiated, unambiguous clarification of the failure cause is

always essential in order to correct design defects in case of a repair job and to take
appropriate precautions against faults in comparable transformers.
4.1.5.5 ASSESSMENT OF EXTERNAL DAMAGE ON SITE
As a first step, the external condition must be checked for abnormal changes. Particular
attention must be paid to:

damage to bushings and arresters

cracks, deformations and discolorations of the tank
external traces of flashover
traces of fire
oil level and oil leaks
failure of auxiliary equipment like fans, pumps, cooling units
trips of protective devices, e.g. pressure-relief devices
maximum readings of the temperature monitoring instruments
damage to other units or system components

All observations should be accurately recorded and wherever possible documented by

pictures. This documentation is of crucial importance both for definitive assessment of
the fault and its cause, and for specifying the repair and remedial actions as well as for
legal and insurance-related assessments. Defective components must be kept safe, to
ensure they are available to experts or insurers, or to enable material analyses to be
performed as necessary for further clarification.

248
When entering the unit on site, all specified safety regulations must be complied with;
the transformer bay must be de-energized and visibly grounded.

4.1.6 DIAGNOSTIC MEASUREMENTS AND THEIR INTERPRETATION [128][129][130]

[131]
The second important step, essential in both fault cases, concerns diagnostic tests that
can provide indications of the internal status and operational capability of the
transformer. The most important conventional measurements are:

gas-in-oil analysis
breakdown voltage of the oil and further laboratory examinations to indicate
the condition of the oil
insulation resistances and tan of the windings
transformer ratio
winding resistances
short-circuit impedance
excitation at low voltage

If there is a suspicion of partial discharges, PD detection and localization measurements

should be made. These tests, of course, can be conducted only if the transformer still is
serviceable and able to be excited at least up to the rated voltage.

To check for winding displacements and deformations an FRA measurement is

recommended. FRA allows recording the transfer function of the windings in the
frequency range of line frequency up to approx. 2 MHz, and comparing the result with a
reference characteristic.

If there is a suspicion of increased moisture content, it may be advisable to measure the

polarization spectrum using the DFR (Dielectric Frequency Response) method or
comparable methods. The appropriate modeling tools should be used to estimate the
average moisture content in the transformer insulation. If such a model is not readily
available a comparison with reference spectra from previous measurements will at least
enable a trend to be discerned.

The results of diagnostic measurements must be compared with previous

measurements taken in the factory or on site (fingerprints). Any deviation may allow
important conclusions to be drawn regarding the internal conditions and further
serviceability of the unit, so that correct decisions can be made.

The following list of diagnostic measurements and procedures makes no claim to

completeness. It can offer merely a rough survey of the recommended procedures, the
limits of their application and their possible contribution to fault clarification. Conducting
routine measurements does not require any specialized knowledge. Frequently the
measuring equipment available on site is simple and not as sophisticated as in a well
equipped modern test field. Nevertheless most of the routine tests listed below can be
performed using standard equipment and procedures. For more information see section

249
3.1 to 3.3. The application of more complex diagnostic procedures and the interpretation
of their results should, however, be left to appropriately qualified experts.
4.1.6.1 ROUTINE MEASUREMENTS ON SITE

4.1.6.1.1 Oil analysis

The breakdown voltage of the oil, measured in accordance with the relevant standards,
e.g. IEC 60156, can, when compared to previous reference values, provide a hint to
increased moisture content and particles. In particular, soot and carbonized cellulose
particles may indicate an internal flashover. Further investigations into the oil condition
must be carried out according to the standards by an appropriately qualified laboratory
(see section 3.2.1).
4.1.6.1.2 Insulation resistance and tan
A low insulation resistance between the winding systems may indicate massive damage
to the main insulation. The same applies for the insulation resistance between windings
to ground. In conjunction with an increased tan , this may be a pointer to increased
moisture content in the insulation or carbon tracking in the insulation.

The insulation resistances between winding systems or against ground are measured
on site using a mega-ohm instrument with an integrated DC voltage source.

For measuring tan in conformity with IEC 60076-1 or for the North American region, in
accordance with IEEE Std. C57.12.90 (see section 3.2.6).

TT = test object
MO = mega-ohm meter

Figure 4-2 : Circuit diagram for measuring the insulation resistance

4.1.6.1.3 Measurement of transformer ratio

Deviations from the transformer ratio measured in the factory indicate undoubtedly a
short-circuit or a conductor break in winding sections, and thus a fault that cannot be
cured without repairing the winding concerned.

The measurements must be taken in a no-load circuit with low excitation by directly
measuring the voltages at the terminals of the primary and secondary windings. When

250
interpreting the results, the connection diagram and the vector group of the voltages
must be taken into account. In the case of complicated circuits, e.g. special
transformers with phase-shifting windings, this measurement routine should be carried
out by experienced testing personnel.

Figure 4-3: Circuit for measuring the transformer ratio using the voltmeter method

4.1.6.1.4 Measurement of winding resistances

This measurement routine should be performed phase by phase. It can also be used to
detect open circuits or short-circuits in windings. In addition the fault can be localized in
a discrete winding. Here, too, a comparison with the factory test report can be
particularly helpful.

For measuring the resistance, a constant DC source with >20 A is required. If this is not
available, a powerful battery can also be used. The voltage should be measured
(particularly in the case of small resistances of high-current windings) via separate
cables for current and voltage measurement connected directly at the terminals, to
eliminate the influence of the external connections. For converting the results to the
reference value at 75/85° C, the temperature of the unit during the measuring routine
must be checked. In most cases, the windings, by reason of their small time constant,
will have cooled down sufficiently by the time the measurements are taken, so that the
mean oil temperature displayed can be used as the reference value.

Rx = unknown winding resistance

Rd = adjustment resistor
S = switch with protective spark gap
B = DC voltage source

Figure 4-4: Circuit for measuring the winding resistance

4.1.6.1.5 Measurement of short-circuit impedance

Deviations between measured short-circuit impedances and the reference values
indicate changes in the effective number of turns, e.g. caused by short-circuits in a

251
winding or winding section. Moreover the test result may provide indications for
localizing the fault if each phase is being measured separately. Minor deviations may
indicate deformations without shorts in the windings. To perform the measurements, a
line-frequency adjustable low-voltage source must be available, which supplies the HV-
winding with the LV-winding short-circuited.

TT = test object
K = short-circuit link
CT = current transformer
VT = voltage transformer

Figure 4-5: Circuit for measuring the reactance

4.1.6.1.6 Excitation at low voltage

Excitation of the transformer by feeding the HV-winding with open LV-winding at low
voltage (a few 100 V) provides the most reliable indicator of internal short-circuits if the
current measured (no-load current) rises inappropriately in a short-circuit-like pattern.

As under 4.1.6.1.5, a line-frequency adjustable voltage source is required. The

connection is the same as for the no-load measurements, in 1- or 3-phase form,
depending on the type of transformer involved. For a comparison with the factory
measurement, the current measured can be linearly extrapolated to the rated voltage.
4.1.6.2 SPECIAL DIAGNOSTIC MEASUREMENTS

4.1.6.2.1 Gas-in-oil analysis

During operation the oil is submitted to aging. In this process, which is closely
dependent on temperature, light hydrocarbons (CNHM) and carbon oxides (CO, CO2) are
split off from the complex oil molecules. Since most internal faults go hand in hand with
a significant (mostly local) rise in temperature (in the case of electrical discharges to
well over 1000°C), this will result in a substantial change of the steady slow increase of
the decomposition gases created by the normal aging process. Specific ratios between
individual gas components are related to characteristic fault scenarios. The relevant
standards provide statements on the spectrum of decomposition gases, caused by
normal aging, and on the spectra typical for different temperature ranges and fault
categories. Gas-in-oil analyses thus constitute an excellent, empirically validated tool for
early detection and investigation of faults in a transformer. Sampling and spectroscopic
analysis of the oil in order to determine the content of harmful gases should be
performed by appropriately qualified specialists from a recognized oil laboratory in

252
conformity with the relevant standards, e.g. IEC 60567. The interpretation should be
done by experts in accordance with IEC 60599.

For rough fault classification, the following guideline applies:

Acetylene (C2H2) is an indicator for high temperatures (700 to >1000°C), e.g. hot spots
on metallic surfaces and high-current discharges (arcing).

Typical hot spots are caused by:

loose connections of current conducting conductors
circulating currents caused by stray-fields in construction components or in
parts of the tank
loops in the earthing system of the core-and-coil assembly
increased resistance of breaker or tap selector contacts
defective insulation of core lamination (core burning)
short-circuits between parallel conductors or “cold” soldering joints of cable
and bus-duct connections

While all other gas components are also created during the normal aging process,
acetylene, even in small quantities, will almost always indicate a fault. This does not
apply to transformers with on-load tap-changer without complete separation of the
expansion tanks for tank and breaker, where switching gases, which always exhibit a
high content of acetylene, may pass over into the main tank oil. The same applies in the
case of a leaking diverter switch compartment.

Carbon monoxide (CO) and carbon dioxide (CO2) are produced by thermal aging of
cellulose. Abnormally high portions of these gases indicate increased thermal
decomposition of solid insulation, either from local overheating of adjoining metallic
surfaces or as a consequence of high-current discharges burning the isolation.

Hydrogen (H2) as a dominant component indicates the presence of partial discharges.

For a more precise analysis, further higher-molecular components of the decomposition

gases (CH4, C2H6, C3H8) and their quotients must be incorporated in the assessment
procedure. For the evaluation, standardized evaluation schemes and criteria are
available, e.g. the quotient criterion as defined by Rogers, Duval’s triangle, or the IEC
criterion (see [125]).

IEC Standard 60 599 contains typical values for the gas contents and increase rates in
normally aging transformers, which may serve for purposes of comparison if no
comparative values from previous analyses are available. These evaluations permit
more accurate conclusions to be drawn on the nature of the fault concerned and the
temperature range involved.

253
Another important criterion are the increase rates for the individual gas components;
particularly when a decision has to be taken on how long a faulty transformer can safely
continue in operation until the requisite corrective action can be initiated. In such a case
the installation of a gas monitoring system for permanent on-line control of the
development of gas is useful and recommended.

There are numerous publications available dealing with gas-in-oil analysis, which are
useful for more comprehensive information, e.g. [132] [133] [134][135].

Table 4-2 - Interpretation of gas-in-oil analyses in accordance with IEC 60599

IEC Code Fault type C2H2/C2H4 CH4/H C2H4/C2H6

PD Partial-discharges ns <0.1 <<0.2

D1 Low-energy discharges >1 0.1 – 0.5 >1

D2 High-energy discharges 0.6 – 2.5 0.1 – 1 >2

T1 Thermal fault ns >1 <1

T = <300°C

T2 Thermal fault <0.1 <1 1–4

300°C < T <700°C

T3 Thermal fault T > 700 °C <0.21) >1 >4

1)
higher values of C2H2 indicate temperatures of >1000°C
ns = not significant

4.1.6.2.2 Measurement of partial discharges

In addition to gas-in-oil analysis, PD measurement has become acknowledged as the
most meaningful diagnostic procedure for detecting and locating electrical faults inside
transformers that have not yet suffered a total failure. For conducting and evaluating PD
measurements on new transformers, there are binding standards, e.g. IEC 60 076-1,
IEEE C57.12.00, IEEE C57.12.90, but these do not cover the possibilities opened up by
the modern diagnostic methods. In the past 20 years, procedures have been developed
that even under on-site conditions lead to unambiguous fault detection and in many
cases to localization of the fault concerned. Compared with PD measurement in the
(shielded) test bay, however, special precautions must be taken in order to eliminate the
influence of external disturbance factors.

PD measurements are helpful only if the transformer concerned is still operational, i.e.
can be excited from the grid or by a separate power source up to at least rated voltage.
If an appropriate high-voltage source is available, then the measurements can also be
performed with applied voltage. Well-equipped service companies meanwhile provide
adequate transportable power sources and measuring equipment.

254
The measurements permit conclusions to be drawn on the type of discharge involved,
and by means of propagation time analyses to a certain degree on the location of the
PD source (see Case Study 1, chap. 4.1.8.1).

Measurement of the electrical PD pulses can be supplemented by acoustic methods,

especially to localize the source of the discharge.

PD measurement on site requires special measuring equipment and skills, and should
always be performed by experienced specialists. The results should always be
interpreted in collaboration with transformer experts, thoroughly familiar with the internal
construction and the design details of the individual unit under test.

For more information, see section 3.3.4.

4.1.6.2.3 FRA method
Since the transformer winding system constitutes an oscillating system, comprising
ohmic resistances, reactances and capacitances, each winding has its own typical
frequency response, characterized by the quotient of the input voltage and the voltage
measured at the output. If, due to geometrical changes in the windings or by short-
circuit of winding sections, the reactance changes, then the frequency spectrum will
also change in a characteristic manner. For that reason deviations from the reference
spectrum of previous measurements are reliable indicators of winding deformations.

The measuring circuit reacts very sensitively to external disturbances and interference.
In the frequency range from approx. 1 kHz to 1.5 MHz, however, reproducible and
meaningful results can be expected, if standardized measuring equipment and
measuring conditions are used [137].

Because of the extreme sensitivity, it is not easy to interpret the results, particularly if
the deviations from the reference spectra obtained in previous measurements are small.
At present, this measuring procedure cannot supply unambiguous statements regarding
the nature, the extent and the position of the geometrical changes in the windings and
regarding the degree of risk to the transformer. If no references are on file, a
comparison with the spectra of the other phases or with measurements taken from
transformers of identical design and construction may prove helpful [136],[137].

In conjunction with further measurements, e.g. the change in reactance or in the audible
noise behavior, an experienced expert may, however, be able to draw conclusions on
the operational risk involved.

For the assessment of the type of fault, the CIGRE Study Committee A2 – Power
Transformers has listed spectra for various fault and deformation types [136].

For more detailed information on the application of FRA, see section Error! Reference
source not found..

255
 4.1.6.2.4 Measurement of polarization effects for assessing the moisture
The determination of moisture content in the oil-paper dielectric inside the transformer
constitutes a difficult problem. For measuring the water dissolved in the oil, the Karl-
Fischer method is a well recognized procedure, but the results are not suitable to draw
conclusions on the moisture content in the cellulose, since there are many parameters
involved in determining the distribution of moisture between the oil and the solid
insulation.

In recent years, various approaches have been pursued to use polarization effects for
determining the moisture content.

The recovery voltage method (RVM), in which the recovery voltage is measured during
pulsed stress of the insulation system with DC voltage, thus determining a polarization
spectrum, has proved to be ambiguous, since it does not take due account of the
individual, inhomogeneous configuration of the insulation system. Therefore the
procedure cannot be recommended. [138], [139].

Recent development of procedures for measuring the polarization current in the time
and frequency domains promises better success. The method of polarization and
depolarization current measurement DPC [140], [141] operates in the time domain,
while frequency domain spectroscopy (FDS) measures in a broad frequency domain of
several kHz. For these procedures, the CIGRE WG D1.01.14 has developed an
interpretation scheme that in conjunction with the appropriate software enables
statements to be arrived at on the aging status and the water content.

Some of the above mentioned methods are still under development and the
interpretation of the results is best left to experts.
4.1.6.3 INSPECTION OF CORE-AND-COIL ASSEMBLY ON SITE
To make a decision whether a repair job is possible on site, or perhaps even without
moving the transformer at all, opening the transformer on site and getting inside to
inspect the core-and-coil assembly may be necessary. The aim of this inspection is to
determine and document the extent of the damage. If possible, any simple repair work
required can be performed inside the tank.
4.1.6.3.1 General preconditions
Depending on the suspected location of the damage, the first step to take is to lower the
oil level before bushing domes or inspection flanges can be opened to enter the tank.
Depending on the weather conditions, a cover should be provided, so that no moisture
(rain) or foreign parts and dirt can get inside the transformer. Whenever the work is
interrupted, the openings must be closed, at least provisionally.

If no suitable enclosed space (power house, assembly hall) is available for the
inspection and if the repair work takes several days or weeks, it may be necessary to
erect a permanent tent and to ensure a slight overpressure of dry air inside the tent.

256
 4.1.6.3.2 Safety precautions
Inspection of a transformer, especially entering it, is not without potential danger for
health. For this reason in addition to the common safety regulations for high-voltage
systems (see chapter 10), some safety measures specific for transformer internal
inspection should be taken. Ladders have to be safeguarded; work platforms must be
fitted with a rail to prevent falls. In order not to get lost, all tools, lamps, measuring
instruments, cameras being used inside the transformer must be securely attached
outside or to the wrist or belt of the person who is performing the inspection or repair
work. If the inspector wears glasses, they must also be safeguarded against loss!

Before going inside the transformer, its interior must always be ventilated with fresh air,
so as to expel the oil vapor and to ensure an adequate supply of fresh air for breathing.
The person entering the transformer should wear a safety belt and when entering be
secured with a rope by at least a second person positioned at the entry point, so that
he/she can be rapidly rescued in the event of an injury or a sudden attack of nausea.

The working clothes of the inspector - an oil-proof overall is recommended, one that
also covers the shoes - has to be clean, so as not to bring any contaminants into the
transformer; all pockets must be emptied.

The inspector should be comprehensively informed about his/her task and the risks
involved, and should study drawings in order to have a clear idea of the accessibility,
the construction of the core-and-coil assembly, the wiring and cabling and significant
design elements before he/she begins the examinations.
4.1.6.3.3 Checks to be conducted
In most fault cases, the advance information and the investigations and diagnostic
routines beforehand will already have provided indicators to the nature and at times the
location of the fault concerned. Depending on the type of fault involved, different
characteristic damage patterns can accordingly be expected.

In the event of internal flashovers and partial discharges, the traces of these flashovers
or partial discharges must be investigated at the parts involved, and their progress
tracked. It is of particular interest to find out the position of the roots, and in the case of
PD defects the nature of the typical discharge patterns that may provide hints to the
direction of the discharge.

In the case of bushing faults, it is particularly the connections between winding and
bushing that need to be examined. Loose screwed connections and faulty potential
connections of electrostatic shielding devices are frequent causes of bushing faults.

In the case of short-circuit defects, it is necessary to check whether windings and bus-
ducts are deformed, shifted or broken, whether the clamping structure is damaged,
whether perhaps the yoke has been displaced. Often, a short-circuit has finally resulted
in an internal flashover, so that in addition to mechanical damage, flashover traces,
copper pearls, burned paper or pressboard and soot deposits can be found.

257
In the case of faults in and around the breaker, the breaker components located inside
the transformer and the connections to the regulating winding must be examined for
mechanical damage and shifts and traces of flashovers. The contacts of the tap selector
must be examined for burn marks and arcing roots. Screw connections and terminals
must be checked. Carbonized-oil formation on contacts, which have not been moved for
long time, must be noted.

If possible, the core earthing connections must be checked. The core and the core
clamping parts should be examined for deformations, shifts and discolorations of the
core laminations and the core insulation. They are indicators of excessive mechanical
stresses or core burning.

More generally, all irregularities, such as discolorations (pointers to overheating), loose

or detached parts, contamination, incipient rusting at the core, tank or design
components and all detectable deviations from the as-new condition, have to be
recorded.

All findings must be recorded and documented in writing and if possible illustrated by
pictures with all details of their nature and location. Flashover paths should be
documented in the relevant drawings or recorded in sketches. Defective parts must be
secured and kept safe as evidence and for material analysis to be performed later on if
necessary.
4.1.6.4 DISMANTLING THE DEFECTIVE TRANSFORMER
If the investigations on site reveal that the defective transformer has to be removed for
repairs, the fault location must be completely exposed, to identify the cause of the fault.
Dismantling of the transformer permits a more detailed examination of all windings and
the core and all parts, particularly the interior windings, which are difficult or impossible
to access during an on-site inspection.
4.1.6.4.1 Preconditions
For removing and dismantling the core-and-coil assembly, suitable premises and
equipment must be available. Generally this work plus the actual repair job will be
carried out in a transformer factory, preferably the manufacturer’s plant, where all the
requisite documents, equipment and qualified personnel are available. In order to save
transportation cost and time, however, more recently transformer manufacturers are
offering to carry out comprehensive repair jobs on site, and to arrange the requisite
conditions (see section 7.9). If some components, particularly windings, must be
replaced by new ones, these can be manufactured in the factory, to be installed on site.
The packing for transportation must be designed to prohibit the ingress of moisture
during the journey.

The following equipment must be provided:

an enclosed, dust-free room, if possible air-conditioned, with humidity monitoring

suitable, heavy-duty lifting equipment (e.g. mobile crane)
special tools for removing and re-installing windings

258
 clamping tools for clamping windings
drying and oil-treatment equipment
test equipment
suitable storage facilities for removed parts that can be used again
4.1.6.4.2 Inspection

Removal and dismantling must be performed in steps, until the original fault location has
been found.

1. Lifting the core-and-coil assembly from the tank

2. Removing the upper yoke (in the case of shell-type transformers if appropriate
the entire core)
3. Lifting the winding blocks and removing the individual windings
4.1.6.4.3 Inspection of the core-and-coil assembly after lifting out of the tank

Careful examination of the core-and-coil assembly concerns the electrical and

mechanical condition of the insulation system between the phases, between the
individual windings and the core and against the tank. Particular attention must be paid
to traces of flashover and their arcing roots. Note that discharge figures on insulation
surfaces may provide pointers to the point of origin and the direction of discharges.

Damaged and detached parts must be examined to identify the type of damage
involved, electrical and/or mechanical, and their position in relation to their original
location of installation must be documented.

The position and condition of the winding end insulation, clamping rings and clamping
elements and the base plates provide indications of excessive mechanical stresses.
The same applies for the position and deformation of bus-ducts and bus-duct supports.
The compression of the windings must be checked and - if possible - the residual
clamping force must be measured.

Generally, attention must be paid to any discolorations, contamination, soot and sludge
deposits, plus rusting on iron components like core laminations, core clamping parts,
tank and conservator tank:

Particles – metallic, but also non-metallic – may act as triggers for

discharges that have initiated defects.
Sludge deposits are an indicator of advanced oil aging due to high thermal
continuous stress or poor oil quality
Rusting indicates free water in the oil. In the case of transformers with a free
breathing system, that are exposed to severe fluctuations in stress and
temperature, for example in furnace transformers, free water was
occasionally discovered at the bottom of the tank and rust on core
laminations and iron components.

259
Since the on-load-tap changer is frequently involved in the fault history, it has to be
carefully examined for any damage. Particular attention must be paid in this context to
electric marks at the contacts and to flashover traces and soot or dirt deposits at the
insulating components of the tap-changer and the tap selector.
4.1.6.4.4 Inspection of the windings
All windings must be examined for:

mechanical deformations of the entire winding or individual winding sections;

especially the type and position of the deformation must be documented
electrical or mechanical damage to the winding insulation
traces of discharges on adjoining insulation barriers
thermal damage to conductor insulation
flashover marks between windings, coils, part-windings
dirt deposits due to the fault or during operation
shifting or deformation of the winding ends
deformed, shifted or missing spacers and strips
spiral displacement of windings

The nature and extent of the damage will often allow conclusions to be drawn on the
cause of the fault, since typical fault patterns can be assigned to various fault
categories.
4.1.6.5 TYPICAL FAULT PATTERNS OF WINDINGS

4.1.6.5.1 Short-circuit faults

Mechanical damage to windings is the result of short-circuits or repeated high shock
loads. The typical types of mechanical damage are axial tilting (see Figure 4-6 ), radial
buckling (see Figure 4-7) and spiral shifting of the conductors (typical multi-layer and
spiral windings).

normal tilted

Figure 4-6: Typical deformation of the conductors of a coil winding under axial overload (tilting)

260
 Figure 4-7: Radial deformation of an interior coil winding (buckling)

In the case of external short-circuits, the damage may be limited to deformations without
any consequential electrical damage. These faults do not inevitably result in a failure of
the transformer. Without removing the windings they can be detected only by diagnostic
measurements (reactance, FRA).

Internal short-circuits always involve high-current discharges that in addition to

mechanical damage lead to destruction and carbonization of insulation sections.
4.1.6.5.2 Electrical flashover
Electrical flashovers between windings or coils in most cases can be traced back to
high-frequency overvoltages (lightning impulse voltages, resonances). The typical fault
pattern often shows only slight damage to the insulation with meager soot formation and
carbonization of the surrounding insulation sections. The electric marks on the
conductors involved are often merely the size of a pinhead (see Figure 4-19, chap.
4.1.8.2).

AC voltage defects in windings, by contrast, are characterized by severe current marks,

even extending to melted conductors, and by extensive destruction and burns of the
insulation and heavy soot deposits (see Figure 4-15, chap. 4.1.8.1).

For the interpretation of the cause of the fault, it is important to document the precise
position of the fault location in the winding, so that the anticipated local stresses can be
calculated for fault analysis.

261
Flashoversover extensive insulation surfaces outside the windings, e.g. from the
terminal of the bushing along the insulated HV bus-duct and parts of the main insulation
against ground may indicate switching overvoltage stresses or static charge
phenomena (static electrification) [142].

Flashover paths and discharge figures on insulation barriers of the insulation between
windings allow considerations on the starting point and the direction in which the
discharge has propagated (Figure 4-8).

Figure 4-8: Discharge figures on the surface of paper insulation

Discharge channels below the surface in the insulation material (wormholes) indicate
moisture in the insulation, which has initiated partial discharges and finally resulted in a
complete breakdown. Sometimes X-wax can be found in enclosed pockets in the solid
insulation, e.g. between paper layers of extended insulating wraps of bus-ducts. This
decomposition product of the oil is produced in the presence of moisture by long-time
partial discharges of low strength. Due to the increase in tan , X-wax formation may
lead to thermal instability of the compact insulation. Such kind of damage has quite
often been observed in aged transformers of old design (see Figure 4-9). But also in
modern pressboard, pockets can be created in the insulation structure if it is improperly
dried. The presence of moisture in these pockets can lead to the generation of partial
discharges.

262
 Figure 4-9: X-wax in a paper insulation

4.1.6.5.3 Thermal faults

If there are indications of thermal overloading as evidenced by dark coloring of the
conductor insulation, then the individual insulation layers must be examined to check
whether the thermal degradation of the insulation has proceeded from inside out or in
the opposite direction. In accordance with the temperature gradient in the transformer,
the paper wrapping close to the conductor material will normally suffer higher thermal
stress, and therefore will be more aged than the outside layers. In the case of insulated
bus-ducts located close to hot points either of the core or of uninsulated metal
components, the outer layers may also be discolored and more severely aged than their
inner counterparts.

Aged paper exhibits a dark discoloration and becomes brittle. Even without more
sophisticated methods in the laboratory, a simple bending test on a strip of paper or
pressboard can indicate reduced mechanical strength and by this confirm thermal
degradation of the cellulose.
4.1.6.6 INSPECTION OF THE CORE AND THE TANK
Concerning the core the following checks are recommended:

migration of core laminations out of their original position

damage/deformation of core laminations
spreading of joints at the corners of the core
discoloration of the lamination indicating hot spots
burnt spots and melting points of the lamination
residual core clamping force of bandages or bolts
position and deformation of the tie rods between upper and lower yoke
discoloration/carbonization of the core insulation
potential connections of the individual core packages and to ground
position of the core and the mounting pads in the tank

263
The tank must be examined for deformations, cracks in welding seams and
discoloration of the paint. Inside the tank, a careful inspection must be made for
flashover and current marks caused by circulating currents in the tank. The mounting
and where appropriate the insulation of stray-field shielding sections at the tank walls
must be checked. Parts that come from the core-and-coil assembly but are lying on the
bottom of the tank must be identified and their position documented.

4.1.7 FINAL ASSESSMENT OF THE FAILURE AND THE FAULT

The compilation of all information, measurements and visual observations, plus, if
necessary, special material analyses, constitutes the foundation for a thorough
reconstruction of the failure history and the definitive analysis of the reason of the fault.
Based on the result of the failure analysis a qualified decision on the action required for
restoring operational capabilities of the system can be drawn. The quality and accuracy
of the analysis are crucial to the time and money expended on the requisite remedial
measures.

When the final, definitive analysis is carried out, it is absolutely essential to check
whether failure risks also apply for transformers of similar or identical design and
construction. If so, the data from those transformers must be analyzed for signs of
incipient faults.

All information, all examinations performed and action initiated, plus the conclusions
shall be documented in a final report.

Chap. 4.1.8 provides three examples of failure analysis describing the procedures
involved, the diagnostic methods utilized, and the final assessment for the cause of the
fault, plus the decisions taken on remedial measures or repair work. This description of
examples is restricted to merely the sequence of the essential steps, the findings and
measurements obtained, and the conclusions drawn from them.

264
 4.1.8 CASE STUDIES

4.1.8.1 CASE 1: EXAMINATION OF A TRANSFORMER AFFECTED BY PARTIAL DISCHARGES

Technical Data of the Transformer

Type: Generator Step Up (GSU) Transformer
Rating: 340 MVA
Rated voltage: 236 ± 11.5 % / 21 kV
Frequency: 50 Hz
Connection: YN d5
Type of cooling: OFWF
Year of manufacture: 1977

History
The transformer went into operation in 1977. Following the explosion of a wall bushing
in the transformer box on the power plant site in 1991, the transformer was subjected to
check measurements. This included a gas-in-oil analysis but it did not reveal any
damage whatsoever in the unit.

At the end of 1996, the unit was removed, and installed at a different location. In order
to transport by rail, the oil level in the tank was lowered by approximately 320 mm. The
removed oil was treated in a separate tank, and put back in the transformer via the
expansion tank after the installation. The unit was put back in service in the middle of
January 1997.

In April 1997, the transformer was again returned to its original location and re-installed.
The oil removed for before transporting the transformer was appropriately treated and
pumped back into the transformer via the expansion tank. The transformer was then put
back into operation.

Immediately after being restarted, an oil sample was taken in order to check the
treatment. The DGA revealed a significant increase in content of H2 and CH4. These
gases continued rising steadily to approximately 800 ppm and 90 ppm respectively by
April 1998.The increase was not regarded as alarming, and the cause was assumed to
be due to incorrect refilling of the transformer with the residual oil after relocation. In
order to eliminate all doubts, a decision was made to degas the oil on site using a
repeated degassing procedure. Six months after this treatment, however, the above-
mentioned gas components increased again to 835 ppm and 70 ppm respectively. No
significant increase in concentration was observed for the other gas components. Table
4-3 shows a history of gas concentrations in the transformer during this period.

265
 Table 4-3 - Listing of the H2 and CH4 values

Date H2 [ppm] CH4 [ppm] Comment

29.04.93 25 5
26.09.95 4
11.04.97 >1000 Before 2nd relocation
13.05.97 338 49
20.06.97 344 49 After restart
29.07.97 536 57
30.09.97 372 54
27.11.97 762 73
15.12.97 662 63
21.01.98 612 73
26.02.98 732 78
23.03.98 675 70
25.04.98 793 88
05.05.98 52 7 After degassing
02.06.98 316 19
01.07.98 466 28
31.07.98 579 37
14.09.98 701 47
12.10.98 770 50
03.11.98 835 70

Diagnostic Measures on Site

The interpretation of the analyses indicated continuous partial discharges and steadily
progressing damage to the insulation system. To clarify the cause, diagnostic
measurements were agreed upon and performed on site.

The investigations were designed to clarify the following questions:

What is the cause of the continuous gas production?

What are the risks involved in continuing to operate the transformer?
Can the transformer be refurbished on site or must it be sent back to the factory
for repair?

The measurements on site comprised

PD measurements using the PRPDA procedure for characterizing and localizing

partial discharges. For the PD measurements, the transformer was disconnected
from the network, and excited on the LV side by the generator.
Measurement of the polarization effects of the oil/paper insulation using the
recovery voltage method (RVM), in order to assess the general condition of the
insulation system.
Measuring the frequency-dependent impedance (FRA method) for checking
mechanical/geometrical changes in the windings.

266
Results of Diagnostic Measurements

PD measurements:
PD measurements were performed as multi-point measurements at all 6 terminals and
at the neutral. The signals showed similar PD activity at rated voltage with apparent
charges of >10,000 pC and >5,000 pC respectively in the HV windings of Phases V and
W. The inception voltage was 34% and 52% of rated voltage respectively, whereas
Phase U was initially PD-free up to rated voltage.

A comparison of the time-resolved signals with those of the calibration showed that the
PD sources were located not in the vicinity of the bushings, but inside the transformer
insulation system in the exit area of the windings. The patterns recorded (see Figure
4-10 and Figure 4-11) indicated discharges without metallic electrodes in the solid
insulation (bubbles, shrink holes, moisture). From the high amplitude of the apparent
charge at rated voltage and the large number of PD pulses (12 to 13 pulses per period
of the sinusoidal voltage), it was concluded that several different PD sources had to be
assumed inside the insulation system. The dependence of the PD inception voltage on
the setting of the on-load tap-changer position indicated defects close to the exit of the
HV windings.

After approximately half an hour during the first measurement routine at 52% of rated
voltage, an asymmetrical accumulation of PD signals synchronous with the supply
voltage was observed in Phase U as well. The amplitude of the apparent charge, at a
maximum of 500 pC, was significantly smaller than in Phases V and W, and the number
of pulses, too, was merely 1 to 2 pulses per minute. It was hypothesized that this
second PD source was caused by a conductive particle in the vicinity of Phase U, but
would not endanger safe operation of the transformer.

These results of PD measurements corresponded very closely with the results of the
gas-in-oil analyses.

Acoustic signals could not be measured, which in turn confirmed PD sources inside the
insulation system. Therefore it was impossible to locate the PD sources by acoustic
means.

The size of the apparent charges measured, and the high number of PD pulses per
period in Phases V and W led to the conclusion that the insulation system had already
been damaged, and that there was a real risk of its complete destruction, including an
internal catastrophic breakdown, if the transformer continued operation.

From the result of these measurements it was hypothesized that the cause for the PD
activity following two relocations of the transformer, was mistakes during the oil filling
procedure, since during the preceding 19 years of operation there had been no signs of
irregularities.

267
 Figure 4-10: PD measurement, PD activity at Phase V HV winding
PD patterns and conventional oscillographic diagram

268
Figure 4-11: PD measurement, PD activity at Phase W, HV winding
PD patterns and conventional oscillographic diagram

269
Measurement of the Polarization Effect:
Polarization effects in the oil/paper insulation and the aging status of the insulation were
checked by measuring the recovery voltage (RVM).

In general the RVM method does not allow a quantitative statement on the residual
moisture content in the insulation system. Since no reference measurement was
available either, the result was qualitatively compared with measurements taken on
transformers of similar construction.
In accordance with the results, the state of the insulation was assessed as good
considering its age (see Figure 4-12).

Figure 4-12: Results of the RVM measurements

Analysis of the Transfer Function (Frequency Response of the Impedance):

The results of these measurements are presented in Figure 4-13. They show very good
congruence in the superimposition of the transfer functions for all three phases. Phases
U and W are almost congruent in the range up to 200 kHz. The slight variation of Phase
V in this frequency range is caused by the difference in capacitive coupling of the
middle phase. In the range > 200 kHz, the impedance characteristics differ more
substantially, since the influence of the different capacitive coupling becomes more
significant. The FRA measurements did not give any indications of mechanical changes
in the windings.

270
Figure 4-13: Frequency-dependent impedance of Phase W and superposition of the impedances of
Phases U and V

Decision:
Based on the results of the PD measurements, an urgent recommendation was given to
take the unit out of operation for further investigations in the factory, in order to finally
detect and eliminate the PD sources.

Further Investigations in the Factory

In the repair shop the transformer was prepared for repeating all routine tests as
specified in DIN VDE 0532, with the residual oil being topped up under vacuum. All
tests at 75 % of the rated test levels were passed successfully. The PD measurements,
however, confirmed the results determined on site. The inception voltages were
somewhat higher, and the PD levels a little lower.

In view of these results, the decision was taken to open the transformer and to find the
PD sources by gradually dismantling the core-and-coil assembly.

Findings:
After the core-and-coil assembly had been lifted out of the tank, considerable quantities
of welding cinder and paint residues were found on all horizontal surfaces, particularly in
small crevices. The paint particles, some of which had an area of up to 10 cm² (Figure
4-14), were not identical to the paint on the inside tank wall. The pipes supplied for the
cooling system also exhibited a different inside coating.

271
 Figure 4-14: Cinder and paint particles

As the windings were dismantled still further, after the upper yoke had been removed,
considerable quantities of cinder and paint particles were also found on the yoke collars
and inside the windings (Figure 4-15).

Figure 4-15: Particles inside the HV winding, Phase U

On the upper static shielding rings of the HV windings of Phases U and W, and the yoke
collars above them, faint traces of discharges were found (Figure 4-16).

272
Figure 4-16: Discharge trace (left) and cinder particles (right) on the static shielding ring of the HV
winding, Phase U

Small traces of discharges were also found on the static shielding ring and the
associated yoke collar of the HV winding, Phase V.

These findings confirmed the results of the diagnostic examinations, particularly of the
extensive PD measurements and their interpretation. In general the windings were in an
excellent condition, and did not exhibit any deformations whatsoever, which confirms
the result of the FRA measurements. The insulation system showed only some signs of
aging, despite the long period of operation.

Cause of the Fault

The rise in the key gases H2 and CH4 was caused by partial discharges, which were
initiated by the paint and cinder particles that had washed into the windings. The
conclusion to be drawn from the data, the results of the diagnostic measurements on
site and in the repair shop, and the findings obtained from removing and dismantling the
core-and-coil assembly, is that these particles must have entered the transformer during
the course of its double relocation. It is known that modifications were carried out to
adapt the cooling system at the receiver station. These included welding and painting of
some components. This hypothesis is supported by the fact that the first increase in the
content of decomposition gas was observed immediately after the transformer had been
relocated.

Due to the directed oil flow of the OD cooling system, the particles were flushed directly
into the windings, where they were deposited on the horizontal surfaces. At the most
electrically stressed points of the HV windings in the vicinity of the upper static shielding
rings, the conductive cinder and paint particles led to local field disturbances, resulting
in weak but continuous partial discharges at rated voltage, which after only a relatively
short time damaged the solid insulation.

Based on these findings, despite the generally good aging condition, the decision was
taken to replace the windings and the main insulation, since it was not possible to

273
ensure that the foreign particles could be completely removed by means of cleaning
procedures.

After the transformer had been repaired and started up again, the diagnostic
measurements were repeated on site. No measurable internal partial discharges could
be detected. Consequently, the transformer was put back in service.

Final Remarks
This case study demonstrates in exemplary form the systematized, step-by-step
approach for investigating an incipient fault using modern diagnostic methods, here
primarily PD measurement on site.

274
4.1.8.2 CASE 2: ANALYSIS OF A FAILURE CAUSED BY OVERVOLTAGES AT NO-LOAD SWITCHING OPERATION

Technical Data of the Transformer

Type: 3-phase substation transformer
Rating: 40 MVA
Rated voltage: 115 ± 13 x 1.945 / 21 kV
Connection symbol: YN d5
Frequency: 50 Hz
Year of manufacture: 1976

History and Fault Progress

Shortly after the transformer went into operation, a flashover occurred across the
protective spark gap of the HV bushing of Phase V against ground during no-load
switching operation to the 110-kV network.

Routine measurements involving insulation resistance (mega-ohm test), winding

resistances and transformation ratio (TTR) were performed. No deviations were
detected from the factory results and no faults could be detected. Following these
measurements, the transformer was energized via a generator. The on-load tap-
changer was moved through all its positions at 50 % and 100 % of rated voltage. Finally
the unit was repeatedly switched on at no-load and loaded to full load. No further
irregularities occurred.

In order to clarify the fundamental cause of the external flashover, switching tests were
conducted about 2 months later. These tests resulted in another breakdown, with a trip
of the Buchholz relay. When the transformer was subsequently energized using a
generator, the Buchholz relay was again tripped, at just 45 % of rated voltage. After that
the transformer was sent to the factory for examination.

Observations at the On Site Switching Tests

During the switching tests, it was observed that closing of the three poles of the 110-kV
circuit-breaker was randomly out of synchronism by up to 9.5 msec. The peak
overvoltage observed in a total of 22 tests varied substantially with a maximum of 3.9
p.u. which corresponds to a peak voltage of 600 kV.

Findings:
When the windings of Phase V were dismantled, a flashover was found half-way up
between the two layers of the LV winding. The LV winding is designed as a 2-layer
spiral winding with adjacent oil channels and a pressboard barrier between the two
layers.

Figure 4-17 shows the damage to the windings. The severe burns in the copper
conductors of both layers, and the severe soot deposits, are clearly related to the AC

275
voltage flashover during the restart at 45 % of rated voltage following the preceding
switching voltage discharge at the switching tests.

Figure 4-17: Root of the AC voltage flashover on the inner layer of the LV winding of Phase V

After the repair work on the LV winding of Phase V, all voltage tests were repeated in
order to check the dielectric strength of the insulation. During the surge voltage test of
the HV side, another flashover was observed in the LV winding of Phase U. In this case
as well, the flashover marks were half-way up the winding between the two layers.

Figure 4-18 and Figure 4-19 show a typical surge voltage defect with roots of just the
size of a pinhead on the conductors, slight traces of discharge around the discharge
channel of the pressboard barrier, and minimal traces of burns.

Figure 4-18: Root of the surge voltage defect on the inner layer of the LV winding of Phase U

276
 Figure 4-19: Surge voltage breakdown of the layer insulation of the LV winding of Phase U

Analysis of the Cause of the Failure:

The flashover in the low-voltage winding of Phase V during the switching operations is
attributable to impermissibly high switching overvoltages when the transformer was
connected in no-load state. The synchronism error of the circuit-breaker was
responsible for the extreme level of the overvoltage. Whether the winding had already
been damaged beforehand by the flashover of the protective spark gap could not be
decided. The investigations performed revealed no indications of any damage. The
stationary rated voltage stress of 10.5 kV at the fault location, however, was so small
that even after a perforation by a discharge the layer insulation was comfortably able to
withstand this stress.

In the case of non-synchronized switching of the circuit breaker, the 3-limb core is
magnetized up to saturation in single-phase mode. If the neutral of the HV winding is
not grounded, high overvoltages may be induced. Only after the third breaker pole has
closed will the associated winding be returned to line potential.

The size of the induced voltage is governed by the remanence condition of the core and
the switching instant referenced to the line voltage. If high remanence coincides with the
zero transition of a phase voltage, and in addition there is a high synchronism error,
then as a consequence of non-linearity and magnetic transient phenomena, high
overvoltages must be expected. This is what was confirmed by the switching tests.

Modern cores made of cold-rolled grain-oriented steel sheets exhibit a very steep
magnetization characteristic, with a narrow hysteresis loop. The remanent
magnetization can match 80% of the rated induction depending on the instant of
switching off the voltage.

In the case under discussion, the arcing of the protective spark gap (set to 750 mm) of
the HV bushing of Phase V indicates switching overvoltages of more than 500 kV. For
rod-rod spark gaps with 750 mm spacing, the 50% breakdown switching voltage at
positive polarity is more than 500 kV and about 600 kV at negative polarity. These

277
figures may even be substantially exceeded due to the high scatter for long-wave
surges.

The test level of the transformer was 550/630 kV for lightning impulse voltages and 450
kV for switching voltages. It must be assumed that the stress encountered during the
flashover of the protective spark gap had been substantially higher than the protection
level, possibly above its test level.

Due to the exterior flashover, the HV winding must have been stressed with a chopped
wave, which by overshooting may have reached a level up to 140 % of the breakdown
level of the spark gap. The HV winding withstood this stress. However due to the
voltage transmitted to the LV side, layer insulation of the LV winding of Phase V
suffered a flashover. Depending on the relation of the winding capacitance and
inductance, there is a transient up-swing of the outer layer versus the inner layer. As a
result, typically the maximum stress occurs between the layers halfway up the winding.
The location of the fault in both of the defective LV windings involved confirms the
theoretical considerations.

The defect in Phase U follows the same logic; however, it cannot be unambiguously
decided whether the flashover occurred as a consequential fault of the defect in Phase
V prior the subsequent circuit breaker switching tests.

Actions
After the defective windings had been repaired the transformer was reconnected to the
system. As a result of the fault analysis, synchronism of the circuit-breaker was ensured
by appropriate means. Since then, no troubles have been observed due to overvoltages
caused by no-load switching operations.

Remark
This case study shows that switching under no-load conditions is of major importance.

By using core material with an extremely steep magnetization characteristic and

accordingly high remanence voltage, and due to the step-lap lamination of the core (low
shear of the hysteresis loop), high overvoltages which may substantially exceed the test
voltage level can occur under unfavorable switching conditions. The most important
parameters in this context are the switching instant related to the voltage half-wave and
synchronism errors of the circuit-breaker. For this reason, measures to ensure
controlled switching of the individual breaker poles are highly recommended.

278
4.1.8.3 CASE 3: FAULT ANALYSIS ON A GENERATOR STEP-UP T RANSFORMER FOLLOWING AN INTERNAL
FLASHOVER

Technical Data of the Transformer

Type: generator step-up transformer
Rating: 200/100/100 MVA
Rated voltage: 353.63 ± 2x8.63/13.8/13.8 kV
Connection symbol: YN d1 d1
Frequency: 60 Hz
Year of manufacture: 1991

History and Fault Progress

As the name plate data indicate, the unit involved is a 3-winding transformer feeding the
power of 2 gas turbines into the grid. The HV side is connected to an overhead
transmission line network via a 345-kV SF6 substation.

The 3-winding transformer has two separate LV windings, constructed as 2-layer spiral
windings and arranged axially one above the other. The common HV winding also is
axially subdivided and connected in parallel with the HV entrance in the middle. It was
designed as an interleaved coil winding. The conductors of the outermost 4 coils at both
ends of the HV winding are subdivided in such a way as to minimize additional eddy
current losses by the radial component of the stray field. The regulating winding is
likewise interleaved, and divided into two halves connected in parallel.

Approximately 6 months after startup of operation, the transformer was shut down
following a ground fault. No cause for the fault was found. The transformer was put back
into operation after routine measurements had been performed.

Seven months later, during normal operation, the differential protection relay of Phase B
was activated by the overcurrent function, but without an instantaneous trip of the unit.
To clarify the cause of the fault, the following measurements were performed:

winding resistance on the LV side; the results were not significant, since the
generator bus-ducts were included in the measuring circuit
insulation resistance of the windings between each other and against ground
voltage test with 80 kV DC over 10 minutes

These measurements did not indicate an internal fault.

Examination of the oil, and specially a gas-in-oil analysis, was not performed.

The transformer was reenergized under no-load conditions, i.e. with the HV side
disconnected from the grid. Immediately after being switched on, the unit was
instantaneously tripped, triggered by:

279
 differential protection of Phases B and C
overcurrent protection of Phase C
Buchholz relay
tank overpressure protection

After 6 half-waves, the unit was switched off by the HV circuit-breaker. Immediately
afterwards, a fire alarm was initiated in the transformer bay.

Findings
The data from the fault recorder confirmed an internal flashover to ground with a high
fault current and a collapse of the voltage in Phase B. A first inspection on site revealed
that the tank had split open at the cover flange, and was bulged at the HV side. A
considerable amount of oil spilled and was ignited by the internal arc. The fire in the
transformer bay was successfully extinguished within a short time by the sprinkler
system and the fire brigade.

The explosion-like incident had sheared off the foundation bolts, and shifted the
transformer approx. 50 mm sideways. A provisional inspection of the interior of the
transformer revealed severe damage to the regulating winding of Phase B, the
associated HV bus-duct and the bus-duct framework.

For further investigations and clarification of the fault cause, the transformer was
transported to a repair facility, where it was completely dismantled. The following
damages were observed:

The windings of Phases A and C were undamaged with only slight deformations
at the HV connections, caused by the collapse of the framework of the regulating
bus-ducts
In Phase B, the axially split regulating winding had collapsed completely; the
inner windings were radially buckled, and the outer ones were stretched. This is
a typical fault pattern for interleaved windings, with the current flowing in the
opposite directions through radial adjacent turns (Figure 4-20)

280
 Figure 4-20: Deformations of the regulating winding, Phase B (lower part)

The main insulation between the regulating and the HV windings was bulged
from inside and cracked longitudinally, due to a flashover starting from the
bottom end of the HV winding towards the HV line lead entry (Figure 4-21).
The last 4 coils at the bottom end of the interleaved HV winding experienced
severe mechanical distortion; several turns had burnt open. Like in the regulating
winding, adjacent conductors had shifted upwards and downwards out of their
original position in the coil and penetrated the axial adjoining coil. This is again a
typical deformation observed after a short-circuit in interleaved coil windings. The
transition section between the two radially disposed strands and the two axially
disposed strands in one of the coils was never found, suggesting that it was
consumed by the arcing fault (Figure 4-22).
The whole area in which the flashover to ground had occurred was severely
sooted and covered with charred pieces of insulation.

281
 Figure 4-21: Damage of the main insulation between the HV and regulating winding of Phase B

Figure 4-22: Deformed end coil

Analysis of the Failure Cause

The initial warning of the differential relay when triggered by the overcurrent function
was in itself sufficient to justify the conclusion that a short-circuit had occurred in the
entire Phase B winding or part of the winding. The high effective reactance of the sound
part of the HV winding had limited the fault current at the HV terminal to a level below
the activation level of the differential relay. In accordance with the ratio of 1/11 between
the number of turns in the short-circuited coils and the remaining sound coils, the fault
current was able to trigger an overcurrent warning, but it was far below from matching
the rated short-circuit value.

282
At any rate a pre-damage of the Phase B HV winding must be assumed. This initiated
the total flashover when the unit was switched on again.

The failure evidence strongly indicates that the most probable scenario for initiation of
the fault was a strand-to-strand insulation breakdown at the transition from two radially
disposed strands to two axially disposed strands in one of the discs toward the bottom
of the B-phase. The transition point constitutes a discontinuity with reference to fast
transient phenomena, which may lead to considerable increase in voltage due to
reflections. The risk of high overvoltage stress is particularly relevant in the case of
switching operations with SF6 breakers and in SF6 substations.

There is no doubt that at the no-load switching operation after the poor investigation (no
DGA samples taken) of the previous overcurrent warning, the previous turn-to-turn fault
was re-established and quickly spread to involve more discs and finally ended up in a
full line-to-ground fault. Either the inrush current or switching overvoltage or both of
them caused the catastrophic flashover along the HV winding between the HV entry and
the end of the winding, thus resulting in the serious damage caused by a single phase
ground fault.

Analytic network studies for determining the switching overvoltages to be anticipated on

the system, plus investigations of the transformer’s oscillatory characteristics and its
natural frequencies, did not produce any unambiguous results that would explain the
cause of the fault beyond any doubt.

The part-winding defect was likely not due to incorrect execution of the connecting point
between the end coils and the normal coils of the HV winding. This is because the unit
had been in operation for one year before the fault occurred. Moreover, an examination
of the connection points at the other windings did not indicate any incorrect
workmanship. The design and dimensioning of the conductor insulation was checked
by an experienced independent expert, who attested to high safety margins.

Remark
This case demonstrates how important gas analysis can be for assessing fault
situations. When the first differential protection trip was investigated, the person
responsible for the investigation failed to perform a gas analysis, which would definitely
have resulted in strong indications of the developing fault. Additional ratio and
impedance measurement and perhaps an FRA examination would have provided
unambiguous indications of an internal fault, so that the total loss of the transformer and
the fire on site could have been avoided.

283
5 ONLINE DIAGNOSTIC MONITORS FOR TRANSFORMERS
AND KEY ACCESSORIES
Power transformers and their accessories at critical nodes in electricity networks and
industries are subjected to heavy stresses during their lifetime. Load peaks –
predictable as well as unexpected – generate high temperatures which shorten
component lifetime. In the worst case, sudden failure can occur, causing havoc in the
network or losses of production. It is because of this risk, and the penalties that can
attach to it, that utilities give such a high priority to controlling and monitoring the status
and condition of their transformers [143]. It has also become clear that periodic
sampling of key condition indicators is ineffective at capturing the onset of some serious
problems in transformers. This is where the application of online monitors comes in. By
constantly measuring and accessing the appropriate parameters, these devices allow
asset managers to intervene before failure, malfunctioning or reduction of lifetime can
occur. Increasingly, for many utilities and industries, the watchword is ‘early detection of
failure conditions.’ In this section, we provide a discussion of some of the key
parameters that are useful to monitor on transformers and other key accessories.

5.1 POWER TRANSFORMER (TANK & CORE)

The key parameters for power transformer online monitoring (Table 5-1) include gases-
in-oil, moisture-in-oil, oil/ambient temperature, load current, winding hotspot
temperature (calculated from oil/ambient temperature, load current profile, and design
data), partial discharge, and motor current of cooling pumps/fans [144].

Table 5-1 : Power Transformer Monitoring Needs [145]

Natural Outcome if Left Signals Used for
Natural Aging Process Factors that Accelerate Aging
Unchecked Online Monitoring
Paper cellulose decomposes into Water ingress from outside;
CO, CO2, H2O, acids, and glucose internal water from paper, heat, Continuous degradation of Water (H2O) in oil and
(which further breaks down into and oxygen; and continuing paper insulation component. gases in oil (CO, CO2 ).
furans). presence of acids.
Localized cumulative Hydrogen (H2), partial
Partial discharge (corona)
insulation damage and discharge RF currents,
(<300 ºC).
eventual failure. and acoustic detection.
High localized stress on oil,
Hydrogen (H2), methane
Thermal fault (>300 ºC paper and bus work leading
(CH4), ethane (C2H6),
and <700 ºC). to damaged components, and
and ethylene (C2H4).
Oil decomposes into various gases. functional failure.
Rapid decomposition of oil
and paper into gases and Hydrogen (H2) and
Arcing fault (>700 ºC).
explosion likely if not stopped acetylene (C2H2).
quickly.
Water (H2O) dissolved in
Accumulated water dissolved in Oil will saturate and become oil and low oil
oil. free when due point reached.
temperature range.
Excessive heat in core due to Very rapid breakdown of Core temperature
elevated load currents. insulation papers in core monitoring.
Electrical current flows. Core distortion, paper Frequency response
Short circuit fault currents. damage, and loss of clamping analysis (FRA). This is
pressure. rarely available online.

284
It should be noted that although online monitoring of partial discharges (PD) on power
transformers is offered by a handful of vendors, it is still an evolving technology. The
difficulties include the ability to eliminate interferences and consistent interpretation of
the results. Due to the relative high costs PD monitoring is usually required only on
“problem transformers”.

Cooling control systems that integrate current and temperature monitoring may be a low
cost solution for power transformer retrofit. Hotspot temperature, remaining insulation
life and dynamic ratings can all be estimated from the current, temperature and
transformer design data [146,147,148].

5.2 LOAD TAP CHANGER

The simple means of monitoring the condition of load tap changers is to monitor the
temperature difference between the LTC and the transformer main tank. There are,
however, advanced monitoring parameters, including drive motor torque, drive motor
current (contact wear calculations), and vibration acoustic pattern that allow more
refined diagnosis of LTC problems to predict maintenance/overhaul needs based on
actual condition and not on the number of switching operations.

Table 5-2 : Load Tap Changer Monitoring Needs [145]

Natural Outcome if Left Signals Used for Online
Natural Aging Process Factors that Accelerate Aging
Unchecked Monitoring
Drive motor torque, drive
Drive mechanism sticky or erratic High wear and eventual drive
motor current, and vibration
in operation. mechanism failure.
acoustic pattern.
No secondary voltage control
Drive mechanism totally jams Drive motor torque and
with likely damage to drive
(locked rotor). drive motor current.
motor.
Slower increase in OLTC
Breakdown of insulation system. Gradual heating of tap temperature (compared to
Stops over correct tap contact but changer oil and cumulative main tank temperature)
Damage to mechanism. with insufficient contact pressure. damage to contact. when remaining on this tap
Wear of mechanical components. setting for longer periods.
Faster increase in OLTC
Rapid heating of tap changer
temperature (compared to
Stops between two adjacent tap oil due to series resistor
main tank temperature)
positions. remaining in circuit
with heating rate
continuously.
dependent on load current.
Low oil level leading to arcing Serious damage or
Header tank oil level.
fault. explosion.

LTC monitoring is often integrated with transformer main tank monitoring.

5.3 BUSHING & CT

The key parameters for online monitoring include partial discharge, capacitance, and
power factor or tan (Table 5-3). Due to the difficulty of PD monitoring, the only key
parameters left are the capacitance and tan . There are also some systems that
monitor insulation leakage current at the bushing tap. Dramatic changes in the signature
of this current provide indication of problems in the bushing.

285
 Table 5-3 Bushing & CT Monitoring Needs [145]
Natural Outcome if Left Signals Used for Online
Natural Aging Process Factors that Accelerate Aging
Unchecked Monitoring
Rapid temperature swings or
extreme environmental Crazing allowing dirt build-up
conditions, conductive oil by- and water ingress with Changes in tan and
Damage to porcelain. reduction of insulation leakage current.
products built up on porcelain
inside walls, and oil leak leading properties.
to paper drying out and cracking.
Localized partial discharge
Partial discharge RF
promoting further paper (and
Core paper deterioration. Elevated operating temps. currents and changes in
oil) deterioration (possible
tan and leakage current.
runaway condition).
Localized partial discharge
Partial discharge RF
Similar mechanism as within promoting further oil (and
Oil deterioration. currents and changes in
transformers. paper) deterioration (runaway
tan and leakage current.
condition).
Design/manufacturing defects, Increase in bushing C1
Higher capacitance leading to
Short circuits between foil layers. paper/oil aging, and conductive (and possibly C2)
increased capacitive current.
ink migration. capacitance.
Localized insulation damage
Partial discharge (corona) Partial discharge RF
Small voids within insulation with formation of carbon
caused by tiny flashovers inside currents and changes in
system. and/or shorts between foil
voids in insulation. tan and leakage current.
layers.
Frequent thermal cycling
results in excessive seal
Differing coefficient of expansion Usually will show up as an
wear, in turn leading to oil
of bushing components and increase in tan and/or
Large temperature variations. leak and/or water ingress.
rapid cooling can create bubbles partial discharge RF
Also end paper damage &
in oil. currents.
possible corrosion in tap
chamber.

5.4 EXAMPLE MONITORING SYSTEMS

A limitation of many transformer monitoring systems is that they are not able to control
or make decisions and recommendations based on the available data, forcing engineers
to spend a great deal of time sorting and interpreting the information they receive.
ABB’s Transformer Electronic Control (TEC) monitoring system [143], offered for new
transformers, and addresses this issue. To achieve the goal of making power
transformers “intelligent” and maintenance-free, ABB created and integrated a common
electronic interface to exchange information with the following apparatus:

Monitoring and diagnostics devices of the transformer and components

Transformer control cabinet
Tap changer motor-drive
Voltage regulation system
Overall protection system

TEC receives all the information it needs for transformer control from these sensors;
other necessary parameters are calculated. Through this interface, TEC provides exact
status information to enable utilities to extend transformer lifetime and save costs by

286
reducing maintenance and increasing availability. It does this by generating a model of
the transformer and its working condition and then comparing the measured parameters
with the simulated values. Discrepancies are detected and potential malfunctions and/or
normal wear in the transformer and its ancillaries are indicated.

For retrofit applications, ABB also proposes the T-Monitor, described in Table 5-4. In
addition to the standard data acquisition of various sensors and built-in models, an
implemented prognosis /diagnosis tool helps the operator in taking the right decisions.
With a remote control access option ABB can support the operator for troubleshooting.

Table 5-4 provides general descriptions of some of the systems available on the market
for monitoring various parameters and accessories on transformers. Note that this is
only a small sample of the various devices and systems that are available from a
multitude of suppliers.

287
 Table 5-4: Examples of Typical Monitoring Systems for Transformers

Monitoring Sample Pictures General Description

Systems/Devices
A common electronic interface to new
ABB transformers to provide add-on
functionality. The system is composed
of several sensors designed and built
into a new transformer. The system
measures or estimates the following
parameters:
Hotspot at HV/LV/TV and estimate
of hotspot temperature
Cooling control
OLTC contact wear
ABB WEB interface
Transformer Electronic
Control (TEC) Monitor Moisture in oil measurement
Bubbling temperature
Insulation aging due to moisture
Transformer temperature balance
Oil level in conservator
Online gas-in-oil interpretation
OLTC temperature balance
OLTC torque
A proven retrofit solution that provides
adequate predictive power by means of
easily fitted add-on sensors and
models that use the available
information about the transformer and
its component design. The system
measures the following parameters:
Top oil and ambient temp
Coolers temp
Dissolved gas and moisture in oil
Load current
ABB
T-Monitor Bushings
OLTC (Temperature difference
between main and OLTC tanks,
motor torque)
Acoustic partial discharge

The system performs the following

calculations (on-line models):
Winding hotspot calculation
Insulation aging calculation
Overload capability calculation
Moisture saturation in the oil
Moisture in the insulation
Bubbling temperature based on
hotspot calculation
Contact wear in the OLTC from
current at OLTC
Cooling optimization
Communication: Ethernet, LAN, Web

288
 Monitoring Sample Pictures General Description
Systems/Devices

The HYDRAN 201R Model i intelligent

fault monitor reads a composite value
of gases, in ppm, generated by faults
and provides output capability for
communications. It consists of a
HYDRAN 201Ti intelligent transmitter,
a HYDRAN 201Ci-1 one-channel
controller, and HYDRAN HOST
software. The HYDRAN 201Ti
transmitter attaches to a valve on the
transformer to be monitored. The
HYDRAN 201Ci-1 controller handles
communications and provides display,
alarms, and analog outputs.
GE Energy
HYDRAN 201R + M2 The HYDRAN M2 reads a composite
(Courtesy of GE Energy) value of gases [ppm] and the moisture
content in oil [%] at a corresponding
temperature.

The Calisto system reads dissolved

hydrogen in oil and moisture content in
oil.

Morgan Schaffer
Calisto
Dissolved Hydrogen and
moisture monitor

289
 Monitoring
Sample Pictures General Description
Systems/Devices

The Serveron Transformer Monitor

offers accurate and repeatable online
DGA of the 8 critical fault gases and
other key parameters. Correlation to
real events is accomplished through
time stamping of all gas data relative to
SERVERON transformer load, ambient temperature,
On-line Transformer oil temperature, and moisture-in-oil
Monitor measurements. Hourly sampling,
(Courtesy of Serveron) trending, and alarm signals are
available for all measured parameters.
Field-proven in utilities worldwide, the
gas chromatograph technology in the
Serveron Transformer Monitor offers
high reliability and low cost of
ownership.

The TRANSFIX Transformer Gas

Analyzer offers accurate and
repeatable online DGA of the 8 critical
fault gases as well as moisture.
Hourly sampling, trending, and alarm
KELMAN signals are available for all measured
TRANSFIX parameters. It allows extensive local
Transformer Gas Analyzer and remote communication options.
The system requires only minimal
maintenance and no consumables are
needed (no carrier gases).

290
 Monitoring
Sample Pictures General Description
Systems/Devices

The AQUAOIL 400 system is a stand-

alone unit for field installation on a
transformer valve. It allows the user to
monitor online the relative humidity in
the oil and its changes during load
GE Energy variation.
AQUAOIL 400
(Courtesy of GE Energy)

The Vaisala sensor reads the moisture

in oil and temperature.
It allows the user to monitor online the
relative humidity in the oi.
VAISALA
HMP228
Moisture and Temperature
Transmitter for Oil

Doble’s Moisture in Oil sensor

measures the relative saturation and
temperature of the insulating fluid and
calculates the moisture concentration
in parts per million. The DOMINO’s
Doble stainless steel probe is placed directly
DOMINO into the insulating fluid through an
(Courtesy of Doble access point on the transformer wall,
Engineering) typically a sampling or drain valve.
The IDD for Moisture estimates the
moisture concentration in paper and
the average moisture concentration
over time. The IDD for Moisture also
tracks the dielectric breakdown
strength and stores this hourly.,

291
 Monitoring
Sample Pictures General Description
Systems/Devices

The IDD, Intelligent Diagnostic Device

for Bushings and CTs, is a cost-
effective solution to continuously
evaluate the condition of bushings and
CTs while in service. This IDD
measures the electrical signal at the
Doble bushing and CT taps. The conditions of
Intelligent Diagnostic the bushings and CTs are evaluated by
Device (IDD) – Bushing Tap summing the leakage currents
Adapter measured at each tap. The analysis
(Courtesy of Doble requires monitoring all bushings or CTs
Engineering) in a three phase set. One IDD can
monitor up to four sets of bushings (2
sets of CTs) associated with the same
apparatus. Tap adapters specifically
designed for the particular bushing/CT
are mounted to the tap, allowing the
IDD to measure the leakage current.

The InsulGard system monitors

bushing insulation integrity on-line.

Cutler-Hammer
InsulGard G2

TM
The CENTURION On-line insulation
performance monitor detects any
changes in oil dielectric strength due to
moisture content or contamination
(carbon contaminants/burning or
metallic and ferrous contaminants). It
Weidmann
TM provides warning when it is time to take
CENTURION
remedial action and indicates the need
(Courtesy of Weidmann
for further remedial action. It provides
Electrical Technology)
information through local annunciation
or integration with a data collection
system for long-term trending of oil
degradation. It can be used on
Transformers, Load Tap Changers, Oil
Circuit Breakers and Regulators.

292
 Monitoring
Sample Pictures General Description
Systems/Devices

Standard temperature index

calculations: the difference in
temperature between the LTC
compartment and the main tank over a
specified period of time.

Standard temperature differential

calculation: the difference in
temperature between the LTC
compartment and the main tank.

GE Energy Advanced differential calculation:

LTC-MAP 2130 Compares any two sensor inputs for
(Courtesy of GE Energy) differential, averages differential, and
moving window index.

Integration over time of the square of

the current feeding the motor allows to
evaluate the condition of the
mechanical system driving the LTC.

Advanced moving window calculations:

For any sensor input, shows the
moving average, moving maximum,
moving minimum, moving trend, and
moving normal deviation.

The TM100 system monitors the

condition of the LTC

MR
Tap-changer TM100
monitoring system

293
6 PREVENTIVE MAINTENANCE OF TRANSFORMERS
6.1 BASIC AGEING PROCESSES [149]
6.1.1 INTRODUCTION
The main point of concern in ageing and life expectancy of transformers is the condition
of the insulation system, which is typically based on organic products. The following
organic products are found in an oil immersed transformer:

Mineral oil and

Paper and pressboard (cellulose)

The organic products in a transformer degrade over time and finally they lose the
capability to withstand the stresses a transformer might see in daily life (short circuits,
energizing, vibration, etc.). It is possible to reverse the ageing of mineral oil through oil
reclamation that can restore the material properties close to the values when new.
Ageing of paper insulation however is an irreversible process and is considered one of
the life-limiting processes of a transformer. The main factors that contribute to the
degradation processes of the organic materials are:

Temperature
Moisture
Oxygen
Acidity

The lifetime of this combination of mineral oil and paper in a transformer is very much
dependent on the operating temperature, oxygen content, acidity of the oil and the
moisture content in the insulation. Temperature is mainly dependent on the transformer
design, the loading, the cooling facilities, and the ambient temperature. Changing these
parameters is not easy and normally involves large investments. Moisture is
accumulated in the paper insulation of the transformer and has different sources. Poorly
maintained breathing apparatus on open breathing expansion tanks or damaged
gaskets on the cover can be a source of water from outside as can exposure of the
insulation material to air during a repair operation[150]. Also, the depolymerization
(ageing, degradation) of the insulation paper and the ageing of the oil create water as a
by-product inside the transformer [151,152].

New research also shows that the acid content plays a major role in the degradation
processes, in which the low molecular acids are of main concern. The acids are
produced as by-products of the oil and paper degradation. To extend the technical life
and to increase the reliability of the transformer, the moisture and acid levels in the
insulation should be kept as low as possible.

294
When a transformer leaves the factory, the insulation system is dry and almost free of
acids. This will change over the time as water and acids are produced through
degradation of the cellulose material and accumulate in the transformer. As a result, the
speed of degradation of the paper and oil will increase.

Removing moisture and acids from the transformer can slow down the ageing process
and thus extend the lifetime of the insulation system. These removal processes are
preferably performed on site, as moving a large power transformer to a transformer
workshop equipped with a vapor phase drying plant incurs considerable costs in both
time and money. Onsite drying and online oil reclamation are two processes that will
extend the remaining lifetime of the insulation system on site.

6.1.2 PAPER DEGRADATION

Ageing of the paper insulation in the windings is irreversible, and considered one of the
life limiting processes of a transformer. As the paper ages its mechanical properties
(tensile, bursting and folding strength) are reduced. This was first described by
Montsinger [153], who gave the rule that the rate of change of mechanical properties
increases with temperature and doubles (the life is halved) for every 6-8 degrees
Celsius increase according to the formula:

Rate of ageing = constant·e

Where is the temperature in Celsius and p is a constant.

Paper (and pressboard) consists mainly of cellulose and some percent of hemicellulose
and lignine. The reduction in mechanical strength is due to chain scissions of the
cellulose molecules. New oil-impregnated paper has an average chain length of 1000-
1200 polysaccharide rings in series – denoted degree of polymerisation (DP). The
tensile strength depends on the DP as shown in Figure 6-1. Conventionally a DP value
of 200 is used as an end of life criterion

Figure 6-1: Tensile strength dependence of DP for Munksjø kraft paper

295
Instead of using tensile strength to estimate ageing, one can use the scissoring of the
cellulose molecules ( ) , which in turn reduces the DP value and the Arrhenius
dependence according to Ekenstam [154]:

= DPnew·A·exp(-E/(RT))·t = DPnew·k·t

The equation shows how the number of chain scissions ( ) increases with time (t). The
expression A·exp(-E/(RT)) describes the ageing rate k. R is the molar gas constant, T
the absolute temperature and E the activation energy, which describes how the ageing
varies with temperature. The higher the value of E, the faster the ageing rate increases
with temperature. A is a factor that depends on the chemical environment of the
cellulose.

It has long been established that the ageing of cellulose can be divided in two
processes: oxidation and hydrolysis [155,156]. For both processes the ageing rate
increases with temperature. There are strong indications [157] that the temperature
dependence is different for the two processes, and it may well be that oxidation
dominates at lower temperatures and hydrolysis at higher temperatures.

Under oxidation, the ageing agent is oxygen from air ingress. The ultimate end products
of oxidation are the same as for combustion, i.e. water and carbon dioxide. However,
the mechanism of low temperature oxidation is quite different from that of combustion.
The oxygen concentration is an important parameter that determines the rate of
oxidation. However, most experimental studies show that the ageing rate is not so
strongly influenced by oxygen content. Typically, the overall degradation rate will no
more than double in experiments with oxygen present, compared to when oxygen is
totally excluded. It can therefore be concluded that the importance of oxygen to ageing
is limited.

The other major mechanism of paper ageing is hydrolysis. The significance of water
content is paramount: a humidity of 3-4% may increase the degradation rate of paper by
a factor of 10 or more, compared to dry paper. This underlines the importance of
assessing both moisture and temperature of the insulation system. It must be
mentioned that these studies are from laboratory samples and may not be directly
transferable to transformer life assessment.

Newer theories on hydrolyses proposes that the process is due to acid catalysis; H+ (or
rather H2+O) ions from carboxylic acids that are dissociated in water serve as catalysts
to the chain scissoring of the cellulose molecules [158]. Since both hydrolysis (and also
oxidation) produces carboxylic acids and water, this makes hydrolysis an auto-
accelerating process.

The various acids produced in a transformer influence ageing differently. This is

demonstrated in an experiment where five different acids were added to the oil up to a
neutralization value of about 0,4 mg KOH/g [159].

296
 1200
Formic
Acetic
1000
Levulinic
Naphtenic
800
Stearic

DP-value
No acid
600

400

200

0
0 200 400 600 800
Time [hours]

Figure 6-2: Ageing of paper at 130 °C in oils with an acidity of 0,4 mg KOH/g

As can be seen from Figure 6-2, the high molecular weight acids (stearic and
naphthenic acids) do not accelerate ageing. The effect of the other acids on ageing rate
is an increase with decreasing molecular weight. The reason for the different influence
on the ageing rate is mainly due to the hydrophilic nature of the lower molecular acids.
The acids with lower molecular weights dissolve easier in the paper than the stearic and
naphthenic acids [160].

The significance of these findings for maintenance is important. The ageing accelerating
substances – water and low molecular acids - will tend to dissolve better in paper than
in oil. Removal of these substances from the cellulose is therefore what the winding
maintenance activity should aim for. Oil reclaiming will certainly remove some of these,
but other methods may be more efficient.

It is in this context important to note that the measurement of the neutralization value as
is standardized in IEC 62021-1 and ASTM D974 mainly detects acids formed by oil
ageing. A new measurement technique is needed for the detection of the dangerous
hydrophilic acids. Presently there are no standards methods for estimating the content
of these acids in a transformer. One technique may be “water washing” of an oil sample
to remove the low molecular weight acids from the oil sample. The difference in
neutralization number between a measurement of a “pure” and a “water washed” oil
sample may yield a descriptor for the low molecular weight acids.

297
 6.1.3 ON-SITE DRYING METHODS

6.1.3.1 T RADITIONAL METHODS

Circulation of hot dry oil
Dry oil is circulated through the tank and moisture that is extracted from the oil is
absorbed in a vacuum degasifier. The method can be efficient in cases where the
insulation is made of thin structures and the weight of the insulation is low.

Heat /Vacuum Cycles

This technology incorporates the following procedures:

Heat core and coil with hot oil circulation

Drain the oil
Vacuum treatment
Cycle repetition (if necessary)

Hot oil spray

This technology incorporates the following procedures:

Bring oil level to the bottom part of the core

Oil spraying under vacuum
Increase vacuum level after spray is stopped to ensure final dry out
Atomizing nozzles for the oil spray are recommended

Combined oil-spray / Hot air / Vacuum / Oil circulation / Cycles process

Pre-starting procedures:

Drain the oil

Vacuum pre-dry the tank, core and coil
Bring oil level to the bottom part of the core with dry, clean, stable, well-soluble
oil

Heating process:

Oil-spray through spreading pipes installed above the windings under vacuum.
Periodically circulate dry air through the tank to maintain the surface
temperature.

Drying process:

Cycle vacuum treatment with circulation of transformer oil through the oil heater /
filter followed with a cycle of oil spraying at low vacuum to maintain average
drying temperature.

298
6.1.3.2 ON-SITE DRYING WITH LOW FREQUENCY HEATING (LFH) IN COMBINATION WITH HOT-OIL SPRAY
[161]
Temperature and vacuum are the main factors for drying speed and drying quality. For
optimized drying the transformer should be heated at the same time as vacuum is
applied (as is done during the vapor phase process). With the combination of the low
frequency heating (LFH) technique and hot oil spray or hot oil circulation, this can also
be achieved on site.

The low frequency voltage is necessary in order to reduce the applied voltage on the
transformer when it is under vacuum. The reason is that the breakdown voltage of the
insulation system is much lower under vacuum than under atmospheric pressure. This
is also known as the Paschen law. In order to prevent hotspots during drying (due to
reduced cooling), the applied current for heating power transformers should not exceed
50% of the nominal current. Consequently, the copper losses will be no more than ¼ of
the nominal losses and the risk of potential hot spots at points with higher resistance will
be negligible.

Plant and process concept

In order to heat up both the low and high voltage windings, a frequency of approx. 1 Hz
is applied to the transformer. With the combination of LFH drying and conventional hot
oil spray method, the whole transformer can be heated very uniformly. The LFH system
heats the windings from the inside and the hot oil spray supports the heating process by
heating outer parts of the insulation system.

Short circuit HV.

LV.
Vacuum control
valve
LFH Converter

Spray
nozzles

Oil treatment plant

Vacuum
pumps

Figure 6-3: Plant concept for a mobile LFH drying process in combination with hot oil spray.

299
 6.1.4 OIL RECLAIMING

6.1.4.1 ONLINE OIL RECLAIMING TECHNOLOGY [162]

The essential elements of the online reclamation technology are that the absorbent is
automatically reactivated after each cycle and that the transformer remains in operation.
The reactivation allows for using much more active absorbent than with the classical
type of reclaiming plants, where the Fullers earth needs to be replaced frequently and
thereafter has to be disposed of. To achieve a long lasting effect, the complete oil
volume is typically circulated 8 to12 times over the Fullers earth. The final step is to
restore the inhibitor content.
6.1.4.2 COMPARISON WITH OIL CHANGE
There are several attractive features of online reclaiming compared to an oil exchange.
For safety reasons it is sometimes necessary to de-energize the transformer when the
equipment is being connected and disconnected. However, the process can be
performed when the transformer is energized. This can present some obvious
economical advantages.
From a technical point of view the continuous “washing” of the solid insulation
represents a great advantage. During ageing of the oil, large amounts of acids, sludge
and other degradation products are absorbed by the paper and other cellulose material.
These contaminants can later be redissolved into the new clean oil. In the case of
reclaiming, the contaminants are constantly being removed from the oil during the
process and they can be permanently removed from the cellulose material. In the case
of oil replacement without proper cleaning of the active part, the residual degradation
products will cause a substantial shortening of the life of the new oil. Without an
appropriate washing procedure, the oil will typically be degraded again after only a few
years (See Figure 6-4).
6.1.4.3 LONG- TERM STABILITY
The long-term stability and oxidation stability of reclaimed oil has been questioned. One
reason is that the conventional reclaiming process (without reactivation of the
absorbent) with the daily exchange of the Fullers earth was often performed using a
limited amount of absorbent. As a result, the ageing by-products were not effectively
removed from the cellulose material and were reabsorbed into the reclaimed oil. The
effect is a recontamination of the processed oil within a very short time, as shown in
Figure 6-4.

300
 Aging of transformer oil after reclamation and oil
change
Severely
aged oil0,3
mg KOH/g, Acidity level
0,25

0,2

0,15

0,1

0,05
New oil
0
Before After process 3 mnd. 2 years 4 years 6 years 10 years
Time

Online reclamation Little absorbent Oil change

Figure 6-4: Evolution of total acidity in oil after online reclamation, oil change and reclamation
with little absorbent

The modern reclaiming technology with the reactivation of the absorbent makes it
economically feasible to use much more active absorbent material. Typically 5-10 times
more absorbent is used compared with the old conventional systems. This leads to a
much better cleaning effect of the paper and to an excellent long-term stability of the
reclaimed oil.

For some transformers, there is data available for up to 10 years since reclamation. The
data shows very little change in acidity and color after 10 years in operation since the oil
was reclaimed. For a successful reclamation process with excellent long-term stability
approximately 700 kg of active absorbent is necessary to treat 1000 kg of oil.

To prevent excessive production of acids due to oil degradation, it is recommended that

the oil be reclaimed before it is severely degraded. Low molecular weight acids do
migrate into the paper insulation and become much harder to remove than from the oil.

301
6.2 GENERAL MAINTENANCE OF TRANSFORMERS [163]

The primary purpose of transformer maintenance is to ensure that the internal and
external parts of the transformer and accessories are kept in good condition (“fit for
purpose”) and are able to operate safely at all times. A secondary, equally essential
purpose is to maintain a historical record of the condition of the transformer.

Transformer maintenance can be done periodically or as condition-based maintenance.

On a periodic schedule, the frequency of inspection and maintenance procedures will
vary with the rating of the transformer, but the intervals suggested below are
recommended as minimums. Condition-based maintenance is usually the most
economical way of doing maintenance. Recommended maintenance is then done
based on one or more of the following: inspections, analysis of oil samples, electrical
measurements, test of equipment, measurement of temperatures by using a heat-
sensitive camera, and/or monitoring (offline and/or online).

For personal safety reasons, only a limited amount of maintenance activities should be
performed on the transformer when it is in operation. Be sure to follow the
manufacturer’s recommended safety requirements before any maintenance activity is
undertaken.

6.2.1 RECOMMENDED SCHEDULE OF MAINTENANCE ACTIVITIES

CAUTION: Before performing any maintenance work near the transformer

terminals, be certain that the transformer is de-energized. Ground the transformer
terminals before entering the area at the top of the transformer. Failure to de-
energize the transformer may lead to equipment damage, severe personal injury,
or death.

6.2.1.1 MONTHLY MAINTENANCE SCHEDULE

1. Check and record the ambient temperature.
2. Check and record the transformer liquid temperature and note the maximum
value since the last reading.
3. Check and record the transformer winding temperature and note the maximum
value since the last reading.
4. Check and record the transformer load current and note the maximum value
since the previous reading.
5. Check and record the line voltage and note any variation from rated value since
the previous check.
6. For transformers equipped with Sealedaire®, check and record the reading of the
pressure-vacuum gauge. If the gauge remains at or near zero when the oil
temperature varies, the transformer should be checked for leaks. This is an
important maintenance check which will verify the integrity of the transformer
seal.

302
6.2.1.2 QUARTERLY MAINTENANCE SCHEDULE
1. Check and record the readings on all indicating instruments, such as the liquid
level of the main tank, all oil-filled compartments, top oil temperature, and
winding temperature. The maximum reading and the present reading should be
noted on the temperature indicators. If the transformer is equipped with an
Inertaire® oil preservation system, check and record the tank pressure and the
remaining pressure in the nitrogen bottle feeding the system.

Examine the piping to the coolers or radiators and all bolted pipe joints for signs of oil
leakage. Tighten any loose fittings and repair any oil leaks.

CAUTION: Some locations on the cooling equipment may be near the

transformer line connections. It may be necessary to de-energize the
transformer to work in these areas. Use proper safety procedures.

2. Examine the coolers or radiators for accumulation of dirt and foreign material that
might impede airflow. The coolers or radiators can be cleaned by directing a
stream of low-pressure water over the surfaces. On FOA coolers, the water
should be directed to the front side of the cooler to wash any dirt toward the rear.
Be certain the fans are shut off before starting any cleaning operation. The
frequency of cleaning will vary depending on the conditions at the installation
site. Annual cleaning is generally sufficient, but installation subjected to salt spray
or heavy dust and dirt will require a more frequent schedule.

3. Inspect the control cabinet for the following conditions:

CAUTION: The control circuits may have dangerous voltage levels. De-
energize the auxiliary power source before working on any control
components. Failure to do so may cause personal injury or equipment
damage.

Control-circuit voltage
Collections of dirt or gum
Excess heating of parts (evidenced by discoloration of metal parts,
charred insulation, or odor)
Binding or sticking of moving parts
Corrosion of metal parts
Remaining wear allowance on contacts
Excess slam on pickup
Proper contact pressure
Loose connections
Condition of flexible shunts
Worn or broken mechanical parts
Excessive arcing in opening circuits
Excessive noise in ac magnets
Evidence of dripping water or liquids falling on controls

303
 Operation, including proper functioning of timing devices and sequencing
of devices
6.2.1.3 ANNUAL MAINTENANCE SCHEDULE WITH T HE T RANSFORMER DE-ENERGIZED
1. If the transformer is equipped with forced oil cooling, check the oil circulating pumps,
noting any unusual noise or flutter of the oil flow gauge. Evidence of noise, uneven
oil flow, unbalanced phase current, or heating of the pump motor may require
removal of the pump from the transformer. Disassembly and inspection procedures
are given in the pump instruction leaflet.

2. If the transformer is equipped with air cooling, examine the fans to ensure that there
is no debris covering them or between the blades. Check to make sure that each fan
is operational and that the blade rotation is correct. If the cooling banks are staged,
check that the proper set of fans come on for each stage. Measure line currents on
the fan motor and check for any imbalances.

3. Examine the pump valves for evidence of leaking around the gland seals. Close and
open the flapper-operating arm. There should be some restriction to the flapper arm
movement if the packing is properly tightened.

CAUTION: Do not attempt to operate the pump valve when the pumps are in
operation. Always shut off the pump motor before opening or closing any
valves. Failure to follow these precautions may cause equipment damage and
personal injury.

Tighten the gland nut if necessary to eliminate any leaks.

Take oil samples from the main tank and any other oil-filled compartment, such as the
load tap changer. Perform general oil quality analyses on the oil sample. Oil samples
may also be taken at this time for gas-in-oil or metal particle analysis. If any test results
are questionable, contact ABB.

Perform insulation resistance tests on each winding to the other winding and to ground
and from all windings to ground and compare with the previous test values. Measure the
insulation power factor and compare with previous test values. Contact the Technical
Support Section if any of the tested values vary significantly from the initial tests.

4. Examine all bushings, arresters, and all the interconnecting hardware for
contamination and signs of electrical tracking. Clean any contaminated areas with a
soft cloth and suitable solvent, then wipe the area dry. Perform power factor and
capacitance measurement on the bushings and compare the values to the test
results made when the transformer was installed.

5. If the transformer is equipped with a load tap changer, inspect the tap changer as
noted in the tap changer instruction leaflet. Detailed information for the inspection
procedures and the frequency of inspection is supplied as part of the transformer
instruction book.

304
6. Inspect any breathers and small screen openings in pressure-relief valves or a
pressure-vacuum breather to be certain they are clean and in operating condition.

7. If the transformer is equipped with a COPS oil-preservation system, remove the

expansion tank breather and check for oil leakage into the bladder. The procedure
for making this inspection is explained in the instruction leaflet for the oil
preservation system.

8. Examine the paint finish, particularly around welded joints and on accessory items
such as the radiators, coolers, and associated piping. Check for paint peeling or
cracking and evidence of rust. Clean the affected areas by wire brushing, then wipe
with a clean dry cloth. Paint the area with the touch-up primer and finish coat
supplied in the transformer details shipment box.

De-energize the auxiliary power source and inspect the control devices in the control
cabinet. Remove grease, oil, or other contaminants with a lint-free cloth moistened in a
nonflammable cleaning fluid. Do not soak the parts with the cleaner, but use just
enough to loosen grease or dirt so that it can be wiped off. For cleaning small parts, a
small paintbrush dipped into the cleaning solution is good for getting into corners and
crevices. Repair or replace any broken or malfunctioning parts, tighten all loose
connections, and eliminate any oil or water leaks into the compartment.

More frequent inspections may be needed in heavily contaminated installations.

6.2.2 MAINTENANCE OF COMPONENTS

6.2.2.1 T RANSFORMER LIQUID AND INSULATION

The task of oil in a transformer is to act as an electrical insulation and transfer heat from
the transformer’s active parts into coolers. Oil acts as a good electrical insulation only
as long as it is satisfactorily dry and clean.

Moisture balance between the oil and the solid insulation implies that most of the
moisture will gather in the paper insulation. Moisture in insulation is one of the dominant
ageing accelerators. It is recommended to dry the insulation when the moisture exceeds
a certain level. Drying of the insulation and oil is recommended for large distribution and
power transformers since this can be technically and economically motivated.
Equipment for drying transformers at site is available, and the residual moisture in the
insulation will be less than 1% after drying with low frequency heating equipment.
During drying, the transformer has to be de-energized. Drying time can vary from one to
two weeks depending on the transformer size, amount of insulation, and initial moisture
level in the insulation.

Testing of oil in transformers should normally be performed 12 months after filling or

refilling and subsequently annually on large distribution and power transformers. ABB
offers different tests and analyses of oil samples, depending of transformer type, size,
service record, and strategic importance for safe electrical supply. Testing of oil in on-

305
load tap changers must be performed according to the tap changer supplier’s
recommendations.

Taking oil samples from hermetically sealed transformers is normally not necessary,
and should only be performed after consultation with ABB. The oil in this type of
transformer is not in contact with the atmosphere and less exposed to moisture. Oil
regeneration/reclaiming of oil may be technically and economically motivated, especially
for large distribution and power transformers. Reclaiming implies filtering, de-gassing,
removing ageing by-products, and adding an inhibitor if required. Reclamation of oil is
performed with the transformer in service (operation). The transformer is only de-
energized for a few hours when the equipment is connected and disconnected from the
transformer. If the oil is in good condition, except from particles present in the oil,
filtering can be recommended for removal of the particles.

Often it is recommended to do both drying and reclaiming on a transformer at the same

time. If this is done at the right time, i.e. before the degradation of oil and insulation has
gone too far, the lifetime of the transformer can be extended by several years.
6.2.2.2 BUSHINGS AND JOINTS
The porcelain insulators of transformer bushings ought to be cleaned during service
outages as often as necessary. This is particularly important for places exposed to
contamination and moisture. Use recommended cleaning agents for cleaning the
bushings. The condition of external conductor and bus bar joints of transformer
bushings should be checked at regular intervals because reduced contact pressure in
the joints leads to overheated bushings, etc. and may cause the adjacent gasket to be
destroyed by the heat. An infrared camera can be used to check the temperatures in
joints, bushings, etc. Maintenance of HV condenser bushings shall be performed
according to the instructions given by the bushing supplier.
6.2.2.3 OFF-LOAD TAP CHANGER (DETC)
The transformation ratio can be adjusted with an off-circuit tap changer when the
transformer is not energized. The control shaft of the off-circuit tap changer is brought
through the cover or the tank wall. The shaft end is provided with a handle, position
indicator, and locking device. When the tap changer is turned, the locking device must
be secured, thus assuring that the off-circuit tap changer has been set to operating
position. Off-load tap changers do normally not require regular maintenance, but it is
recommended that the off-circuit tap changer is moved from one extreme position to the
other a few times during service interruption. This is necessary especially when the tap
changer is moved infrequently. Moving from one position to another is performed either
manually by a hand wheel or by a motor drive unit.

Total expected lifetime depends on the number of operations, normal current, etc.
Inspection/maintenance of tap changers must only be carried out by trained and
experienced personnel. See supplier’s documentation provided.

306
6.2.2.4 ON-LOAD TAP CHANGER
Maintenance of on-load tap changers should be performed according to the instructions
given by the supplier of the tap changer. In addition, it is strongly recommended that
only suitably trained personnel should undertake OLTC examination and maintenance.
On-load tap changers have to be maintained regularly. The maintenance interval and
total expected lifetime depend on the number of operations, normal current, if an oil
filtering unit is provided, etc.
6.2.2.5 MOTOR DRIVE UNIT
Motor drive units have to be maintained regularly. The maintenance interval and total
expected lifetime depend on the number of operations. Only trained and experienced
personnel should carry out maintenance on motor drive units. See supplier’s
documentation provided.
6.2.2.6 OIL FILTERING UNIT
The paper filter in the oil-filtering unit for the on-load tap changer has to be changed
when pressure loss has reached approximately 4 bars on the pressure gauge. See the
supplier’s documentation.
6.2.2.7 COOLERS
Coolers are cleaned by brushing inside the water tubes or by air-side vacuum cleaning
when necessary. The need for cleaning is indicated by increased pressure loss,
decreased temperature-difference oil/water/air in/out, increased transformer
temperature, decreased water flow, etc. See the manufacturer’s documentation.
6.2.2.8 LIQUID CONSERVATOR WITH RUBBER DIAPHRAGM (COPS)
This system consists of an oil conservator with a rubber bag. It is recommended that
the rubber bag be checked every two years for leaks. This is done by opening the
bleeder access at the top of the bag and swabbing the inside of the bag with a stick that
has a cotton cloth covering its end. Care should be taken not to puncture the bad. If
there is any oil on the cotton cloth, this indicates that the bag is leaking and should be
replaced. Also, it is generally recommended that the bag be replaced every ten years.
The silica gel breather should be inspected on a periodic bases and the silica gel should
be changed when approximately of it has changed from blue to red color (old type of
silica gel) or from pink to white (new type of silica gel).
6.2.2.9 GASKETS
The gaskets of the cover and flanges, as well as between bushings and cover, are
usually made of liquid-resistant vulcanized cork sheet, nitrile rubber, or silicone sealant.
If the gaskets are leaking, leaks can usually be sealed by tightening the screws (bolts).
When these gaskets have to be replaced, it is recommended to contact ABB. Liquid
resistant rubber rings are used as gaskets for bushing bolts, shafts, and spindles. All
these gaskets can be tightened and replaced from outside the tank. When tightening the
gaskets, special care must be taken to prevent the breaking of screws (bolts) or else the
gasket “floats away” (if not in a groove) as a result of the heavy pressure. In particular,
stud nuts must be tightened very carefully.

307
6.2.2.10 SURFACE PROTECTION

6.2.2.10.1 Painted surfaces

When repairing damaged paint, the points to be repainted should be cleaned from rust,
dirt, and grease before priming with a zinc-rich primer prior to top coat paint. The final
paint thickness should at least be equal to the original paint thickness. If major paint
damage is present, it is recommended that one contact a specialized surface coating
company.
6.2.2.10.2 Zinc coated surfaces
Zinc coated surfaces have a self-repairing, passivating characteristic. Small damages
such as scratches do normally not need repairing. Larger areas, above 50 mm2, may
need repair. After thoroughly cleaning, apply zinc-rich (between 65-69% zinc by weight,
or >92% by weight metallic zinc in dry film) paint to at least the same thickness as the
original zinc coating. Do not remove any original zinc during cleaning. The paint may be
one-component (preferred) or two-component.

6.2.3 INVESTIGATION OF TRANSFORMER DISTURBANCES

If, during operation, the protective equipment of the transformer gives an alarm or trips
the transformer from the network, one should immediately investigate the reason for it.
Studies may reveal whether it is a question of transformer damage or some other
disturbance in the system.
6.2.3.1 RECORDING OF DISTURBANCES
If a disturbance occurs to the transformer, it is important to make a record of the
following. This will be helpful to investigators in identifying the cause of the disturbance.

Date and time of the occurrence

Data for installed overvoltage protection
Network data: Were connections or other relevant things made when the
disturbance took place? What was the loading like? Are there possible relay
operations which took place elsewhere in the network (e.g. ground fault relay)?
Weather data (thunderstorm, rain, etc.)
Is the gas relay filled with gas? Color and quality?
Is the oil sooty?
Thermometer readings
Were coolers or tank damaged?
Are there visible marks of arcing on the bushings, cover, conservator, or
elsewhere?
Gas-in-oil analysis for power transformers
Any other observations

Special mention should be made above the operation of the protective equipment on
the transformer. Operation of some protective equipment such as gas relay or
differential relay does not always mean that the transformer is damaged. The gas relay
can operate for example when:

308
 An air bubble has been left under the transformer cover. An air bubble is
colorless and odorless.
A short circuit current has passed the transformer. No gas bubbles.

However, if the gas has color or smell, the transformer is damaged.

6.2.3.2 FAULT LOCALIZATIONS ADVICE FOR OIL-IMMERSED TRANSFORMERS
Table 6-1 provides probable causes to several disturbances to oil-immersed
transformers and in some cases, possible corrective actions. In most cases, however, it
may be necessary to contact the manufacturer for further advice on what actions to
take.

Table 6-1: Probable Causes of Disturbances to Transformers

SYMPTOMS PROBABLE SOLUTIONS

CAUSES
Low insulation resistance. Earth fault. Oil Contact ABB.
deficiency.
Unexpected secondary voltage. Primary voltage. Check installation and contact the
Absence of electricity utility.
primary voltage.
Tap changer or Change position or connection.
bolted links
incorrectly
positioned or
connected.
Winding rupture. Contact ABB.
Blown fuse in one Change fuse.
phase.
Non-symmetrical voltages on the Bolted links Check the connections.
secondary side. incorrectly Check installation and contact the
connected in one electricity utility.
of the phases.
Winding rupture. Contact ABB.
LV. installation. Check LV. Installation.
Nonsymmetrical
load on the
secondary side.
No voltage Contact the electricity utility.
applied in one of
the phases on the
primary side.
Triggering of the over-current relay. Short circuit in the Remove the failure in the system.
system on the
secondary side.
Winding rupture. Contact ABB.
Triggering of differential relay during Internal failure in Contact ABB.
operation. the transformer.
Failure in current Check current transformers.
transformers
feeding the relay.

309
SYMPTOMS PROBABLE SOLUTIONS
CAUSES
Spurious triggering during operation. Triggering and Check settings. Check thermometer.
alarm incorrectly
set. Incorrect
thermometer
operation.
Defect Ptl00 Check sensors or thermistors.
sensors or
thermistors.
Relays incorrect Check timing.
timing.
Short circuit in the Remove the failure in the control system.
control system on
the secondary
side.
Abnormal operating temperature Local heating at Clean contact surfaces and retighten.
measured by thermography. the transformer
terminals.
Excessive cable Undersized cables.
heating.
Winding and/or top-oil thermometer Insufficient Check ventilation of premises. Consider
alarm and/or trip. ventilation. installation of cooling fans.
High ambient
temperature.
Transformer Consider load reduction or installation of a
overloaded. transformer with higher power rating.
Reduced oil, Check oil, water, and air circulation.
water, or air
circulation.
Too high oil Reduce load.
temperature.
Measurement of unexpected voltage Ground failure on Remove failure.
to ground. one phase.
High acoustical sound level. Supply voltage Reduce supply voltage or change position
higher than on tap changer. Retighten.
presupposed.
Loose
accessories or
elements.
Reflection from Install sound damping panels. Place the
walls and other transformer in non-parallel direction to the
elements. walls. Use damping pads below the
transformer.
Low frequency. Contact electricity utility.
Oil flow trip. Oil circulation too Open valves in oil circuit.
low.
Oil pump Check oil pump and protection.
protection.
Buchholz-gas relay alarm. Gas bubbles De-energize the transformer.
caused by local If the captured gas is flammable, carry out
overheating. dissolved gas analysis (DGA). Contact
ABB.

310
SYMPTOMS PROBABLE SOLUTIONS
CAUSES
Gas bubbles If the captured gas is not flammable, bleed
caused by the transformer properly and energize.
incomplete
bleeding
Buchholz-gas relay trip. Arcing in active Carry out dissolved gas analysis (DGA).
part. Contact ABB.
Oil level too low. Adjust oil level and repair leakages.
Welding on the transformer is only allowed
if the transformer is filled by oil/inert gas
(nitrogen).
Oil level indicator: alarm high level or Incorrect oil level. Adjust oil level. Repair leakages, if any.
trip low level. Welding on the transformer is only allowed
if the transformer is filled by oil/inert gas
(nitrogen).
Leakage detector alarm. Leakage in Repair/change the cooler.
cooler.
On-load tap changer protective relay Sudden pressure Inspection/repair of tap-changer diverter
trip. rise in tap switch.
changer
compartment.
On-load tap changer out of step trip. Operation of tap Check tap changer, interlocking, and
changer failed. synchronism.
Pressure-relief device trip. Sudden pressure Carry out dissolved gas analysis (DGA).
rise transformer. Contact ABB.
Gas-monitoring alarm. Gas-detection. Carry out dissolved gas analysis (DGA).
Contact ABB.

NOTE: Contact ABB specialists before inspection, adjustment, and repair of vital parts.

311
 6.2.4 INTERNAL INSPECTION
The present methods of oil preservation (Sealedaire®, lnertaire®, or COPS) protect the
interior of the transformer if the systems are functioning properly. Internal inspection of
the transformer above the core and coils is not necessary unless results from the oil
analysis indicated potential problems. Sludging of the oil, low dielectric strength,
moisture in the oil, or the presence of combustible gases are conditions that may merit
an internal inspection of the transformer. Severe system disturbances, incidence of
through-fault, or a circuit breaker operation would also be reason for an internal
inspection of a transformer. Refer to the latest version of IEEE C57.125 [164] for further
direction on performing internal inspections.

WARNING: BE CERTAIN THE TRANSFORMER AND THE AUXILIARY POWER TO

THE CONTROL CABINET ARE DE-ENERGIZED BEFORE BEGINNING ANY OF THE
WORK. GROUND ALL THE TRANSFORMER TERMINALS. FAILURE TO DO THIS
MAY RESULT IN EQUIPMENT DAMAGE AND SEVERE PERSONAL INJURY.

CAUTION: Do not enter any transformer until the gas in the tank is replaced by dry air.
Oxygen content inside the transformer must be at least 19.5 % before entering for
inside inspection. The oxygen content must always be checked. Oxygen contents less
than 19.5 % may cause drowsiness and/or injury and death.

If it is necessary to filter the oil, follow the proper procedures supplied by the
manufacturer.
6.2.4.1 OPENING THE T RANSFORMER

CAUTION: Before opening a transformer or removing an inspection cover, be

certain that the internal pressure has been lowered to atmospheric pressure. The
oil level must be lower than the cover being removed. Failure to observe these
precautions may cause equipment damage or personal injury.

Remove the manhole cover and proceed with the internal inspection. The time the
transformer is open for inspection should not exceed two hours. Dry air must be
circulated in the tank during inspection. Consult the manufacturer’s instruction manual
for the required airflow rates.

Internal inspection and maintenance procedures will require removing and replacing oil
in the transformer. Pump the oil into clean storage containers. The oil level should be
kept as high as possible while the internal work is being done to minimize the exposure
of the insulation and to prevent the entrance of moisture.

CAUTION: Persons entering the transformer must not have loose dirt particles on
their clothing. Clean cloth shoe covers or nitrile rubber overshoes must be worn
by anyone entering the transformer. Failure to observe these precautions may
cause an electrical failure.

312
The amount of oil removed and the length of time the transformer is open will determine
the procedure for treatment of the oil. Open time starts each time the transformer seal is
broken and ends when the transformer is resealed, refilled with oil, or pressurized with
dry gas. The time is cumulative. Feed bottled dry air through a pressure regulator into
the tank while it is open.

If the insulation of the transformer has been exposed to air for some period of time
during the inspection/repair of the transformer, the moisture level in the insulation will
increase. Depending on increase in moisture level, on-site drying of the transformer
should be considered.
6.2.4.2 T HE INSPECTION
The principle objective of the internal inspection is to locate any damage which might
have occurred in service. The inside assembly drawing furnished with the transformer
should be used as a guide during internal inspection. Pay careful attention to the
following areas:

Core:
Open the core ground and measure core insulation resistance to assure
proper core isolation. Multiple core grounds may permit damaging circulating
currents to flow within the core.
Visually inspect the accessible core components for movement, displacement
or looseness of core insulation, and mechanical supports. Look for evidence
of localized hotspots or discoloration on the core.

Cable Leads and Bus Bars:

Check leads, bus bars, and all mechanical lead supports for physical
condition and electrical or thermal damage. Check alignment and lead-to-lead
clearances. Tighten or replace loose or damaged hardware as required. Any
evidence of overheating in the cable insulation should be investigated further.
If the insulating tape is to be removed from a cable for further examination,
measurements of tape dimensions, cable spacing, and clearances must be
documented and the number and dimensions of papers recorded. Save the
removed paper for further examination. Replace the insulation with properly
dried and oil impregnated materials to the original tape build (thickness).
Examine the line lead exits as they emerge from the coils for proper securing
(clamps and/or cotton tape ties) and alignment.
Examine regulating leads where they exit the coils and DETC lead
connections at the DETC deck as above.

Windings:
Inspect accessible winding conductors for mechanical deformation and
electrical damage.
Disc windings are supported mechanically by multiple columns of pressboard
radial or key spacers arranged at intervals within the winding. This

313
 arrangement provides the necessary mechanical support to couple vertical
winding forces (short circuit forces) into the primary mechanical winding
supports and end frame structures. It is important that these vertical spacer
columns are properly aligned and secure. Check the column alignment.
On shell form transformers, check the inter-phase blocking to be certain that
the wedges are tight and secured. Tighten any loose wedges by tapping them
in place with a non-metallic mallet and secure the dowel pins at the top of the
wedges.
Check the winding end insulation items for proper alignment and voids
between the top of the winding and the coil support. Check for any loose
radial spacer blocks. Look for tilt or misalignment of the radial spacer
columns, especially in the top 20% or so of the winding. A mode of short
circuit motion in the outer winding is a tendency for that winding to try to
“unwind” due to the repulsive forces between the outer and inner winding.
The effect on the inner winding is opposite with a tendency to tighten the
winding. Evidence of looseness or misalignment will require an engineering
analysis as to possible corrective actions. Some retightening is possible in
place but access to the inner windings is, of course, limited.
Obtain samples of pressboard for DP analysis. The samples are best taken
from an area near the top of the outer winding and adjacent to the conductor,
as this is the area of greatest heat. Make sure a supply of properly dried and
impregnated pressboard is available to replace any spacer blocks that may
be removed for this analysis or found to be defective.
6.2.4.3 ELECTRICAL T ESTS
If there is evidence of internal damage, contact the manufacturer. The following tests
may be made as part of the internal inspection procedure:

1. A ratio test on all windings and tap positions.

2. If the transformer is filled with oil, an insulation resistance test from each winding to
all other windings and ground and from all windings to ground should be made.
Record the temperature of the oil. These readings should be compared with the
measurements made at the time of the initial installation.
3. Make an insulation power factor measurement and compare the values with the
measurement made at the time of initial installation. The transformer must be filled
with oil for this test.
4. If the transformer has an accessible core ground, disconnect the core ground lead
and measure the resistance from the core to ground. Compare the value with the
measurement made at the time of installation.

CAUTION: Do not attempt to conduct any electrical tests if the oil has been removed
from the transformer. The windings and the associated connections must be under oil
even for low-voltage tests. Failure to observe these precautions may lead to equipment
failure or severe personal injury.

314
 6.2.5 MAINTENANCE OF BUSHINGS
Little maintenance is required other than periodic checking of the oil level as indicated in
the sight glass or by the gage, measuring of the power factor, and checking tile
condition of the top terminals. Bushings exposed to salt spray, cement dust, and other
abnormal deposits are subject to a special hazard and must be cleaned regularly to
prevent flashover and corrosion of parts.

Metal parts are painted for protection against the weather. Care should be taken to
prevent scratching these painted surfaces. The sight glass transparency may become
impaired due to reaction with atmospheric contaminants and should be cleaned
regularly to deter this reaction. In the event the sight glass does become opaque, it
should be replaced in order to maintain good visibility of the oil level.

In the unlikely event it becomes necessary to add oil to a bushing, the fill plug in the
spring assembly or the metal dome can be removed. Insertion of a clean standpipe, with
an outside diameter of slightly less than the diameter of the hole will provide a means of
adding small quantities (two quarts or less) to the bushing. This should return the oil to
the proper level. If not, the bushing must be removed from service and returned for
repair and processing. Follow the procedure outlined below for oil additions in the field.
Obtain the necessary oil from the Components Division Plant, Alamo, TN 38001, or
provide oil that meets the following standard:

Transformer oil processed to have additional requirement of:

1. Moisture content less than 5 ppm

2. Neutralization less than 0.02 mg KOH/g
3. Dielectric breakdown min. 45 kV
4. Power factor less than 0.05%

To prevent oxidation of the bushing oil, the air space above the oil level should be
purged with dry nitrogen and the fill plug replaced immediately afterward.

WARNING:
DO NOT REMOVE THE FILL PLUG WHEN THE BUSHING IS AT AN ELEVATED
TEMPERATURE AS THE OIL INSIDE THE BUSHING MAY BE VERY HOT AND
UNDER HIGH PRESSURE. MAKE SURE THE BUSHING TEMPERATURE IS IN THE
15 TO 35°C RANGE. FAILURE TO FOLLOW THESE GUIDELINES COULD RESULT
IN SEVERE PERSONAL INJURY.

Due to the inconvenience and possible service interruptions resulting from bushing
outages, many users have programs for Planned Preventative Maintenance. We
endorse such programs and recommend:

1. Measurement of power factor and capacitance at the time of installation and

repeating the measurements annually. Field measurements of power factor and
capacitance can differ from measurements made under the controlled conditions in

315
 the factory. Therefore, the power factor and capacitance should be measured at the
time of installation and used as a base to compare with future measurements. You
should contact your ABB Inc. representative for corrective action procedures if:

a. The power factor doubles the original installation value; or

b. The capacitance increases to 110 percent of the original installation value.

The following guidelines may be used to minimize the effect of contamination and
high humidity during power factor and capacitance measurements in the field.

a. Clean the bushing thoroughly with a suitable cleaner to remove any

contaminants that may have deposited on the porcelains and other parts during
shipping or storage. After cleaning, wipe the surface dry to avoid moisture
condensation.
b. Clean and dry the power factor tap insulator to remove any contamination or
condensation.
c. Avoid making power factor measurement in wooden crates to minimize the effect
of surface leakage due to moist wood.
d. Provide sufficient clearance between the bushings and other objects to minimize
the effects of stray capacitance.
e. Do not invert bushing as this may cause entrapment of gas and result in erratic
readings.
f. For information on ground connections and other guidelines, please refer to the
test equipment manufacturer’s Instruction Manual.

2. Examination of the top terminals for loose connection and overheating.

For satisfactory operation of a bushing, it is important that the top terminals are tight
at all times. If any of these parts are loose, overheating of the current-carrying joint
can take place and result in damaged terminal joints. This type of overheating can
deteriorate the bushing gasket seals, which could result in deterioration of the oil-
paper system.

Visually examine the bushings and look for discolored top terminal, external terminal
connector or bolts, and the draw lead cap nut. Look for steam rising from the terminal
during rain. Perform an infrared scan of the top terminals.

If the above examination indicates overheating, remove the transformer from service
and check the power factor and capacitance. Remove the top terminal and examine for
any damage. Examine the gaskets for any sign of hardening. If any signs of overheating
are present then the bushing should be taken out of service as the overheating may
have affected the other sealing gaskets at the top end. If the top terminal cannot be
removed, it has most likely suffered overheating damage at the threaded joint. Remove
the bushing for service.

If the top parts do not show any sign of overheating or damage and the power factor

316
and capacitance values are acceptable, then reinstall these parts by following the
proper instructions. Use new gaskets if they show any sign of damage.

Any repair of high-voltage condenser bushings should be done in the factory because of
the danger of contamination to the insulation if the seal is broken. In addition, the very
high vacuum and clamping pressure require the use of equipment not usually available
in the field.

6.2.6 MAINTENANCE AND SERVICE OF OLTCS/LTCS 29

Table 6-2 provides some general guidelines for maintenance of the various components
in a tap changes. Note that these guidelines do not supersede those provided in the
manufacturers’ maintenance and service manuals.

Table 6-2: General Pointers for LTC Maintenance

LTC Component What to Inspect/Maintain Consequence of Improper/Inadequate
Maintenance
Control Equipment Cam and limit switches Control malfunctions can result in overheating,
Control relays and contactors voltage regulation problems or runaway situations,
Control switches which can cause physical damage to the LTC or
Regulating relays transformer. Clean and inspect wiring, contactors,
Control wiring and terminations auxiliary switches, gauges, indicators, etc.).
Mechanical Components Contact assemblies Excessive wear can lead to a unit failure. Check for
Subject to Wear Drive shafts alignment of mechanism and lubricate mechanism.
Bearings
Gears and sprockets
Cam and toggle assemblies
Mechanism components in the air cabinets
Pins and keys (check for fatigue and
damage)
Other Mechanical Components Liquid level gauge Inadequate maintenance can lead to a failure or
Dehydrating breather lack of protection due to an inoperative device.
Mechanical relief device Moisture in the oil accelerates the formation of
Sudden pressure relay oxide films, which will eventually result in contact
Oil preservation system failure. Moisture in the LTC tank can also lead to
rust damage in the internal tank and mechanism
parts.
Tank Doors Inadequate maintenance can lead to leaks or
Gaskets internal contamination.
Manholes
Oil seals for shaft penetration
Paint finish
Welded seams
Current Carrying and Switching Windings Can overheat due to loose connections, overload or
Components Cables high contact resistance.
Contacts
Resistors
Reactors
Bushings
Insulation Materials Paper These items can track from contamination, break, or
Terminal Boards in the case of terminal boards, leak to or from the
Phase barriers and support structures main tank.
Operating linkages
Operating shafts inter-phase
Oil Carbonization With the oil being in contact with everything in the
Oxidation oil compartment, degradation can accelerate failure
Moisture contamination of all other components. It is important to filter the
Metal arcing by-products oil as part of normal maintenance procedures.

29
See section 1.7.1 for comments on OLTC (IEC designation) or LTC (IEEE designation)

317
 In addition, for OLTC types in which arcing occur in the oil, it is important to thoroughly
clean the component as part of routine maintenance. Attention should be given to
mechanical wear and arcing contact wear. In non-oil-arcing LTCs, it is expected that the
oil will be relatively clean. It is necessary to check the wear gauge on vacuum
interrupters in order to assess contact erosion. The interrupters themselves must be
pull-tested to ensure they are still under vacuum and not filled with oil.
Additionally, the dielectric integrity of vacuum interrupters must be tested with an AC
high potential (HIPOT) tester.

6.2.7 GENERAL QUALITY INFORMATION FOR VARIOUS TYPES OF LTCS

Over the years as ABB service engineers and technicians have worked on LTCs,
several “tricks of the trade” have been compiled. Some of these quality tips are listed in
Table 6-3 to Table 6-10.
6.2.7.1 NORTH-AMERICAN PRACTICES

Table 6-3: Maintenance Quality Tips for GE LTCs

Model Potential Problem Potential Mitigation Action
GE LR300, LR400, LR500 These LTCs are oil-filled under vacuum then This problem can be eliminated by addition
pressurized with dry nitrogen and sealed. of a bi-directional, regulating, dehydrating
It is believed that over time the nitrogen blanket is breather.
absorbed into the oil, resulting in negative
pressure on the tank.
This interferes with vacuum interrupter operation
and results in burning of the main contacts.
GE LR400, LR500, LR700 These LTCs have an interrupter latch that is Use drill bit shanks as GO/NO GO gauges.
adjusted to 0.060" +/- 0.020”. This adjustment is #46 and #60 drill bit shanks work with great
important for good vacuum interrupter tripping results.
action.
The problem is getting the measurement feeler
gauges to fit.
GE LRT-200 – Not oil-tight The instruction book (GEK-29296) notes that the When vacuum processing on the main tank
terminal board between the LTC compartment is being done, the LTC compartment must
and the main tank is not oil tight. be completely sealed.
Vacuum Interrupter Viewing Water collecting between the plate and window Make sure that the gaskets on the cover
Ports can freeze and break the glass. plates over the interrupter wear gauge
This will admit water into the tap changer tank viewing ports are intact.
and contaminate the oil.
GE LoadVac Isolating Warning: The motor drive control circuit provides To insure the safety of maintenance
Transformer - Plugs power to the isolating transformers that are part personnel, the plug that provides entrance
of the vacuum interrupter integrity circuit. into the reactor compartment must be
Even with all transformer bushings isolated and disconnected during most maintenance
grounded, 5 kV can exist at the tap selector. procedures.
See the applicable instruction manual for the
location of the plug.
GE LTC Contact Thimble On most GE LTCs the moving contact fingers For disassembly:
are pressurized by a spiral spring assembly. Rotate the post of the assembly until its pin
A contact thimble is a tool made from steel pipe is vertical.
that aids in the disassembly of the spiral spring Place the contact thimble over the cup
assembly. washer, aligning the slot with the vertical
pin.
Squeeze the tool down over the post until
the pin drops out.
Resetting Lockouts on Lockouts occur for a good reason. Determine why lockouts occur and make
Vacuum Type LTCs Resetting a lockout without investigation is like corrections before resetting.
playing Russian roulette.
Vacuum LTCs: Don't Break Pulling vacuum on the transformer main tank will Always disconnect transformer bushings
Your Bushings deflect the cover and tank walls and cause from external leads and bus prior to vacuum

318
Model Potential Problem Potential Mitigation Action
displacement at the bushing terminals. operations.
Failure to disconnect the external bushing
terminations could result in a strain on the
bushing, causing physical damage or breaking a
seal.
Vacuum Interrupters If you ever disassemble a vacuum tap changer, When the unit is reassembled, you must use
remember that you must document the existing this information to recalibrate the wear
wear on the interrupter if you intend to reuse it. indicator.
Failure to do so may ultimately lead to failure of
the interrupter because of improper calibration.

Table 6-4: Maintenance Quality Tips for Westinghouse LTCs

Model Potential Problem Potential Mitigation Action
On the UTT (not UTT-A or UTT-B), the air Check your UTTs and replace the gasket
Westinghouse Type UTT : Air
compartment that houses the cam switch before it leaks.
Compartments assembly has a small plate on top of it. The new gasket is 1/8" cork-neoprene approx.
The gasket under this plate may deteriorate with 6X6". ABB S#4155D20H05.
age and eventually leak.
A leak at this location will cause considerable
damage to the cam switch assembly.
This damage is not evident to a careful visual
inspection from the ground until it is too late.
These units have black molded switches that Replacement molded switches are no longer
Westinghouse UTT : Cam
deteriorate with time. They can bind up, breaking available. Therefore, if this problem develops,
Switches (Prior to UTT-A70) parts coupling the cam switch to the tap changer. UTT-B style switches can be adapted for use
The motor then runs until a mechanical limit is on these tap changers.
reached.
This often results in expensive damage to the tap
changer motor mechanism and/or reversing
assembly.
Tracking has caused problems on the Type UTT In addition to flushing the LTC compartment,
Westinghouse Type UTT :
terminal boards. make sure you wipe the boards to remove all
Terminal Board Tracking This problem has been attributed to inadequate carbon deposits.
cleaning during maintenance cycles. Inspect the boards carefully for any signs that
In early stages, carbon tracks can be cleaned with tracking has begun.
sandpaper. Proper maintenance will minimize the
If the tracking is allowed to progress unchecked, a possibility of tracking on the Type UTT
flashover could result. terminal boards.
Westinghouse Type UTT: Many UTT models previous to the UTT-B model UTT-B contacts were developed to eliminate
Conversion Kits have experienced coking failures. this hazard and are available for all earlier
models.
The balance coils on W estinghouse Type UTH If the coils have crimped terminals, check the
Westinghouse Type UTH:
LTCs have been manufactured with terminals terminations very carefully to ensure that they
Balance Coil Leads either crimped or brazed to the leads. There is are not loose.
potential for the crimped leads to loosen. The coils are located on the LTC transfer
switches.
Westinghouse UVT The dehydrating breather on the W estinghouse Replacing the plastic vent with a brass filter,
Type UVT LTC has a plastic vent plug with a wire ABB part# 72A4551H01, will eliminate this
that when improperly maintained can prevent the potential problem.
breather from operating.
The faster operating LTCs (UVT , UVW ) have very To set the more complicated sequences on
Westinghouse UVT , UVW
these tap changers, timing wheels are
critical timing sequences.
required.
Westinghouse Type UVT: Certain vacuum LTCs require vacuum filling to If the main tank is oil-filled, do not tie the LTC
GE LoadVac prevent damage to the vacuum interrupter bellows tank to the main tank and pull vacuum over
due to trapped air. the main tank oil.
The terminal boards will withstand full vacuum
if the following conditions are adhered to:
The board temperature must be less than
50°C.
The pressure in the maim tank gas space
must be 0 psi.

319
 Model Potential Problem Potential Mitigation Action
Westinghouse URS: Drive With age these motors appear to lose power and New URS motors are no longer available at
Motor the LTC will not operate properly. any price.
Motor rewind shops often find nothing wrong with We recommend calling other users looking for
the motor. a used motor that is useable or at least can
In fact the rotor running bar(s) are open at the be rebuilt.
shorting ring and restrict the development of
torque.
Westinghouse and ABB "17” link type switches in ”shorted” position on The switches must be left in the open position
Controls LTC controls means non-functioning circuits. for normal operation.
Many W estinghouse and ABB LTC controls are Be sure the “17” switches are open except for
equipped with one or more "17" sliding link type certain servicing operations.
shorting switches. These are shipped in the
shorted position for safety reasons.

Table 6-5: Maintenance Quality Tips for ASEA and Reinhausen Resistive Type LTCs
Model Potential Problem Potential Mitigation Action
Reinhausen Suction Tubes Entrapment of air in high-stress areas can result The "S' pipe is provided with a bleed screw for
in a failure. venting.
Commonly missed are the suction tubes in
Reinhausen tap changers.
This tube connects the "S" pipe to the bottom of
the diverter tank for draining.
ASEA & Reinhausen Tighten the kerosene plug before the main tank
Resistance Type LTC is sealed. DO NOT over-tighten, or you will
crash the gasket and it will leak.
Make sure the valve between the diverter
compartment and main tank is open during
vacuum processing and closed in service.

Table 6-6: Maintenance Quality Tips for Moloney LTCs

Model Potential Problem Potential Mitigation Action
Moloney Type MA & MB The tap changer contacts must move in a quick A little jerk is the desired adjustment for the
jerk when they operate. spring mechanism.
This is often not the case because of a broken
spring at the accelerator drum.
A broken spring is not always obvious and will
result in rapid erosion of the arcing contacts.
Over-tightening of these spring mechanisms will
always result in a broken spring.

Table 6-7: General Maintenance Quality Tips for LTCs

Idem Potential Problem Potential Mitigation Action
Dehydrating Breathers with Some of the dehydrating breathers applied to A malfunctioning regulator can result in excessive
Regulators LTCs have in-and-out breathing regulators. pressure (positive or negative) within the LTC
These regulators can become inoperative due to tank.
a number of reasons. During LTC maintenance it is good practice to
verify proper operation of these regulators.
Hand Crank Through Every Local maintenance procedures often dictate The procedure should include a requirement to
Position hand cranking an LTC through a few positions hand crank the tap changer through all positions
above and a few positions below neutral. raise and lower
Never operate an LTC either electrically or This gives you a feel for the unit through the
manually when the tank is under vacuum. entire range of operation. Binding, misalignment,
Deflection of the tank can cause misalignment, or other problems will be easier to identify.
and operation of the unit could result in
mechanical damage.
Reversing Switch Contacts LTCs often operate within a range of taps that Periodically exercising the tap changer through
does not change the position of the reversing all taps will break up any film or other
switch. accumulated contaminants in the vicinity of the
If the reversing switch remains on one position contacts.
over an extended period of time, the possibility
increases for coking to occur.

320
Idem Potential Problem Potential Mitigation Action
This can result in overheating of the reversing
switch contacts and, if not corrected in time,
failure of the tap changer.
Oil Processing Oil processing may seem like the final Follow manufacturers’ instructions.
unimportant step. If the oil is not filtered properly this to can create
All is lost if the unit fails because air or moisture problems and affect the time the unit can operate
is in the wrong spot. until the next servicing.
Oil Sampling Safety Introduction of air bubbles through the sampling NEVER take oil samples with negative pressure
value can cause a failure resulting in equipment on the tank.
damage, severe personal injury, or death.
Arc-in-Oil Transfer Switch Arc-in-oil transfer switch compartments are Before starting to wipe out, flush, and clean this
Compartments expected to be found in very dirty conditions. compartment, do a very careful visual inspection.
Often, broken parts, water, or other evidence of Starting the cleaning process too soon will
trouble will be seen on the tank bottom. destroy this valuable evidence.

Gaskets Properly applied gaskets provide better seal Always use the correct gasket for the application
and maintain the proper compression.
Cork-Neoprene is generally used for permanent
joints, such as air-filled cabinets and installment
mountings. Proper compression is 43%.
Nitrile rubber is generally used where reusable
gaskets are required, such as inspection covers,
bushing flanges, etc. Proper compression is 33%
% Compression = 100*(Gasket Thickness –
Compressed Thickness)/Gasket Thickness
Silicone Grease A lubricant is often used on gaskets and seals. Do not use silicone grease on load tap changers.
Silicone grease is a good insulator. But if the Put gaskets on dry or use petrolatum.
material migrates to the LTC contacts, high
contact resistance and overheating could result.
Transformer Openings All transformer openings, such as upper filter It is recommended to use Teflon paste for all
press/filling openings must be sealed at all times fittings.
to prevent introduction of water into the Check the integrity of the tank after sealing.
transformer tank. Pressurize the tank to 3 to 4 psig with dry
nitrogen.
Allow the pressurized tank to stand for one to two
hours.
Examine the tank and fittings for leaks by
applying a soap solution to all joints and pipe
fittings.
After determining that there are no leaks reduce
the internal pressure to normal start-up limits.
Terminal Board The design withstand pressure of the board can Before applying vacuum to any LTC with the
be different between different types of LTCs. main tank at different pressures, make sure that
The design withstand pressure of the board can the terminal board is designed to withstand the
be different between LTCs of the same type. pressure applied.
If there are any questions, check with the
manufacturer.

321
 6.2.7.2 EUROPEAN PRACTICE

Table 6-8: Maintenance Quality Tips for ASEA and Reinhausen Resistive Type OLTCs
Model Potential Problem Potential Mitigation Action
ABB diverter switch type UC Correct diverter switch insert. After the diverter switch is placed in the housing
the OLTC should be maneuvered 3 times before
the cover is put on.
ASEA & Reinhausen Tighten the kerosene plug before the main tank
Resistance Type LTC is sealed. DO NOT over-tighten, or you will
crash the gasket and it will leak.
Make sure the valve between the diverter
compartment and main tank is open during
vacuum processing and closed in service.
Reinhausen Suction Tubes Entrapment of air in high-stress areas can result The "S' pipe is provided with a bleed screw for
in a failure. venting.
Commonly missed are the suction tubes in
Reinhausen tap changers.
This tube connects the "S" pipe to the bottom of
the diverter tank for draining.

Table 6-9: General Maintenance Quality Tips for OLTCs

Part Potential Problem Potential Mitigation Action
Pressure relay or oil flow A trip from the pressure relay or the oil flow Open the tap-changer compartment were the
indicator indicator can be caused of a severe fault braking takes place for inspection before the tap-
changer is put in service again
Dehydrating Breathers with Some of the dehydrating breathers applied to A malfunctioning regulator can result in excessive
Regulators OLTCs have in-and-out breathing regulators. pressure (positive or negative) within the OLTC
These regulators can become inoperative due to tank.
a number of reasons. During OLTC maintenance, it is good practice to
verify proper operation of these regulators.
Hand Crank Through Every Local maintenance procedures often dictate hand The procedure should include a requirement to
Position cranking an OLTC through a few positions above hand crank the tap changer through all positions
and a few positions below neutral. raise and lower
Never operate an OLTC either electrically or This gives you a feel for the unit through the
manually when the tank is under vacuum. entire range of operation. Binding, misalignment,
Deflection of the tank can cause misalignment, or other problems will be easier to identify.
and operation of the unit could result in
mechanical damage.
Reversing Switch Contacts OLTCs often operate within a range of taps that Periodically exercising the tap changer through
does not change the position of the reversing all taps will break up any film or other
switch. accumulated contaminants in the vicinity of the
If the reversing switch remains on one position contacts.
over an extended period of time, the possibility
increases for coking to occur.
This can result in overheating of the reversing
switch contacts and, if not corrected in time,
failure of the tap changer.
Oil Processing Oil processing may seem like the final Follow manufacturers’ instructions.
unimportant step. If the oil is not filtered properly this to can create
All is lost if the unit fails because air or moisture problems and affect the time the unit can operate
is in the wrong spot. until the next servicing.
Dielectric withstand Dirty oil that contains moisture will have a Check the dielectric strength of the oil. This test
reduced dielectric withstand is much better then only checking the moisture in
oil content.
Oil Sampling Safety Introduction of air bubbles through the sampling NEVER take oil samples with negative pressure
value can cause a failure resulting in equipment on the tank.
damage, severe personal injury, or death.
Breaking Compartments Breaking compartments are expected to be Before starting to wipe out, flush, and clean this
found in very dirty conditions. compartment, do a very careful visual inspection.

322
Part Potential Problem Potential Mitigation Action
Often, broken parts, water, or other evidence of Starting the cleaning process too soon will
trouble will be seen on the tank bottom. destroy this valuable evidence.

Oil filters Make sure that the oil filter operates in the A wrong operation direction can cause air
correct direction bubbles to enter the bottom in the tap-changer
It is important that all joints are tight in the oil and create failures
filter circuit. A leak in the oil circuit can cause that all oil is
drained from the tap-changer resulting in a
failure.
Gaskets Properly applied gaskets provide better seal Always use the correct gasket for the application
and maintain the proper compression.
Cork-Neoprene is generally used for permanent
joints, such as air-filled cabinets and installment
mountings. Proper compression is 43%.
Nitrile rubber is generally used where reusable
gaskets are required, such as inspection covers,
bushing flanges, etc. Proper compression is 33%
% Compression = 100*(Gasket Thickness –
Compressed Thickness)/Gasket Thickness
Silicone Grease A lubricant is often used on gaskets and seals. Do not use silicone grease on load tap changers.
Silicone grease is a good insulator. But if the Put gaskets on dry or use petrolatum.
material migrates to the OLTC contacts, high
contact resistance and overheating could result.
Transformer Openings All transformer openings, such as upper filter It is recommended to use Teflon paste for all
press/filling openings must be sealed at all times fittings.
to prevent introduction of water into the Check the integrity of the tank after sealing.
transformer tank. Pressurize the tank to 3 to 4 psig with dry
nitrogen.
Allow the pressurized tank to stand for one to two
hours.
Examine the tank and fittings for leaks by
applying a soap solution to all joints and pipe
fittings.
After determining that there are no leaks reduce
the internal pressure to normal start-up limits.
Terminal Board The design withstand pressure of the board can Before applying vacuum to any OLTC with the
be different between different types of OLTCs. main tank at different pressures, make sure that
The design withstand pressure of the board can the terminal board is designed to withstand the
be different between OLTCs of the same type. pressure applied.
If there are any questions, check with the
manufacturer.

Table 6-10: Maintenance Quality Tips for Vacuum OLTCs

Model Potential Problem Potential Mitigation Action
Resetting Lockouts on Lockouts occur for a good reason. Determine why lockouts occur and make
Vacuum Type OLTCs Resetting a lockout without investigation is like corrections before resetting.
playing Russian roulette.
Vacuum OLTCs: Don't Break Pulling vacuum on the transformer main tank will Always disconnect transformer bushings
Your Bushings deflect the cover and tank walls and cause from external leads and bus prior to vacuum
displacement at the bushing terminals. operations.
Failure to disconnect the external bushing
terminations could result in a strain on the
bushing, causing physical damage or breaking a
seal.
Vacuum Interrupters If you ever disassemble a vacuum tap changer, When the unit is reassembled, you must use
remember that you must document the existing this information to recalibrate the wear
wear on the interrupter if you intend to reuse it. indicator.
Failure to do so may ultimately lead to failure of
the interrupter because of improper calibration.

323
7 REPAIR, REFURBISHMENT AND RETROFIT
Power transformers are the most important components of electrical energy supply
networks. Although these components have high reliability, failures can occur due to
external factors such as short circuits in the grid. Failures can also occur due to ageing
effects of the active part, especially the insulation system, or ageing of important
accessories such as bushings or on load tap changers. .

A theoretical failure statistic versus age of various types of transformers is shown in

Figure 7-1. Figure 7-2 shows the contribution of various components in/on a transformer
to the failure rate

Failure rate (%) Probability of Failure

100%
90% Generator transformer
80% Net transformer
70%
Industrial transformer
60%
50%
40%
30%
20%
10%
0% Years

0 20 40 60 80

Figure 7-1: Bath-tub failure rate curve

Figure 7-2: Occurrence of failures in different parts of power transformers

In case of failure, the first action is to identify the type of failure. The aim is to localize
the failure itself as well as its root cause. In order to fully understand the reason for the
defect, it is usually necessary to investigate if any abnormalities had occurred in the grid
prior to the failure.

The investigation into the failure location inside the transformer is performed using
different methods of condition assessment. These have been explained in greater
detail in other sections of this handbook.

324
After performing different measurements at site, it is necessary to decide whether the
repair should be performed in a repair factory or at site. In cases with minor damage,
especially in case of a failure related to connections, bushings or tap changers, the
repair should be done at site. In cases with more extensive damage or when windings
are involved, several criteria need to be considered. These include the practicality of
transporting the unit, the transportation time and risk, the consequences of non-delivery
of energy during the repair period and redundancy of the failed units i.e. availability of a
spare unit.

Several years ago, ABB started to develop processes to perform major repairs,
including replacement of windings and repair of the core at site. To date a total of about
200 units, including Utility, Industrial and HVDC transformers and reactors have been
repaired. In many of these cases the repair included an upgrade of the unit to a higher
rating. The trademarked name for the ABB solution for repair at site is
TrafoSiteRepairTM and is marketed globally.

TrafoSiteRepairTM is often the preferred alternative if the distance to a repair factory is

considerable. In such cases, the total time that the transformer is out of service can be
reduced by eliminating the long and risky transport of the transformer to a repair facility.
Additionally, TrafoSiteRepairTM could be the preferred method of repair because of
difficulties in acquiring transportation authorizations by road or rail.

It needs to be mentioned that a refurbishment or “preventive” repair can be beneficial if

no failure has occurred, in cases where the ageing of the active part has reached a
stage where the continued reliability of the transformer is unacceptable.

The TrafoSiteRepairTM process is very similar to that in a repair factory. Therefore, all
critical elements of site repair have the same quality requirements as for factory repair
and should be performed with the same methods and tools where feasible.

To perform the different phases of repair operations at site, ABB sends experienced and
skilled personnel to the site. The staffs at the site have the same skills as factory staffs
performing the same operations plus the additional experience of performing
sophisticated operations in the field.

7.1 PREPARATION PHASE

Before the repair can be performed, components of the transformer such as bushings,
etc. should first be removed.

Factory repair: The transformer has to be prepared for transportation to a repair

shop, where dismantling of the active part will be performed. If necessary, an
additional high voltage test is performed before untanking the active part in order
to investigate the failure and gain more information on the nature of the damage
as well as its location.

325
 Site repair: A site repair usually does not require extensive transportation of the
transformer. If a suitable workshop is within close proximity of the transformer it
may be used for the repair. If not, a temporary workshop is needed and the
transportation path for the transformer to the temporary workshop has to be
planned.

The method for hauling the transformer to the temporary workshop is chosen
based on available means at site and the location of the workshop. If a spare
foundation may be used as a base for the workshop, means for hauling the
transformer may already be available.

If a temporary workshop is set up, all required tools and equipment will need to
be shipped to the site to be available in time for the different phases of the repair.
A more detailed description of how a temporary workshop is set up is described
below in the section “Facilities for site repair”.

7.2 UNTANKING AND DISASSEMBLY OF ACTIVE PART

After draining the oil, the active part can be untanked and disassembled. If required, a
detailed inspection of the active part can be performed.

Factory repair: In a factory repair, the active part is removed from the tank using
a fixed overhead crane with a high lifting capacity.

Figure 7-3: 500 ton crane inside a workshop

326
 Site repair: If an overhead crane is not available or if the crane does not have
enough lifting capacity, special lifting equipment is brought to the site to untank,
disassemble and reassemble the active part.

7.3 REPAIR OF THE TRANSFORMER

After the final decision on the scope of the repair, all required materials and components
are ordered from qualified suppliers. While waiting for these supplies the work proceeds
to remove all damaged parts. It might be necessary to remove all windings from the
core after the upper core yoke is removed. If the winding is to be replaced, the new
windings are manufactured in a transformer factory according to the design specification
for the transformer.

Figure 7-4: Vertical and horizontal winding machine

Factory repair: For a factory repair, the delivered parts are installed upon
receipt. In case of a winding replacement, the windings are dried and assembled
on the core using the fixed crane. If repair of the core is required, the core is
placed in a horizontal position on a core stacking table. This allows the repair or
exchange of core sheets to be performed.

Site repair: For a repair based on the TrafoSiteRepairTM,concept all windings

and other components are supplied from one of the ABB transformer factories.
Before shipment to site, windings and insulation components are dried using a
vapor-phase process. During transportation, the windings and insulation
components are stored in enclosures supplied with dry air, so they are ready for
installation at site. To keep the insulation dry during the repair and assembly of
the active part, the windings and insulation parts are sealed and supplied with a
continuous flow of dry air. If repair of the core requires the core to be placed in a

327
 horizontal position, a mobile core stacking table is brought to the site. Mobile
lifting equipment is used to tilt the core stacking table.

7.4 ASSEMBLY AND TANKING OF THE ACTIVE PART

The assembly and tanking of the active part is performed in the same way for both
factory repair and site repair. However, if an overhead crane is not available during site
repair, special lifting equipment is brought to the site. After the assembly of the active
part, various low voltage tests on the active part are performed in order to guarantee
quality.

7.5 DRYING
Factory repair: After assembling, the whole active part is dried in a vapor phase
system, in which vacuum is applied with a certain time and temperature profile. In
addition, the active part is surrounded by kerosene vapor, which is why this
process is called vapor phase drying.

Figure 7-5: Vapor phase chamber

Site repair: After the tank cover has been replaced and sealed, the transformer
is put under a low overpressure of dry air to avoid any possible contamination
(mainly moisture) until the drying phase begins. There are different drying
methods that may be used for drying the transformer at site:

o The most common method is the hot oil circulation and vacuum process. It
is a proven method that provides a good quality of drying within a
reasonable processing time (typically between two to 4 weeks). The drying
time can be reduced by including a hot oil spray system in the process. In
both cases, the transformer tank must be “vacuum tight”. If this is not the

328
 case, then the quality of the drying (remaining moisture in the solid
insulation) will be limited and the required drying time will be significantly
extended.
o The most efficient method in respect of drying time and quality of drying is
the Low Frequency Heating method (LFH) in combination with hot oil
spray. The windings are heated by driving a low frequency current through
the high voltage winding of the transformer while the low voltage winding
is short circuited. Simultaneously hot oil is sprayed on the outside of the
winding and on the insulation parts. The drying is performed under periods
of heating and different levels of vacuum to achieve the most efficient
drying. The advantage of this process is that a Low Frequency (LF)
current heats the transformer coils from the inside, while the hot oil
simultaneously heats the outside insulation. Since the windings and the
insulation parts were dried using the vapor phase system and handled
under dry air during the storage and installation periods, the moisture
content should be quite low. The total drying time is typically one to a
maximum of two weeks.
o If an LFH plant is not available, simple hot oil spray or hot oil circulation
may be used. However, these processes may considerably extend the
drying period.

7.6 FINAL ASSEMBLY

The final assembly of the transformer in the field uses the same process as in the
factory. If the high voltage test performed after the final assembly is successful, the
transformer does not need to be disassembled again for transportation since it is
already at the site. The welding of the cover is, however, normally done after the test.

If the repair was performed in the factory then the transformer is typically disassembled
and prepared for shipping. This may involve the removal of the bushings and other
accessories and draining all the oil from the tank. The process is the same as is used
for the shipment of new transformers from the factory.

7.7 HIGH VOLTAGE TESTING [165]

As a final quality control the last step of a repair is the high voltage test, in order to
prove that the insulation can be stressed with voltages that are higher than in operation.
The tests are performed according to IEC or ANSI-Standards, in which the test
procedures and test levels are described.
Factory repair:
A repair factory is equipped with a high voltage test field for performing mainly the
following most important high voltage tests:
- Applied voltage test
- Induced voltage test with partial discharge measurements
- Lightning impulse test
- Determination of no load losses
- Determination of load losses

329
Site repair:
After the site repair the most important high voltage (HV) tests are performed using
a mobile test field.

The system is installed in a 40 feet container and allows performance of the

following high voltage tests for transformers up to 600 MVA:
- Applied voltage test
- Induced voltage test with partial discharge measurements
- Determination of no load losses (depending on the transformer
characteristics)
Additionally, impulse tests and determination of load losses can be performed using
modular extensions in additional containers, but this has not been used to date
because of the high costs involved.

Further, dissolved gases in oil are monitored during the first year after repair.

The testing process described above has proven to be an effective way of testing
transformers repaired at site in order to ensure a high standard of quality. Over the last
fifteen years, ABB has successfully repaired more than 200 transformers at site without
any subsequent failure in operation.

7.8 QUALITY PLAN [166]

The repair processes in both workshop and at site follow documented procedures that
are described within ABB’s quality system. This is adapted from the quality process
used for new transformers manufactured by ABB.

7.9 FACILITIES FOR SITE REPAIR

A site repair of a power transformer based on the TrafoSiteRepairTM concept should be
performed indoors, in a facility that makes it possible to maintain a controlled
environment with the required levels of cleanliness and orderliness. The facility should,
as far as possible, allow for performance of all critical steps of the repair inside the
facility. This includes the heavy lifting of the active part and windings for core type
transformers. For very large shell type transformers, it may be difficult to handle the
lifting of the phases without opening the roof of the repair facility. For such cases, the
timing of those activities will be dependant upon the weather conditions. In addition,
special measures have to be taken to protect the phases from moisture and
contamination.

Based on experience gained within the ABB service centers that have performed site
repair projects, it is estimated that for approximately 50 % of repair projects a facility is
available at the repair site, normally a maintenance shop owned by the customer. For
the remaining 50 % of the projects it is necessary to set up a temporary facility. When a
permanent facility is available at site, the repair area has to be separated from the rest
of the facility in order to maintain cleanliness requirements.

330
To keep the repair time as short as possible the set up time for the facility should be
minimal.

7.9.1 TEMPORARY WORKSHOPS

If no suitable building is available at the site and a temporary workshop is required,
there are various alternatives available for construction of a temporary facility. ABB has
mainly used two alternatives.
7.9.1.1 STEEL BUILDINGS
Large halls composed of a steel structure and a cladding of corrugated sheet steel or
aluminum can be constructed. This type of building is primarily used when more than
one transformer will be repaired at the same site. The advantage with this type of
building is that it can be made to fit exactly the need and built by local constructing
companies practically anywhere in the world. It can also be designed to withstand
severe weather conditions. In some cases the steel buildings used for site repair have
been left at the site for future use.

Figure 7-6: A steel building used as temporary workshop by ABB in Brazil

7.9.1.2 LARGE T ENTS

Large rigid tents consisting of a steel structure and claddings of a flexible sheet material
such as PVC are available to buy or rent through both global and local suppliers. These
types of tents may be set up in a very short time, normally less than a week. They are
also designed to withstand severe weather conditions such as strong winds and snow
load.

Figure 7-7: Typical tent structure that may be used as temporary workshop

331
7.9.1.3 FOUNDATION FOR A T EMPORARY WORKSHOP
A temporary workshop needs a foundation that is designed to handle the weight of the
transformer and built according to the local regulations. In some cases there may be a
spare transformer foundation available to build the facility on. The spare transformer
foundation may also have a system to handle the oil. However, it is recommended that
the oil be drained from the transformer before it is moved to the repair facility. An
advantage of using a spare transformer foundation is that it may already have the
means for hauling the transformer.

7.9.2 FACILITIES FOR HEAVY LIFTING

The major heavy lifting during repair of a core type transformer is the lifting of the active
part during untanking and tanking of the transformer. The active part of the largest
transformers can weigh up to 400 metric tons. In the factory an overhead crane is used
to do the lifting. At a site, where an overhead crane is not available or does not have the
required capacity, mobile lifting systems are available from global suppliers. For smaller
transformers, the lifting may be performed with a mobile crane.

The lifting of windings required for disassembly and reassembly of the active part can
usually be performed using a mobile crane.

Figure 7-8: Active part assembled by using a mobile crane for lifting of winding

7.9.3 MOISTURE CONTROL

A dry air generation system is normally installed at the temporary workshop to keep the
insulation dry during assembly of the active part. It also guarantees a permanent over-
pressure in the tent to avoid any possible contamination due to either moisture or dust.
In addition, the windings and all insulation parts are sealed from the ambient air and
continuously supplied with dry air.

332
 7.9.4 OIL PROCESSING
Required oil processing equipment including tanks is brought to the site.

7.9.5 DRYING EQUIPMENT

As described above, different processes for drying may be used. The required
equipment for the chosen process is shipped to the site in time for the drying phase of
the repair.

7.9.6 HIGH VOLTAGE TEST EQUIPMENT

ABB uses a mobile high voltage test field based on state of the art technology (see
Figure 7-9), making it possible to perform high voltage tests at site for practically all
types and sizes of transformer. The system used is the worldwide first mobile HV test
system for transformers based on a frequency converter technique in contrast to
conventional motor-generator sets, which are quite heavy and therefore not easily
transportable. Motor-generator sets need extensive maintenance and are usually less
robust than frequency converters based on semi-conductor techniques. The frequency
converter system also provides more flexibility for testing at different voltages and
frequency ranges. The mobile high voltage test field is shipped to the site during the
testing phase.

Figure 7-9: The high voltage mobile test system upon arrival at a customer site

333
8 ENVIRONMENTAL ASPECTS [167]
Environmental consciousness has increased manifold in today’s society. Meeting the
environmental requirements in the design of new substations and in the renovation and
maintenance of existing substations is an increasing challenge for the electricity
industry. Successful resolution of community acceptance issues and compliance with
environmental regulations may become major milestones in the process of erecting new
or renovating existing substations. In their increasing efforts to satisfy customers and to
gain community acceptance for their facilities, utilities have employed various
environmental management systems to address these important issues. They have also
employed impact evaluation studies to assess the impact of substations on the
environment.

Most efforts to date have focused on the environmental impact of new substations.
However, existing substations have had and may continue to have a significant impact
on the environment, if they are not adequately addressed. Furthermore, the impact of
existing substations on the environment may be growing due to changes in the
surrounding land use, operation of ageing equipment, changes in environmental laws
and regulations, and changes in the community perception of substations. Sometimes
the environmental aspects are directly related to the safety of people, mainly in the case
of fires that produce fumes and toxic gases.

8.1 CONTAMINATION OF OILS WITH PCB (POLYCHLORINATED

BIPHENYLS)30
8.1.1 GENERAL
The following trade names have been used to identify fluids (askarels) that contain
polychlorinated biphenyls (PCB): Aroclor, Chlophen, Pyralene, etc. They are a group of
synthetic, fire-resistant, chlorinated aromatic hydrocarbons used as electrical insulating
liquids. Under arcing conditions in these fluids, any gases produced consist mostly of
non-combustible hydrogen chloride with lesser amounts of combustible gases.

Transformers containing PCB contaminated oils shall not be considered waste whilst in
service. Should the oil became accidentally contaminated, there are several processes
and techniques available for both on-site and off-site decontamination of PCB
contaminated oils. These processes are based on chemical reactions between PCBs
and a reagent to remove the chlorine present. All PCB decontamination methods, either
off-site or on-site, shall be applied by skilled companies complying fully with local
regulations. Off-site decontamination techniques are limited because of concerns over
the safe transportation of contaminated equipment and liquid to an authorized oil
processing facility. Moreover, such processes are subject to local regulations.

30
IEC 60422-2005 - Mineral insulating oils in electrical equipment – Supervision and maintenance guidance

334
Warning
In some countries, the risk of presence of high concentrations of PCB in transformer oils
still exists. Therefore, it is necessary to verify the concentration of PCB in all apparatus
before treating the oil with an oil treatment machine which may be used for the
treatment of oil of new transformers. Note that by legislation, new transformer oil is
supposed to be free of PCBs.

8.1.2 DEHALOGENATION PROCESSES USING SODIUM AND LITHIUM DERIVATIVES

These processes are typically applied in batch and use reagents based on metallic
sodium, sodium hydride, lithium hydride and additives, for the dehalogenation of PCB in
the oil. This type of process is typically run under pressure and medium to high
temperature (150 °C – 300 °C). This temperature is higher than the flash point of the oil
(140 °C – 150 °C) and therefore introduces safety risks.

WARNING
Proper measures shall be taken to minimize the risk of fire or explosion,
especially in the presence of wet oil.

8.1.3 DEHALOGENATION PROCESSES USING POLYETHYLENEGLYCOL AND POTASSIUM

HYDROXIDE
This process, developed to overcome the problems associated with metallic sodium,
uses a liquid reagent based on polyethyleneglycol (PEG) and an alkaline metal
hydroxide such as potassium hydroxide (KOH). This type of process, which is run at
temperatures of 130 °C – 150 °C, has a limited efficiency on some types of
contaminants (e.g. Aroclor 1242).

8.1.4 DEHALOGENATION IN CONTINUOUS MODE BY CLOSED CIRCUIT PROCESS

This process uses a solid reagent consisting of a high molecular weight glycol mixture
(a mixture of bases and a radical promoter or other catalyst) for chemical conversion of
organic chlorine to inert salts, on a support made of high surface area particulate
matter. This process normally runs at 80 °C – 100 °C and has the capability to
decontaminate equipment on-site through continuous circulation of the oil in a closed
system (without draining the oil or using auxiliary tanks). This uses the solvent capability
of the oil for continuous extraction of PCBs from solid materials inside the equipment.

8.2 ELECTROMAGNETIC COMPATIBILITY (EMC)

8.2.1 INTRODUCTION
Electromagnetic fields (EMF) are an important issue to consider when assessing the
environmental impact of existing transformers in installations or substations.
Regulations and guidelines concerning EMF levels may exist, mainly for new
installations. However, when installations or substations are extended or modernized,
regulations regarding EMF levels would then come into consideration. Typical sources
of electric and magnetic fields in installation and substations include: transmission and

335
distribution lines entering and exiting the substation, buswork, transformers, air-core
reactors and switchgear.

8.2.2 METHODS TO REDUCE EMF LEVELS IN EXISTING SUBSTATIONS

For an existing transformer, when EMF levels need to be addressed, it is usually
recommended that measurements be taken of both electrical and magnetic fields.
However, in both enclosed and open-air substations, the highest magnetic field levels
can be found directly underneath or above the incoming and outgoing overhead or
underground lines.

Although the reduction in EMF levels within existing substations cannot be done without
costly intrusive and extensive modifications, there are common techniques in practice.
Reductions in electric field levels can be accomplished by:
Increasing the height of the buses
Decreasing the phase spacing and bus diameter
Lowering the operating system voltage, and
Making use of vegetation as shielding

Methods for reducing magnetic field levels include:

Increasing the distance from the sources
Optimizing phase configuration to achieve magnetic field cancellation
Minimizing currents by increasing operating system voltages
Minimizing power transfers, and
Providing reactive load transfers and/or alternate line power flow paths

In addition, magnetic field levels can be further reduced by balancing currents on lines
and by shielding conductors and buses.

The health effects of electromagnetic fields have been studied extensively over the last
40 years. The results of these studies have been inconclusive and there is no general
consensus on the possible adverse health effects of these fields. In spite of this, the
electrical utility industry has taken a decidedly proactive stance and is committed to
reducing public exposure to EMF. This applies mainly to the area of distribution
transformers.

8.3 AUDIBLE NOISE

8.3.1 INTRODUCTION
There are two major sources of noise in substations: the continuous noise generated by
the operation of power transformers and reactors and the momentary noise produced
by the operation of high voltage circuit breakers or load interrupters. Other noise
sources in substations include Corona discharges and arcing during operation of
switches.

By far the most important source of noise is that generated by power transformers and
reactors. These pieces of equipment generate a continuous humming noise that might
be disturbing for communities living near the substation. Expansion of urban and

336
suburban areas in the last couple of decades has resulted in some of these substations
being located within and in direct proximity to residential areas. In these new situations,
the noise level generated by the equipment in the substation might not be found
acceptable and corrective measures are often required to reduce the level of noise to
acceptable levels. In addition, public concern regarding industrial noise has increased
over the past few decades and new, more stringent regulations and bylaws have been
introduced to limit noise levels in residential communities.

8.3.2 BACKGROUND

8.3.2.1 CHARACTERISTICS OF T RANSFORMER NOISE

The primary source of noise from transformers is due to magnetostriction of the iron
core. A secondary but much lower source arises from the electromagnetic forces
between the individual turns of the windings. The principal frequency of the resulting
vibration is twice that of the supply frequency (100/120 Hz) and, because the
magnetostriction characteristic of iron is nonlinear, harmonics (200/240, 300/360 and
400/480 Hz) are also generated. The harmonic content of the noise plays a major role in
contributing to the annoyance of the noise as perceived by individuals. The flux density
is controlled by the magnetizing current and the total noise output is proportional to the
exciting voltage times the magnetizing current. Consequently, the noise output remains
essentially constant for a given voltage even though unpredictable variations in the
radiation pattern occur with time. The noise output is normally unaffected by load.
8.3.2.2 PROPAGATION OF SOUND
In an unobstructed outdoor environment, the energy from a point source of sound
propagates as according to the "inverse-square law" - this means that as the sound
spreads outwards from the source, the energy decreases as the square of the distance.
Thus for each doubling of the distance, the sound energy decreases by a factor 4, i.e. 6
dB. This theoretical attenuation usually applies quite well for distances up to
approximately 150 m. Beyond this distance it is subject to the effects of ground and
atmospheric absorption, inhomogenities associated with turbulence, moderate and high
winds and temperature gradients.

In addition, the increase in noise level propagated due to temperature inversion seldom
remains constant. It normally varies with time over periods ranging from a few seconds
to a few minutes due to changing conditions in the atmosphere. This modulation of the
noise, particularly of the type generated by transformers, is another factor that increases
the subjective reaction as compared to that from a steady noise source.

8.3.3 CRITERIA FOR COMMUNITY NOISE LEVELS

There are a wide variety of regulations and bylaws in use in different corners of the
world to control for audible noise levels in the community. Some of the larger cities have
quantitative regulations of varying degrees of complexity. Generally, transformer noise
should be slightly audible during the quietest period of the day and inaudible during the
rest of the day when people are normally active. Experience shows that transformer
noise levels 10 dB above the lowest ambient frequency result in complaints from the
community while a 5 dB increase does not normally produce a response. A 5 dB

337
cushion, however, is considered to be the minimum required to accommodate the
temporal variations in radiation pattern and the atmospheric effects previously
discussed.

8.3.4 REQUIREMENTS
Existing substations have to comply with newer, more restrictive noise regulations
sooner or later, especially when complaints are filed about the elevated noise level of a
certain substation.

Methods used to mitigate noise problems in existing substations will depend on factors
such as:
a) The level of noise exceeding the approved level for the area in which the
substations is located
b) The economics of various solutions – the analysis has to take into account the
life cycle cost of the installation, as well as, the benefits the utility may acquire by
addressing concerns of the community
c) Operational and maintenance implications of installing a certain sound mitigation
solution around transformers and/or reactors
d) Issues related to the ability to construct the sound mitigating solution (can it be
done with the equipment live, is there a need for rerouting of power and/or
control cables, etc.)

Table 8-1 gives typical noise limits specific to particular settings.

Table 8-1: Typical Noise Limits

Location Time of Day Noise Limits
Purely Residential Day 45-50 dB
Early Morning/ Evening 40-45 dB
Night 35-45 dB
Mixed Residential Day 50-60 dB
Early Morning/ Evening 45-50 dB
Night 40-50 dB
Commercial/ Industrial Day 60-65 dB
Early Morning/ Evening 55-65 dB
Night 45-55 dB
Industrial Day 65-70 dB
Early Morning/ Evening 60-70 dB
Night 55-70 dB

8.3.5 METHODS OF SUBSTATION NOISE CONTROL

The first step in the process of mitigating a noise problem at an existing substation is to
determine the noise reduction required at the point of the recipient. This is usually the
most critical location on the property line of the station. Noise reduction is the total level
of untreated transformer noise minus attenuation with distance minus lowest ambient
level or the permitted community level. Once the noise reduction has been established,

338
the most appropriate measures for producing the required noise reduction must be
selected.

The following represents the most widely used methods available for minimizing noise
levels created by transformers at existing substations. The newer methods are
described in greater detail.

a) Replacement of Existing Old Transformers with new units with low noise levels.
Manufacturers have made significant steps toward reduction of basic noise levels of
power transformers and reactors. Replacement of an old unit with a new, low noise
unit might prove to be the best solution if the replacement is dictated also by other
factors (end of life of transformer, history of failures of the unit, chronic oil leaks,
etc.). Levels up to 10 dB below the standard levels are practical and the costs range
up to 1 percent of the cost of a standard transformer per decibel depending on the
size. Higher reductions are not normally economically viable compared with other
methods of control.

b) Landscaping: Planting of tall trees on the outside of the fence line in the direction of
the desired noise reduction is one of the solutions that would provide moderate
noise reduction. If space is available around an existing substation, landscaped soil
berms covered with grass and bushes on the crown is another noise reducing
solution. This solution also provides a means of blending the substation into the
community.

c) Simple Open Roof Barriers: The level of noise reduction obtained with this solution
depends on the height of the barrier above the transformer and its relation to the
elevation of the neighbourhood that is targeted for noise reduction. Typically, 8 to 13
dB noise reduction could be achieved with such a barrier. The barrier may be
constructed from a variety of materials, such as steel plate, cement asbestos sheet
or masonry, etc.

d) Sound Enclosure: This enclosure is installed around all four sides of a transformer.
Depending on the level and directions of noise reduction needed, the enclosure can
be with or without a roof. The roof of such a sound enclosure has to be custom
designed for a particular transformer. Adequate space must be provided between
the tank of the transformer and the walls of the enclosure for maintenance staff to
pass. Also, sufficient space must be provided to enable the opening of the doors of
the control box of the transformer. Reductions of up to 20 dB are possible if proper
attention is given to the details of the construction. Coolers of the transformer are
installed outside the enclosure to ensure the design rating of the transformer is not
compromised.

e) Low Frequency Sound Insulation Panel (LFSI): An effective countermeasure for

an existing shunt reactor and/or transformer is the Low Frequency Sound-Insulation
Panel (LFSI panel), designed to minimize low frequency noise production. The new
soundproof panel is composed of sound absorbing materials and a sound-insulation

339
 board with additional weights that are necessary to reduce the vibration created by
the sound-insulation board.
f) Tight Fitting Enclosure: This solution comprises a total steel plate enclosure
(including a steel roof) around the transformer. This solution is commonly known as
the “tea cozy” solution. In this arrangement, the walls are installed close to the tank
with typically a 10 to 15 cm gap filled with acoustically absorbing material.
Strategically placed doors allow access to the control box of the transformer, the tap
changer, etc. Such enclosures could provide a noise reduction of up to 22 dB.
g) Active Sound Cancellation: This method uses a separate sound source to cancel
the transformer noise. This noise source is produced via an amplifier and speaker
system and is equal in amplitude and opposite in phase to the unwanted noise. ABB
has demonstrated that this method is feasible and especially efficient to reduce low
frequency tones. It may provide useful reductions in the range of 10 dB [168].

8.4 RELEASE OF INSULATING OIL

8.4.1 INTRODUCTION
Older substations were built under environmental regulations that were less stringent
than present regulations. Often there were no requirements to contain any potential
leaks of electric insulating liquids within the substation perimeter. Neither were there
limits to the level of contaminants in the drainage systems of these substations.
Containment and control of oil spills at substations are becoming an increasing concern
for most electric utilities. Beyond the threat to the environment and potential legal
consequences, clean-up costs associated with oil spills could be increasingly
unacceptable.

The probability of an oil spill occurring in a substation is very low. However, certain
substations, due to their proximity to ground water resources, open water or designated
wetlands, the quantity of oil on site, surrounding topography, soil characteristics, etc.,
have or will have a higher potential for discharging harmful quantities of oil into the
environment.

8.4.2 USE OF SYNTHETIC ESTER

Petroleum based insulating fluids have been in use in transformers for over a century. It
is estimated that 8 to 10 billion gallons of transformer oil are in service around the world
today. Due to its excellent performance, availability and low cost, mineral oil has little
competition. Concern over fire safety resulted in the use of high temperature mineral
oils in critical applications. A small percentage of transformers use synthetic fluids such
as silicone and synthetic polyalphaolefin (PAO) hydrocarbon fluids. In recent years,
mainly due to environmental concerns, the utilities have been interested in a fully
biodegradable insulating fluid particularly for use in transformers located in areas where
oil spills would contaminate water. The natural hydrocarbon fluids are only
approximately 30% biodegradable; silicone fluid has very low biodegradability. PAOs
can have as much as 70% biodegradability. An ester fluid developed from

340
pentaerythritol and saturated carboxylic acids (examples: Midel® 7131) is fully
biodegradable, and has found application in traction and power transformers.

8.4.3 USE OF NATURAL ESTER [169]

Vegetable oils are plentiful in nature, and therefore should be considered ideal raw
materials for a fully biodegradable insulating fluid. However, a literature search would
show that vegetable oils were considered in the past only for capacitor use. Capacitors
require gas absorbing fluids; hence synthetic aromatic fluids are commonly used.
Vegetable oils, though not aromatic, have unsaturation, hence gas absorbing
properties. The development of a transformer fluid from vegetable oils caught the
attention of researchers only in the 1990s. ABB has developed a vegetable oil based
transformer fluid under the name of BioTemp. Several technical papers have been
published on this fluid since 1997.

The new fluid has an obvious advantage over conventional transformer fluids for use in
locations where oil spills are of great concern. Currently, large oil spills, whether
mineral oil or vegetable oil, require clean-up. However, for smaller spills clean-up may
not be mandatory for vegetable oils. Since distribution transformers carry only limited
amounts of oil (up to a few hundred gallons), there will not be any large spills. The EPA
still needs to relax some of the restrictions imposed for vegetable oil spills and clean-up
before their widespread use become economical for utilities. Other regulatory agencies
at the state level also may need to change their rules for oil spills.

341
9 ECONOMICS OF TRANSFORMER ASSET MANAGEMENT
9.1 FAILURE STATISTICS FOR POWER TRANSFORMERS

Failure of a transformer refers to a lack of performance of its function so that it has to be

taken out of service. Among other things, failure can occur in transformers as a result of
electrical stress from lightning and switching surges, insulation degradation resulting
from overloads and/or contamination, and faults that occur downstream of the
transformer secondary bushings. However, most catastrophic failures in transformers
are the result of faults downstream of the secondary bushings. Depending on the
magnitude and duration of the resulting fault current, large mechanical forces could
develop inside the transformer windings. The transformer’s ability to withstand such
forces depends on several factors, including the design and the condition of the
insulation. For example, low-impedance transformers tend to fail more than higher
impedance transformers since they will experience more severe fault currents. On the
other hand, systems designed with neutral grounding reactors or arc-suppressing
reactors would tend to fail less often because of the limited fault current. On account of
this, and because auto transformers are designed to have low impedance, they will tend
to fail more often than multiple-winding transformers.

In the sections that follow, typical failure rates and causes for large populations of
transformers are presented. It is important to note that a utility company may experience
failure rates that are vastly different from the numbers presented here. As mentioned
above, failure rates depend on many factors, some of which may be specific to a given
utility (maintenance practices, system design practices, geography, etc.). It is important
that each company keeps accurate records of failures and analyses of root causes of
failures. This will allow the development of effective corrective measures to prevent
similar future failures.

9.1.1 CIGRE SURVEY OF FAILURES IN LARGE POWER TRANSFORMERS [170]

In the late 1970s CIGRE working group WG12.05 undertook a task to conduct an
international survey to pinpoint the main causes of transformer failures and to evaluate
transformer outage times. The survey involved transformers and reactors designed for
networks with a highest system voltage not less than 73 kV, without any limitations on
rated power, not older than 20 years and installed on generation, transmission, and
distribution systems. The working group received data from 13 countries (Australia,
Austria, Belgium, Canada, Czechoslovakia, France, Finland, Italy, Japan, Switzerland,
United Kingdom, USA31, and the former USSR). The analysis covered more than 1,000
failures that occurred between 1968 and 1978 relating to a total population of more than
47,000 transformer years with a maximum peak of 7,000 units in 1978.

31
The data received from the USA was in summary form and was therefore not integrated into the CIGRE
study.

342
The failure rate, in percent per year, is calculated according to the following equation:
ny
100
Ny
where :
n y Total number of failed transformers in a given year
Ny Total component years of transformers installed on the system during the year

In this particular study, the failure rates were calculated for different categories of
transformers: by voltage class (60-100 kV, 100-300 kV, and 300-700 kV); by type (GSU,
substation transformers, and auto transformers); by regulation (LTC or no LTC); and by
age (0-5, 5-10, and 10-20 years). Other than for the classification by age of unit, only
failures involving forced outages were considered. Figure 9-1 shows the summary of
failure rates by type and voltage class of the transformers in the study.

14

12
60 - <100 kV
100 - <300 kV
10
Failure Rate (% / year)

300 - <700 kV
All Voltages
8

0
GSU/ GSU/ Substation/ Substation/ Auto/ Auto/ ALL/ ALL/
LTC No LTC LTC No LTC LTC No LTC LTC No LTC
Transformer Type

Figure 9-1: Failure Rate of Transformers With and Without LTC

The following comments were made by the working group concerning the results of the
survey:

343
 For GSU transformers with LTC, the failure rate is considerably affected by the
increase in the winding voltage. This also happens in the case of auto
transformers, while in the case of substation transformers, there appear to be no
significant variations.
The lower failure rates shown for auto transformers with LTC are somewhat
surprising. Auto transformers without LTC show higher failure rates than the units
with LTC. A detailed analysis showed that a concentration of the reported failures
were from groups of units from specific networks. This inevitably skewed the
failure rates towards higher values that are not entirely representative of this
class and type of transformer.
A general failure rate, irrespective of the voltage class or function, is of the order
of 2 %/year.
When no distinction is made with regards to whether a unit has LTC or not, the
failure rates seem to increase with voltage (and therefore probably with rated
power).

The data was also classified by age of the units and by voltage class. The resulting
failure rates are shown in Figure 9-2. The data shows that, other than for the voltage
class, 300 - < 700 kV, there is no significant increase in failure rates with the age of
the units. For the lower voltage units, failure rates tend to decrease slowly with age.

5
Age of Unit (years)
0-5
4
Failure Rate (% / years)

>5 - 10
>10 - 20
3

0
60 - <100 100 - <300 300 - <700
Voltage Class (kV)

Figure 9-2: Failure Rate as a Function of Unit Age and Voltage Class

9.1.2 CANADIAN ELECTRICITY ASSOCIATION FORCED OUTAGE REPORT [171]

344
This section contains a summary of forced outage statistics reported by the Canadian
Electricity Association for all transformer failures in Canada and covering the period
January 1, 1998 to December 31, 2002. The data is presented for different voltage
classes and three phase bank (one three phase bank or three single phase bank)
configurations. The total number of component years for this period was 24,211 years
(roughly half of the number for the CIGRE study in the preceding section). The
distribution of component years by voltage class and type of three phase bank is shown
in Figure 9-3.

9000

Three Single Phase Units

8000
One Three Phase Unit
7000
Total Component Years

6000

5000

4000

3000

2000

1000

0
109kV 110-149kV 150-199kV 200-299kV 300-389kV 500-599kV 600-799kV
Voltage Class

Figure 9-3: Distribution of Component Years by Voltage Class (CEA Forced Outage Report)

Figure 9-4 and Figure 9-5 show the average failure rates for each of the voltage
classifications and the mean duration it takes to return the bank to service after a forced
outage. From the reported data, the following comments can be made:

If all the failures are considered without regard to voltage class, three phase units
on this system fail twice as often as single phase units in a bank.
It takes approximately twice as much time to restore a three phase unit back to
service as it take to restore single phase units in a bank.
The highest failure rates in one three phase bank and three single phase banks
occurred in the 150-199 kV voltage class. These units comprise only 2% of the
entire population of units reported.
The failure rate for the entire set of reported data is on the order of 5.5%.

345
 20%
Three Single Phase Units
18%
One Three Phase Unit
16%
All Tank Configurations
Failure Rate (per year)

14%

12%

10%

8%

6%

4%

2%

0%
109kV 110-149kV 150-199kV 200-299kV 300-389kV 500-599kV 600-799kV ALL
VOLTAGES
Voltage Class

Figure 9-4: Failure Rate by Voltage Class of Transformers on the Canadian Power Grid

4500

4000
Three Single Phase Units
3500 One Three Phase Unit
Mean Duration (Hrs)

All Tank Configurations

3000

2500

2000

1500

1000

500

0
109kV 110-149kV 150-199kV 200-299kV 300-389kV 500-599kV 600-799kV ALL
VOLTAGES
Voltage Class

Figure 9-5: Mean Duration of Failures by Voltage Class of Transformers on Canadian Power Grid

The failure data was reported according to the failed subcomponent. Figure 9-6 shows a
distribution of contributions to failures by subcomponents according to voltage class and
without regard to bank configuration. The subcomponent category classified as “other”

346
is the catch-all category for cases where the origin of the failure is not clearly evident. Of
the remaining categories, LTCs contribute to the most failures for most voltage classes.

60
Bushings (Including C.T.'s)
Windings
On-Load Tap Changer
50
Core
Leads
% Contribution to Failures

Cooling Equipment
40 Auxiliary Equipment
Other

30

20

10

0
109kV 110-149kV 150-199kV 200-299kV 300-389kV 500-599kV 600-799kV
Voltage Class

Figure 9-6: Failures by Subcomponent and Voltage Class

9.2 ECONOMICS OF TRANSFORMER MANAGEMENT FOR FLEETS

AND SPECIFIC UNITS
9.2.1 INTRODUCTION

In a time of increasing competition and deregulation within the electrical power supply
industry, increasing attention has to be paid to cost cutting and economical issues.
Transformers are important elements of a power system. They are important not only to
power system performance and reliability of supply, but also to the financial
performance of companies. The technical complexities of transformers, as well as their
high capital costs and long lifetimes, are important elements in the power system asset
management process.

347
 9.2.2 GENERAL CONCEPT FOR ECONOMICS OF TRANSFORMER MANAGEMENT

The financial aspects related to service activities should be integrated into a more
general concept where the following issues are considered:

Risk management
Specification and purchase
Operation and maintenance
Decision process: repair - refurbishment - replacement

CIGRE has issued a guide where all these steps have been considered, based on
practical experiences accumulated in multiple countries [172]. ABB has developed
different simulation models to study transformer populations [173] or specific units
[174]. The economic analysis associated with the decision to maintain, refurbish or
replace can best be done on a life cycle cost (LCC) basis. A traditional approach to
LCC treats future costs as being fixed and their net present value (NPV) is considered.
The cost of refurbishment or replacement is set for particular times in each case. In
case of refurbishment, variables such as failure rate shall be adjusted to the condition of
the units and an improvement level shall be admitted in relation to the maintenance
action taken. The basic LCC equation is:
n
LCC CA CE CI (CPM CC M CO P COo CR ) CD
0

where CA = cost of apparatus

CE = cost of erection
CI = cost of infrastructure
CPM = cost of planned maintenance
CCM = cost of corrective maintenance
COP = cost of operation
COO = cost of outages
CR = cost of refurbishment or replacement
CD = cost of disposal

The criteria to be considered for an economic analysis shall also considered:

Safety
Condition
Age
Operation condition
Availability
Maintainability
Environmental aspects
Legislation
Risk

348
These criteria are not necessarily independent; some overlap will occur. For example, a
particular risk may be related to an environmental issue; obsolete equipment may still
be maintainable and so on. Risks associated with safety overlap economics, system
availability, maintainability, environmental and legislation criteria. Also, the owner may
choose to apply some or all of the criteria to individual transformers.

The transformers represent a large monetary value and this asset must be managed,
from both short and medium/long term perspectives. A significant part of typical
transformer populations is now 30-40 years old and is approaching an age where the
reliability becomes questionable. However, no reliable information is available regarding
the lifetime of transformers. Nevertheless, the future investment needs will in some way
be dependent on the age distribution. It is obvious that there will be limitations in
available capital for future investments. This leads to a need to lower the peak of the
investments and to distribute the investments over a longer period of time. There is
therefore a need to formulate and implement long term refurbishment and replacement
strategies. Such strategies must be based on sound financial considerations and should
be communicated and quantified in economical terms. The return on investment should
be maximized under given constraints, e.g. a specified level of availability. Along with
the trend towards penalties for unavailability, the risk costs have become increasingly
important and should be a part of the formulation of the types of strategies that are
discussed. Furthermore, the implementation of such strategies will require new tools
and methods to identify the highest risk units within the population [175,176].

9.2.3 DESCRIPTION OF THE SIMULATION MODEL

The basic principle of the model is to minimize the total cost function for the entire
population with respect to the total investment need. It is hence the sum of all capital
expenditures over the studied time period that is minimized. The cost function is a sum
of capitalized cost components. Cash flow effects and availability of capital are not
considered.

The input data to the model are:

- financial information, e.g. internal rate of return, depreciation time, current energy
price
- population data, e.g. number of units and total installed MVA per year
- general transformer data for the population, e.g. average load rate, losses, failure
frequency as function of time, repair time
- general financial transformer data for the population, e.g. average acquisition
value, maintenance costs, repair costs, consequential costs upon failure, cost
and effect of refurbishment

The main cost components of the model are cost of capital, operational and
maintenance costs, cost for life extending actions, cost of losses and risk costs

349
including costs for repair, loss of production/revenues, and other costs, e.g.
environmental.

Risk costs are treated as the failure rate times the consequential costs of the failure.
The difficulty in all such calculations is that the shape of the bathtub curve, i.e. the
failure frequency as function of time, is not known in detail. In this work, it is assumed
that the owner of a network, together with the equipment manufacturer, can make fair
estimates of best and worse case scenarios for the bathtub curve of that particular
population. Based on the use of these scenarios, we can build an estimate of the future
investment needs. The model is designed in such a way that the estimated bathtub
curve can be modelled and entered as input.

In the model, it is assumed that the technical lifetime of a transformer can be extended
by refurbishment actions [177]. In the basic version, the types of action are not specified
but it is up to the user to define the cost of such actions and to estimate the effects. In
the advanced version, the types of action and their costs can be specified as well as the
improvement in lifetime.

The cost of losses influences the evaluation in the sense that a new transformer has
lower losses than an old transformer. Performance of power transformers has gradually
improved over the years. However, a significant step was taken with the introduction of
grain oriented, cold rolled core steel in the mid 60’s.

9.2.4 CASE STUDY BY A UTILITY

To illustrate the benefit of such a survey, here are the results of a study performed by
ABB for a Swiss utility. A condition assessment study was followed by an economic
survey. The population is composed of 50 medium sized network and generator step-up
transformers.

According to the wishes of the utility, the three scenarios below were considered:

No maintenance: no preventive actions except the very basic and mandatory actions
such as fixing oil leakages or changing the air-drying compound.

Light maintenance: basic maintenance plus oil and gas analysis, oil filtering and
drying, periodic on-load tap changer overhaul.

Heavy maintenance: mid life refurbishment – after 15 to 30 years in operation

depending on the condition of the unit defined according to a condition assessment
survey. The transformer is untanked in order to check the active part, retighten the
electrical connections and reclamp the winding blocks. If required, the oil is regenerated
[178] and the active part is dried out [179]. Each accessory is checked and corrective
actions are taken accordingly.

For each of these three scenarios, the optimization process was run to define which
transformer should be maintained or replaced by a new one and when. The graphs

350
presented in Figure 9-7 to Figure 9-9 show the capital that should be spent per year to
maintain (blue bars) and replace old units (red bars) in order to operate the network
during the next 25 years with the requested reliability level.

Figure 9-7: Scenario 1: No maintenance. Net Present Value = 4.5 MEUR

Figure 9-8 : Scenario 2: Light maintenance. Net Present Value = 3.8 MEUR

Figure 9-9 : Scenario 3:Heavy maintenance. Net Present Value= 2.2 MEUR

The net present value of the costs to operate, maintain and renew the fleet over the 25
coming years varies between 2.2 and 4.5 millions Euros depending on the scenario. In
this specific case, refurbishing the transformers (scenario 3: heavy maintenance) after
15 to 30 years depending on their condition is considered to be the best scenario from a
financial point of view.

351
 9.2.5 CONCLUSIONS
The tool described in this section is a very important element in the strategic planning
process. It is used to evaluate future investments and to support decisions for
maintenance and repair. Two important aspects of the analysis should be emphasized:
a) the risk of failure should be evaluated precisely based upon the existing diagnostic
tools. This risk has the largest impact on the replacement policy and is often not taken
into account in the decision making process; b) maintenance actions should be defined
not only in terms of costs but also in terms of the benefits, such as lifetime extension.

352
10 HEALTH AND SAFETY ASPECTS / RECOMMENDATIONS
10.1 PREAMBLE

The information under this section is provided as advice in order to inform the reader of
the important Health & Safety issues that need to be addressed before commencing
works on or near high voltage systems. The recommendations represent a minimum set
of requirements when working on or near electrical equipment rated at voltages above
1000V. If higher standards are required by national regulations and/or local companies,
then those higher standards should be adhered to.

10.2 INTRODUCTION

Electrical power systems contain large amounts of energy. Electrical arcs, as well as
causing radiation burns, can vaporize materials causing fire and explosion. Voltages
above 50V can cause fatal interruption to heart rhythms and cause internal tissue
damage.

In order to maintain safety, it is necessary to work to a framework of rules to suit the

wide range of site situations. The effect of human factors, as well the physical
environment in which work takes place should be considered in the application of these
rules.

10.3 SCOPE

These recommendations are applicable to all Service Providers involved in the supply of
site related services on power and distribution transformers, as well as work on near
high and medium voltage systems and equipments.

This policy is based on the internal ABB “7 Steps principles” , and defines the minimum
safety rules to be followed when working on or near electrical equipment rated above
1000V.

It is strongly recommended that all Service Providers shall follow the safety rules below,
or equivalent, for all site operations.

Each Service Provider may define additional rules and requirements based on local
legislation, capabilities, skills, and the nature of services offered.

353
10.4 DEFINITIONS

Medium or High Voltage: Above 1000V ac.

Electrical Equipment: All medium and high voltage products, including power
and distribution transformers.
Power System: A connected network of electrical devices which, in the
event of a short circuit, contains sufficient energy to cause
serious burns to people or physical damage to the plant,
other equipment or property.
Service Provider: Any organization working on or near Electrical Equipment.
Nominated Person: Qualified Person nominated to have the responsibility for
safety and control of a work party.
Qualified Person: Has the skills and knowledge related to the construction
and operation of electrical equipment and installations.
Understands the hazards of electricity and how to protect
him / herself from these hazards. Is undergoing on-the-job
training and who, in the course of such training, has
demonstrated an ability to perform duties safely at his or
her level of training and who is under the direct
supervision of a qualified person is considered to be a
qualified person for the performance of those duties.
Trained Qualified Person: Qualified Persons shall, at a minimum, be trained in and
familiar with the following: the skills and techniques
necessary to distinguish exposed live parts from other
parts of electrical equipment, the skills and techniques
necessary to determine the nominal voltage of exposed
live parts, the clearance distances for qualified and
unqualified persons and the corresponding voltages to
which the qualified person will be exposed, and the
decision-making process necessary to determine the
degree and extent of the hazard and the personal
protective equipment and job planning necessary to
perform the task safely.
Work Party: Persons working or testing under the control of a
Nominated Person.
Work Package: Defined list of service related tasks to be completed by the
Work Party within a set amount of time.
Switching: Operation of equipment designed to alter the electrical
state of an electrical installation such as connection,
disconnection, altering the flow of current, or system
operating voltage, inserting a switching device from a test

354
 to a service position, withdrawing a switching device from
a service to a test position.
Competence: A combination of skills, knowledge, and practical
experience, to enable a person to carry out tasks
effectively and safely.
Customer: Company in charge of the equipment and it’s operation.

10.5 SAFETY MANAGEMENT

These rules focus on electrical safety aspects and are to be implemented in association
with other general safety guidelines (such as personnel transportation, manipulation of
oil, use of Personal Protection Equipments, work at height, etc…).

Each Service Provider shall have a documented electrical safety management policy
(safety-related work practices program). The policy or program shall contain information
on:
a) The application of electrical safety rules,
b) Training and competency,
c) Written authorization of people,
d) Review and audit of compliance.

Each Service Provider shall have suitable written procedures to ensure that electrical
safety rules are applied to all site operations.

10.6 DOCUMENTATION

The following documentation is recommended to be produced by the Nominated Person

prior to commencing work:
a) Risk Assessment or Electrical Job Hazard Analysis Sheet (see sample Appendix 5,
section 9.9.5),
b) Safety Check Sheet (see sample Appendix 7, section 9.9.6),
c) Safety Permits to work (see sample Appendix 8, section 9.9.7),
d) Pre-job briefing or Tail-Gate Meeting form,
e) Energized Electrical Work Permit if applicable (see sample Appendix 9, section
9.9.9).

These documents must be retained in job files for verification and audit.

355
10.7 ELECTRICAL SAFETY RULES
10.7.1 GENERAL RULES

a) All Electrical Equipment shall be treated as live unless made safe and
released for work as described in section 9.7.3 below.
b) Electrical equipment is subject to Electrical Safety Rules when it is capable of
being energized from a power system through operation of switchgear, or by
the replacement of fuses or links.
c) Electrical Safety Rules must, at all times, be applied in such a manner as to
prevent danger to persons switching, working on, or in the vicinity of live
apparatus.
d) Each person within the Work Party must have the Competence to undertake
the scope of work planned for them.
e) Normally the Customer personnel shall undertake all actions to safely
disconnect and earth (ground) Electrical Equipment to be worked on.
f) Normally Service Provider personnel shall not work on live equipment.
g) Work packages which require work on or very near live equipment must
follow relevant additional rules in section 9.7.4.
h) Personnel on Site must at all times wear appropriate Personal Protection
Equipment (PPE) . Flame resistant (FR) clothing rated 8 cal/cm2 shall be
worn as an everyday uniform. NFPA standard 70E will be used to determine
additional protection in the absence of local standards.
i) Other PPE may be used according to local regulations and risk assessment
controls.
j) There must be safe and proper means of escape in the event of failure of any
live apparatus.
k) The Nominated Person must establish prior to starting work relevant
telephone numbers for emergency services.
l) Appropriate safety precautions shall be taken when working on high
equipment to avoid injury through falling.
m) No un-authorised person may enter the work area.

10.7.2 COMMUNICATION AND CONTROL RULES

a) For each Work Package a Nominated Person must be appointed in writing.

The Nominated Person must understand the limitations of his work and the
safe working area.
b) The Nominated Person must have sufficient training and experience to be
competent in the discharge of those responsibilities.

356
 c) Irrespective of whether the Customer has completed all or part of the steps to
safely earth (ground) Electrical Equipment, the Nominated Person shall
ensure and verify that all Electrical Safety Rules have been complied to.
d) The Nominated Person has the authority not to proceed with work should he
consider that compliance to Electrical Safety Rules has not been met.
e) Only the Nominated Person has the authority to issue a Permit to work.
f) All site work packages must have a written scope of work which clearly
defines the tasks to be completed, the relevant time frame to complete tasks,
the resources to be utilized as well as relevant contact details of the
customer’s representative.
g) There must be arrangements to ensure communications are clear and not
subject to misinterpretation. Typically the following may apply:
Instructions are written down and repeated to the sender,
Instructions and confirmation times are noted,
The use of standard phrases which can not be misinterpreted,
The use of standard schematics by all persons concerned.
h) Work or switching must not be commenced by pre-arranged signals or time.
i) Where there are numerous persons carrying out switching operations or work
parties on the same circuit, a control person must be nominated and be
responsible for safety co-ordination.

10.7.3 RULES FOR WORKING ON DEAD ELECTRICAL EQUIPMENT

A safe working area must be created and maintained in accordance with the
following principles. All 7 principles must be applied on every occasion:
Step 1 Clearly identify work location,
Step 2 Disconnect completely, secure against re-connection (Lockout /
Tagout),
Step 3 Provide protection against live parts,
Step 4 Take special precautions when close to bare conductors,
Step 5 Verify the installation is dead by approved means,
Step 6 Carry out earthing (grounding) and short circuiting,
Step 7 Issue a Permit to work.

Step 1: Clearly identify the work location

a) Sufficient labels, schematics and plans must be available to clearly identify

the location and the Electrical Equipment to be worked on.

357
 b) Where Electrical Equipment is not easily identifiable, such as cables, suitable
identification arrangements must be made by the Nominated Person.
c) Nearby live Electrical Equipment must be identified as dangerous by the
application of highly visible temporary warning labels.
d) There must be clear and continuous identification of the safe working area. It
must be possible to easily distinguish between safe and dangerous areas
throughout the course of planned work.

Step 2: Disconnect Completely, Secure against re-connection

a) Verify that the Electrical Equipment has been safely disconnected and
earthed (grounded).
b) A suitable electrical gap (see Appendix 2 & 3, section 9.9.2 & 9.9.3) must be
provided across all points at which the apparatus may be made live including
potential in-feeds from low voltage apparatus.
c) The gap must be physically secured from inadvertent or willful re-connection,
by the application of barriers or locks; or removal of fuses to a safe place; or
removal of apparatus from its normal service position; or
disconnecting/blocking stored mechanical energy devices.
d) It shall not be possible for remote protection or control to reconnect the circuit
under work through electrically closing a switching device.
e) A prominent warning notice must be fixed to each point of disconnection.
f) All keys must be kept in a secure place.

Step 3: Provide protection against live parts

a) All live apparatus in the vicinity of the work place must be locked off, clearly
identified and made inaccessible by barriers or other means.
b) A risk assessment must be carried out to ensure that dangerous voltages
cannot be directly or indirectly created on any of the electrical circuits
connected to the point of work from nearby live circuits (see Appendix 2,
section 9.9.2) including the re-arrangement of earth (ground) connections.
c) Special attention must be paid when the scope of work requires phased
outages of different circuits. The sequence of circuits to be made dead must
be clearly understood by all members of the work party. With each change of
circuit from live to dead, or vice versa:
Access control and warning labels must be changed to suit,
Each member of the Work Party must fully understand which new circuits
are now live.
d) Work on Electrical Equipment connected to overhead lines must cease in the
event of a lightning storm.

358
e) On wood pole lines with one or more circuits live, precautions must be taken
against steelwork being or becoming live.
f) On certain high voltage cable systems it may be necessary to take
precautions against large induced voltages, for example by fully insulated
working.
g) Precautions must be taken to prevent danger from low voltage conductors at
the point of work.

Step 4: Take special precautions when working close to bare conductors

a) Special precautions must be agreed with the Nominated Person when the
minimum clearances (Appendix 1, section 9.9.1) cannot be made to
unearthed (unground) conductors during preparation of the safe working area
by applying screens, testing and applying portable earths (ground). Details on
Approach Boundaries distances are given in Appendix 1, section 9.9.1).
b) The minimum safety precautions in the above circumstances are:
Wear appropriate voltage-rated gloves and applicable PPE,
Ensure access is suitable to avoid inadvertent slips,
Be accompanied by a second person who can render assistance.

Step 5: Verify the installation is dead

a) Where the design of the Electrical Equipment allows, it must be confirmed

dead by a suitable testing device at all points of work and all points where
portable earths (grounds) are to be applied.
b) The testing device shall be of a proper category rating, and proved before and
after each test.
c) Where the design of the Electrical Equipment precludes this, other suitable
arrangements shall be agreed with the Nominated Person. Typically: firing a
spike through underground cables, using proximity testing devices on
insulated conductors, operation of switches to earth (ground) or tracing cables
or conductors back to a visible earth (ground) point.

Step 6: Verify earthing (grounding) and short circuiting

a) The Electrical Equipment to be worked on must be connected to earth

(ground) by connections and conductors capable of carrying the full short
circuit current at that point.
b) Where possible the Electrical Equipment shall be earthed (grounded) by a
fully rated switch.
c) Earths (grounds) must be applied between the point of work and all possible
sources of HV supply, and they shall remain in position for the full duration of
work.

359
 d) Earths (grounds) must be applied by a Qualified Person who is wearing all
applicable PPE and FR clothing to match the Hazard / Risk category outlined
in NFPA 70E.
e) Portable earths (grounds) shall be applied to all phases and in such a manner
as to prevent danger from residual charge or induced voltages.
f) Separate earths (grounds) at the point of work shall be connected to create
an equi-potential zone for all persons at the point of work. They may be
moved during the course of the work.
g) Precautions shall be taken to prevent danger from voltages developed across
earth (ground) conductors connected to earth (ground) at different points.

Step 7: Issue a Permit to work

a) Note: The Permit to work is about control of the workplace and people. It is a
summary of the previous protective measures and hazards. It is also a clear
statement of responsibility from the Nominated Person for these protective
measures.
b) The Electrical Equipment to be worked on must be released for work through
the issue of an electrical Permit to work at the point of work. The Nominated
Person issuing this is responsible for creating and maintaining the safe
working area in accordance with 7 steps principles.
c) The Permit to work shall contain clear, legible details of:
The location of the work, and the precautions that have been taken to
establish the safe working area,
The scope of the work,
Hazards in the immediate area,
Completion of Site Risk assessment,
Completion of Safety Check sheet,
Signature of the Nominated Person in charge of the work party, times and
dates of issue and cancellation,
Signatures of the Work Party members.
d) The Nominated Person in charge of the work party shall be responsible for
the safety of his work area and all other persons in his work party regardless
of seniority.

360
 e) Warning signs and demarcation equipment (barricades) are not to be
removed until after cancellation or the issue of further safety documents.
f) The Nominated Person in charge of the work party shall be responsible for
ensuring all persons and tools are withdrawn on completion of work, and
hand over of the Electrical Equipment in proper condition according to the
work undertaken.
g) Where the boundary of the safe working area changes during the course of
the work, further Permits to Work must be issued.

10.7.4 RULES FOR WORKING ON OR VERY NEAR LIVE ELECTRICAL EQUIPMENT

All work packages that require work on or very near live Electrical Equipment,
other than energized troubleshooting must have prior written approval from the
local Service Provider manager, who must be satisfied that personnel have the
appropriate skills to undertake the required task.

The rules in Sections 9.7.1 - 9.7.3 shall be followed where relevant.

The risk assessment shall specifically evaluate whether the work can be done
dead or in some other way, and what additional special safety measures must be
undertaken.

10.7.5 SWITCHING

Note that switching includes insertion or withdrawal of a switching device.

Switching must be undertaken by the appropriate and authorized customer

personnel in control of the system with the approval of the Service Provider
personnel.

Appropriate personal protection equipment must be worn by personnel in the

switching area during switching operation. At minimum flame retardant overalls,
gloves and a safety visor must be worn.

Where possible switching shall be undertaken remotely.

10.7.6 WORK ON OR VERY NEAR LIVE CONDUCTORS

Any work which cannot be done whithin the safe working clearances in Appendix
1, section 9.9.1 must be the subject of a special written procedure (Energized
Electrical Work Permit), and carried out by specially trained and authorised
(Qualified) persons.

Work is considered ‘Live work’ when it is carried out within the live zone
distances set out in Appendix 1, section 9.9.1 (Prohibited Approach Boundary) or

361
 if the work requires the removal of covers, doors or barriers to expose live
conductors.

Work must only be carried out by a (Qualified) person who is properly trained and
authorized in writing, and who is familiar with the function and operation of the
Electrical Equipment concerned.

Specific PPE according to the hazard / risk category outline in NFPA 70E must
be worn by personnel undertaking the work. At a minimum flame retardant
overalls, gloves and a safety visor must be worn.

10.7.7 TESTING AND COMMISSIONING

Where it is necessary to carry out electrical tests on Electrical Equipment, the

safe working area must be established by following steps 1 - 6 as described in
section 9.7.3, before being released for tests by a safety document, and earths
(grounds) removed.

The test area must be under the control of a suitably competent (Qualified)
person and precautions taken to avoid electric shock as in Step 4 in Section
9.7.3.

Additional warning signs and barriers must be used to protect other persons
nearby from test voltages. If signs and barriers do not provide sufficient warning
and protection from electrical hazards, an attendant shall be stationed to warn
and protect employees

Work on Electrical Equipment shall only be done, or supervised, by a person

competent (Qualified) to recognize danger and apply controls from:
Live low voltage wires ac and dc,
Voltages arising from current transformers, protection, pilot cables and earths
(grounds),
Primary power conductors at high or low voltage,
Interference with fully operational power systems,
Test equipment.

10.8 WORK AT HEIGHT: ADDITIONAL SAFETY EQUIPMENT FOR

POWER TRANSFORMERS.
10.8.1 “NO-RISK SYSTEM”

ABB has studied and patented “NO-RISK System” , a collective safeguard to prevent
falls that complies with Directive 2001/45/EC of the European Parliament and of the
Council, concerning the minimum safety and health requirements for the use of work
equipment by workers at work: “...if temporary work at a height cannot be carried out

362
safely and under appropriate ergonomic conditions from a suitable surface, the work
equipment most suitable to ensure and maintain safe working conditions must be
selected…”.

“NO-RISK” is a cable based fall arrest system that can be installed on every kind of
transformer and offers protection to individuals working at height during the
maintenance activities that transformers could require.

Figure 9-10-1: “NO-RISK System” typical application

“NO-RISK System” equipment includes:

One (1) Collective safeguard, certified by ABB, that can be installed on every
transformer by means of patented support structures. The kit consists of:
- Two (2) Supporting brackets to be permanently welded on the transformer
tank,
- Two (2) Safety rope vertical supports,
- Two (2) Safety rope locking devices,
- One (1) Shock absorber,
- One (1) Safety rope,
- One (1) Turn buckle,
- One (1) Box to store the above listed parts.

363
 One (1) Personal safety kit, consisting of:
- One (1) Safety harness,
- One (1) Adjustable rope,
- One (1) Snap hook,
- One (1) Hard hat,
- One (1) Pair of gloves,
- One (1) Reflective vest,
- One (1) Bag containing the kit.

Figure 9-10-2: “NO-RISK System” equipments

364
 10.8.2 “FALL ARREST TOWERS AND BASE PLATES”

10.9 APPENDICES

Notes on use of Appendices:

The following tables are for guidance only. They should be used as part of a thorough
risk assessment. The tables have been derived from ABB documents which contain
much more detail. Where possible consult local standards, which may state different
figures and methods of application to be adhered to.

365
 10.9.1 APPENDIX 1 - MINIMUM WORKING CLEARANCE

The minimum working clearance is the minimum distance, in air, between a live
exposed conductor, and any part of a person’s body, tool, conductor or apparatus that a
person may be touching. The clearance must take into account the nature of the work
and relative positions of the live conductor and the working area.

Where work infringes this clearance then addition control measures should be
implemented to maintain safety.

Where work infringes the live working zone then it shall be considered to be live work.

System Voltage Minimum Working Live Working Zone

Clearance mm
kV mm
<1 700 (2 ft. 3 in.) 200 (0 ft. 8 in.)
10 1350 (4 ft. 5 in.) 350 (1 ft. 2 in.)
20 1400 (4 ft. 7 in.) 400 (1 ft. 4 in.)
36 1580 (5 ft. 2 in.) 580 (1 ft. 11 in.)
70 1900 (6 ft. 3 in.) 900 (2 ft. 11 in.)
132 3100 (10 ft. 2 in.) 1100 (3 ft. 7 in.)
275 3800 (12 ft. 6 in.) 1800 (5 ft. 11 in.)
480 5200 (17 ft. 1 in.) 3200 (10 ft. 6 in.)

NB. Source: EN 50110-1, whereas in this standard:

- Minimum working clearance is ‘Dv’ (Outer limit of vicinity zone),
- Live working zone is ‘Dc’ (Minimum working distance) in this standard.

Approach Boundaries to exposed energized parts for Shock Protection (mainly

applicable in the USA).

(1) Definitions.
Limited Approach Boundary: A shock protection boundary to be crossed by only
qualified persons (at a distance from a live part),
which is not to be crossed by unqualified persons.
Set your barricades here or at flash protection
boundary if it’s larger than 10 feet (3050 mm),
Restricted Approach Boundary: A shock protection boundary to be crossed by only
qualified persons (at a distance from a live part)
which, due to its proximity to a shock hazard,
requires the use of shock protection techniques and
equipment when crossed. (use voltage rated PPE
and Equipment),

366
Prohibited Approach Boundary: A shock protection boundary to be crossed by only
qualified persons (at a distance from a live part)
which, when crossed by a body part or object,
requires the same protection as if direct contact is
made with a live part.

(2) This distance is for shock protection, if flash protection boundary is greater the
barricades need to be at flash protection boundary.

(3) Protection boundary may also need to be extended if conductive materials that
may come intact with energized parts are being handled.

All dimensions are live part to employee:

Nominal System Limited Approach Restricted Prohibited

Voltage Range Boundary 1 Approach Approach
Phase to Phase Location of Boundary 1 Boundary 1
kV barricades to Qualified person Entering this area
protect unqualified wears voltage may result in
persons from rated PPE before serious injury or
electrical shock 2-3 entering this area death
ft. in. (mm) ft. in. (mm) ft. in. (mm)
0.051 to 0.750 10 ft. 0 in. (3050) 1 ft. 0 in. (305) 0ft. 1 in. (25)
0.751 to 15 10 ft. 0 in. (3050) 2 ft. 2 in. (660) 0 ft. 7 in. (180)
15.1 to 36 10 ft. 0 in. (3050) 2 ft. 7 in. (790) 0 ft. 10 in. (255)
36.1 to 46 10 ft. 0 in. (3050) 2 ft. 9 in. (840) 1 ft. 5 in. (430)
46.1 to72.5 10 ft. 0 in. (3050) 3 ft. 3 in. (990) 2 ft. 1 in. (635)
72.6 to 121 10 ft. 8 in. (3250) 3 ft. 5 in. (1050) 2 ft. 8 in. (815)
138 to 145 11 ft. 0 in. (3350) 3 ft. 7 in. (1100) 3 ft. 1 in. (940)
161 to 169 11 ft. 8 in. (3550) 4 ft. 0 in. (1220) 3 ft. 6 in. (1070)
230 to 242 13 ft. 0 in. (3950) 5 ft. 3 in. (1600) 4 ft. 9 in. (1450)
345 to 362 15 ft. 4 in. (4700) 8 ft. 6 in. (2600) 8 ft. 0 in. (2450)
500 to 550 19 ft. 0 in. (5800) 11 ft. 3 in. (3450) 10 ft. 9 in. (3300)
765 to 800 23 ft. 9 in. (7300) 14 ft. 11 in. (4550) 14 ft. 5 in. (4400)

367
 10.9.2 APPENDIX 2 - MINIMUM DESIGN CLEARANCES WHERE POWER LINES CROSS OR
ARE IN CLOSE PROXIMITY

These provide a useful guide when assessing the degree of risk from nearby exposed
power lines. They do not take into account the additional risks of failure created by
extreme weather conditions or poor condition of the equipment.

The clearances are calculated for the upper conductor operating at its maximum likely
temperature and the lower conductor at 25 °C less. They must also allow for a
maximum wind loading swing of 45deg.

System Voltage Minimum Design Clearance

between circuits
kV mm
<1 1000 (3 ft. 3 in.)
10 1800 (5ft. 11 in.)
24* 1900 (6 ft. 3 in.)
36 2000 (6 ft. 7 in.)
70 2300 (7 ft. 7 in.)
132 2700 (8 ft. 10 in.)
275 3700 (12 ft. 2 in.)
480 4400 (14 ft. 5 in.)

* Extrapolated
Source: derived from UK spec. EATS 43-8

Company vehicles or mobile equipment capable of having parts of its structure elevated
near energized overhead lines shall be operated so that a clearance of 10 ft. (3050 mm)
is maintained. If the voltage is higher than 50kV, the clearance shall be increased 4 in.
(100 mm) for every 10kV over that voltage. However, under any of the following
conditions, the clearance may be reduced:
If the vehicle is in transit with its structure lowered, the clearance may be reduced
to 4 ft. (1220 mm). If the voltage is higher than 50kV, the clearance shall be
increased 4 in. (100 mm) for every 10kV over that voltage,
If insulating barriers are installed to prevent contact with the lines, and if the
barriers are rated for the voltage of the line being guarded and are not a part of or
an attachment to the vehicle or its raised structure, the clearance may be
reduced to a distance within the designed working dimensions of the insulating
barrier.

368
 10.9.3 APPENDIX 3 - MINIMUM SEPARATION ACROSS POINT OF DISCONNECTION IN AIR

These distances are between open isolator blades, or distances between fixed points of
disconnection. Where separation is not fixed, such as overhead line spans, greater
clearances will be required (see Appendix 1 & 2). Consideration should also be given to
other conditions which may bridge or degrade the insulation gap.

System Voltage Electrical clearance

(Phase to earth / ground)
kV mm
12 229 (0 ft. 9 in.)
24* 330 (1 ft. 1 in.)
36 432 (1 ft. 5 in.)
66 786 (2 ft. 7 in.)
132 1473 (4 ft. 10 in.)
275* 2800 (9 ft. 2 in.)
480* 4000 (13 ft. 1 in.)

* Extrapolated

Source: derived from EN 60129

369
 10.9.4 APPENDIX 4 - PRINCIPLES OF RISK ASSESSMENT

The following outlines the very basic principles of risk assessment. There is a
considerable amount of information generally available on the topic.

Risk assessment should be carried out at both the planning and implementation stages
of projects. It is the key to the application of rules and to setting up the ‘Safe Working
Area’.

There are 3 principles:

1) Identify the Hazards

These are things which have the capacity to cause harm in any given situation.

2) Assess the Degree of Risk

This is much more difficult. It takes into account the hazard in relation to the
tasks, the potential severity of failure and the likelihood of failure. It must consider
the competency of the people, human factors, technical factors and
environmental factors. The degree of risk will change if unforeseen
circumstances alter the work. In these circumstances the risk should be re-
appraised.

3) Define and Implement Controls

These will be determined by the degree of risk. There is a hierarchy of the
preferred type of controls, starting with removing the hazard and descending to
the use of PPE. Controls may be a combination of physical measures and by
management of people and documentation/rules. They must take into account
human factors.

370
10.9.5 APPENDIX 5 - EXAMPLE OF SAMPLE RISK ASSESSMENT SHEET.

371
 10.9.6 APPENDIX 6 - ELECTRICAL JOB HAZARD ANALYSIS SHEET.

(To be completed by the person doing the task/job by checking the appropriate boxes).

Yes No N/A CONSIDERATIONS FOR TASK/JOB:_______________________

Is the job scope understood? The job scope must be understood before continuing
the analysis and performing any work.
Am I trained and qualified to do the task? Only trained and qualified employees
may perform electrical task or complete this analysis. To be qualified, you must
understand the construction and operation of the particular equipment that you
have been asked to work on and how to avoid the hazards associated with this
equipment and environment.
Have I performed this task before? Contact your supervisor if you feel you are not
qualified to perform the task or complete this analysis.
Have you conducted a pre-job briefing? Before starting each job, the employee in
charge shall conduct a job briefing with all employees involved. The briefing shall
cover at least the following subjects: Job scope, hazards associated with the job,
how to avoid the hazards (work procedures involved, special precautions, energy
source controls, and personal protective equipment requirements), and your role in
performing the work. This job briefing must be performed even if you are
working alone.
Is the electrical equipment Locked and Tagged Out (completely de-energized for
this task)?
Have you verified that all forms of energy have been isolated from the equipment?
For electrical energy, you must first test your voltage detector for proper operation,
verify that no voltage is present within the work area, and then test to see if your
voltage detector is still working properly.
Is there exposed, energized electrical equipment near the work area? If so, check
the appropriate voltage range(s).
_____> 49 to 250 Volts _____> 250 to 600 Volts _____> 600 Volts
Do I have the proper electrical personal protective equipment required to do the
task? Proper electrical personal protective equipment must be acquired, when
applicable, prior to the completion of the analysis and beginning the work.
Electrical jobs may be split into tasks and a Job Hazard Analysis performed for
each task.
Do I have the proper voltage rated tools and test equipment in proper working
order to perform this task? Proper voltage rated tools and test equipment must be
acquired, when applicable, prior to completion of the analysis and beginning the
work.
Is a permit required for this task? If so, which of the following permits is required
for this task?
_____ Energized Electrical Work Permit
_____ Confined Space Entry Permit
_____ Other Documentation
Can I control my environment? You must have control of your environment prior to
the completion of this analysis and beginning work.
____ Close working quarters ____ High traffic area
____ Flammable atmosphere ____ Wet or damp environment
____ Inability to control intrusion/distraction by others
____ Could drop a tool/component onto exposed, energized equipment/bus
bars/conductors

372
 Do I understand that completing this job safely is more important than the need or
pressure to complete this job? You must be able to answer this question “Yes”
prior to the completion of this analysis and beginning work. If you have a problem
answering this question “Yes”, contact your supervisor or the Health & Safety
Department.
Do I have any unresolved safety concerns about performing this task? You must
be able to answer this question “No” prior to the completion of this analysis and
beginning work. If you have a problem answering this question “No”, contact your
supervisor or Health & Safety Department.
NOTE: If you have checked any gray boxes, contact your supervisor or Health & Safety
Department before working on this equipment.

Employee’s Signature: ___________________ Date: ___________ Time: __________

373
 10.9.7 APPENDIX 7 - SAMPLE SAFETY CHECK SHEET

Safety Check Sheet

Customer Name Site Name
Substation Name Job Number
Nominated Person
Telephone Number for Ambulance/First Aid:
Telephone Number for Fire Department:
Telephone Number for Emergency Services:

Check List Yes No Comment

1 Is there safe means of escape in event of switchgear
failure?
2 Is the scope of work documented and clearly defined
3 Do members of the Work Party have the competence
to perform the scope of work?
4 Have all members of work party been issued with
correct PPE for the scope of work?
5 Has a risk review been undertaken to evaluate risks
within the nearby vicinity and on the job?
6 Does the risk assessment consider direct and
indirect possibilities for re-connection of the circuit
under work?
7 Is work place clearly identified and demarcated?
8 Is live equipment nearby the work place clearly
identified?
9 Does each member of the work party fully
understand which circuits are live and which circuits
are dead?
10 During execution of the work is it planned to change
any circuits from dead to live or vice versa?
WARNING : If Yes then with each change:
demarcation and warning signs must be adjusted
accordingly
each member of the work party must fully
understand which new circuits are now live
11 Is the electrical equipment to be worked on
disconnected and safely earthed (grounded)?
12 Have adequate steps been taken to prevent re-
connection?
13 Can remote protection or control operation re-
connect circuit under work?

374
14 Are there warning notices at all disconnection points?
15 Has access to nearby live equipment been prevented
through the positioning of barriers, locks or other
means?
16 Have adequate safety measures been taken to
prevent access to live low voltage conductors?
17 Have voltage transformers been safely disconnected
to prevent high voltage at the point of work being
applied through energising transformer low voltage
side?
20 During preparation of the safe working area, will any
member of the work party come close to live exposed
conductors?
WARNING: If yes then special safety precautions
must be taken during this activity
21 Has a voltage detection tester been used to verify
that all points of work are dead?
22 Has the voltage detection tester been checked for
correct operation before and after use?
23 Have earth (ground) been applied to all phases at the
point of work?
24 Are earth (ground) conductors and connections rated
for maximum fault level at point of work?
25 Is there a written Electrical Safety Permit to work?

Additional checks for work on or near live

equipment
1 Is scope of work approved by local Service Unit
manager?
2 Do members of the work party have correct PPE for
the scope of work?
3 Has a detailed risk assessment been undertaken?
4 Do all members of the work party fully understand
which parts of the equipment are live?
5 Are all tools and test equipment functioning correctly
and calibrated?

Signed: ………………………… Date: …………………………

375
 10.9.8 APPENDIX 8 - SAMPLE SAFETY PERMIT TO WORK.

Electrical Safety Permit to work

Customer /
Plant-Responsible-Person
Contractor /
Nominated Person
Plant / Control Panel

Location and Scope of Work

to be executed

All parts of the electrical system, which are not addressed, are considered to be in live condition!

1. Authorization for Switching Operations

Date, Time, Name, Signatures
The customer (plant responsible person):

if necessary the authorization is handed over to:

2. Permission for Working

The Plant-Responsible-Person gives permission to the Nominated Person to execute above described
work on the defined area / part of the plant.
Work place clearly identified
Circuit disconnected and secured against re-connection
Access to nearby live parts prevented
Installation verified as dead
Circuit correctly earthed (grounded)
Execution and Confirmation by the Plant-Responsible-Person and Checked by the Nominated Person:
Plant-Responsible- Date: Time: Name: Company: Signature:
Person assigned:

Nominated Person took Date: Time: Name: Company: Signature:

over:

376
 3. Release to start Work
The Nominated Person performs the workplace risk assessment, completes safety check sheet and
grants the release to start work
Site risk assessment completed Safety check sheet complete
Plant-Responsible- Date: Time: Name: Company: Signature:
Person released:

Instructed employees, Date: Time: Name: Company: Signature:

own one's and
contractors:

Date: Time: Name: Company: Signature:

Date: Time: Name: Company: Signature:

Date: Time: Name: Company: Signature:

4. End of Work
Inspection and announcement of completion of described and executed work / evacuation of working
area
Remove all tools, equipment and material
Uncover or remove protection of adjacent live parts
Remove earthing (grounding) and short-circuiting
Remove all locks or devices, which were used to prevent re-connection
Remove all signs used for marking the working area
FROM NOW ON THE INSTALLATION IS REGARDED AS LIVE
Inspection by Nominated Person if the installation is ready for re-energizing
BEFORE RE-ENERGIZING ALL PEOPLE NOT NEEDED HAVE TO LEAVE THE DANGER ZONE
Nominated Person hand over the installation to the Plant-Responsible-Person
Re-connection by the person authorized for switching operations
Nominated Person Date: Time: Name: Company: Signature:
assigned:

Plant-Responsible-Person Date: Time: Name: Company: Signature:

took over:

The released work space herewith has been energized!

Original: Nominated Person Copy 1: Plant responsible person

Original to be retained in job file

377
 10.9.9 APPENDIX 9 - SAMPLE ENERGIZED ELECTRICAL WORK PERMIT.

Energized Electrical Work Permit

TO BE COMPLETED BY THE REQUESTER: CHECK

1) Job Location (Customer Name & Location):

2) Description of Work to be performed:

3) Why Equipment cannot be de-energized:

4) Description of Safe Work practices implemented:

5) Location of main disconnect (In case of Emergency):

6) CPR-Trained Stand-By person (Required):

7) Arc flashing rating of Energized Work Area: 0 1 2 3 4

8) Personal Protective Equipment to complete the job:

9) Method(s) to Restrict Access:

APPROVAL TO PERFORM THE ENERGIZED WORK DESCRIBED ABOVE

REQUESTOR JOB TITLE DATE

ELECTRICALLY QUALIFIED PERSON JOB TITLE DATE

YES NO NA
SERVICE MANAGER OR MANAGEMENT REPRESENTATIVE SERVICE MANAGER CONTACTED DATE

378
 References

1. ABB Transformer Handbook, Rev 02, 2004

2. ABB Transformer Handbook, Rev 02, Sect. 7.1.1, 2004
3. Excerpted from ABB Instruction Leaflet 44-666G, “Instructions for Installation,
Maintenanceand Storage of Type “O Plus C” Bushings 115kV and Higher”
4. Pritpal Singh, “C2 Power Factor and Capacitance of ABB Type O Plus C, AB, and
Type T Condenser Bushings”, Doble Client Conference Paper, 2001.
5. ABB Transformer Handbook, Rev 02, Sect. 7.3.2, 2004
6. Rick Youngblood, “Cinergy Taps Into Longer Life”, Transmission & Distribution
World (Online Edition), Nov 1, 2000
7. A. Tahani, "Numerical Modeling of Electrification Phenomena and Its Implication
for Dielectric Integrity", 1994, Ph.D. Thesis, Rensselaer Polytechnic Institute, Troy
NY 12180-3590
8. R. Tamura, et al., "Static Electrification by Forced Oil Flow in Large Power
Transformers", IEEE Trans. on PAS, Vol. 90 (1980), pp.335-343
9. J. F. Roach and J. B. Templeton, “An Engineering Model for Streaming
Electrification in Power Transformers”, Electrical Insulating Oils, STP 998, H.G.
Erdman, Ed., American Society for Testing and Materials, Philadelphia, 1988, pp.
119-135
10. G. K. Frimpong and J. M. Walden, “Streaming Electrification Model Tests”, EPRI,
Palo Alto, CA: 1999. TR-113535
11. Pettersson, L.: “Estimation of the Remaining Life of Power Transformers and Their
Insulation”. (Electra No 133, Dec 1990, pp 65-71)
12. Pettersson, L., Fantana, N.L., Sundermann, U.: “Life Assessment: Ranking of
Power Transformers Using Condition Based Evaluation - A New Approach”,
(CIGRE Paper12-204, 1998)
13. Pettersson, L., Fantana, N.L., Persson,J.O, Walldén, K.I.,: “Condition Based
Evaluation of Net Transformers – Experience from a New Ranking Procedure.”
(CIGRE Paper12-108, 2002)
14. Craig L. Stiegemeier, Mark Perkins, Asim Fazlagic, James Dunn, “Mature
Transformer Management ProgramTM for Arc Furnace Transformers”,
AISTech2004 Technical Conference, Association for Iron & Steel Technology
Conference and Exposition, Nashville, TN, September 16, 2004
15. Asim Fazlagic, Mark Perkins, Brandon Saeger, “Transformer Upgrading Tools and
Practices”, Proceedings of the 2004 TechCon conference, January 26-29, 2004
16. Ramsis Girgis, Mark Perkins, Asim Fazlagic, “Evolution of ABB’s Transformer
Risk/Life Assessment Process”, Proceedings of the 2003 International Conference
of Doble Clients, paper no. TX1

379
17. Mark Perkins, Lars Pettersson, Nicolai Fantana, TV Oommen Steven Jordan,
“Transformer Life Assessment Tools and Methods”, Proceedings of the 2000
International Conference of Doble Clients, Section 8-1
18. M. Perkins, L. Pettersson, N. Fantana, TV Oommen, S. Jordan “Transformer Life
Assessment Tools with Specific Application to Nuclear Station Generator
Transformers” Proceedings of the IEEE EIC/EMCW Conference, Cincinnati, Ohio,
October 26-26, 1999, pp. 685-690
19. Asim Fazlagic, Mark Perkins, “Transformer Life and Condition Assessment -
Overview of Tools and Methods”, 2000 EUCI Condition Assessment of Power
Transmission & Distribution Systems Conference, December 6-8, 2000, Denver –
Colorado
20. Excerpts from: Dielectric Diagnosis of Electrical Equipment for AC Applications
and its Effects on Insulation Coordination – State of the Art Report, Presented by
Working Group 33/15.08, CIGRE 1990 Session, August 26- September 1, pp. 1-
34.
21. E.J. Rogers, L.E. Humbard, D.A. Gillies: “Instrumentation techniques for low-
voltage impulse testing of power transformers”, IEEE Trans. on PAS, May/June
1972
22. E.P. Dick, C.C. Erven, "Transformer diagnostic testing by frequency response
analysis”, IEEE Trans. on PAS, Nov. /Dec. 1978.
23. G. Zafferani, "Contribution of diagnostic techniques on large power transformers to
the Improvement of service reliability", Proc. of AEI Annual Meeting. Trieste, 1980,
Paper II.1.22
24. E. Dornenburg, W.Strittmatter, “Monitoring oil-cooled transformers by gas
analysis", Brown Boveri Review No. 5, May 1974.
25. CIGRE W6 15.01, “Detection of and research for the characteristics of an incipient
fault from analysis of dissolved gases in the oil of an insulation", Electra No. 42,
Oct. 1975.
26. G. Belanger, M. Duval, "Monitor for hydrogen dissolved in transformer oil". IEEE
Trans. on EI, Oct. 1977.
27. P.J. Burton, J. Graham et al., “Recent developments by CEGB to improve the
prediction and monitoring of transformer performance. CIGRE General Session
1984, Report 12-09.
28. W. Lampe, L. Petterson, et al., “Hot-spot measurements in power transformers",
CIGRE General Session 1984, Report 12-02.
29. R. Wilputte, M. Randoux, "Lessons drawn from the routine testing of insulating oils
used in power transformers on the Belgian network", CIGRE General Session
1986, Report 12-07.
30. J.F. Morel, G. Wind, ''Detection et localization des decharges partiellee par ultra-
sons", RGE No. 1, Jan. 1970.

380
31. R.T. Harrold, "Acoustic waveguides for sensing and locating electrical discharges
within high voltage transformers and other apparatus", IEEE Trans. on PAS, Apr.
1983.
32. CIGRE WG 12.01, "Measurement of partial discharges in transformers", Electra
No. 19, Nov. 1971.
33. AIEE Std. 505-1955, "Test code for power-factor (dissipation factor) testing of
power transformers"
34. ANSI/IEEE Std. 21 – 1976, "General requirements and test procedures for outdoor
apparatus bushings”
35. EEMAC Standard GL1-3 -1979, "Power transformer and reactor bushings"
36. IEC Publ. 137 (1984), "Bushings for Alternating Voltages above 1000 V”
37. Doble Client Committee on Circuit-Breakers and Bushings, "Bushing Field-Test
Guide”, Document BG661
38. A. Bargigia, M. De Nigris, C. Masetti, “Most recent developments in surge arrester
testing", Proc. 56th Int. Conf. of Doble Clients. Boston, Apr. 1989.
39. S.Shirakawa et al., "Maintenance of surge arresters by a portable leakage current-
detector", IEEE Trans. on PWRD, July 1988
40. J. Lundquist, L. Stenstrom et al., "New method of the resistive leakage currents on
metal-oxide surge arresters in service", IEEE PES Summer Meeting. Long Beach.
CA, July 1989, Paper 89SM817-8 PWRD.
41. Y. Du, M. Zahn, B.C. Lesieutre, A. V. Mamishev and S. Lindgren, ”Moisture
Equilibrium in Transformer Paper-Oil Systems”, IEEE Electrical Insulation
Magazine, Vol. 15, No. 1, pp. 11-20, January/February 1999.
42. J. Fabre amd A. Pichon, “Deteriorating Processes and Products of Paper in Oil –
Application to Transformers”, 1960 CIGRE Conference, Paris, France, Paper 137,
1960
43. L. E. Lundgaard et. al., ”Ageing of oil-impregnated paper in power transformers”,
2002 IEEE Power Delivery Transactions – Derived from data presented in Table 2
44. H. P. Moser, Transformerboard, Special Print in Scientia Electrica, Translated by
EHV-Weidmann Limited, St. Johnsbury, Vermont, Section C, 1979
45. W. A. Fessler, et. al., ”Bubble Formation in Transformers”, EPRI Report EL-5384,
EPRI, Palo Alto, CA, August 1987
46. FIST 3-5 Maintenance of Liquid Insulation: Mineral Oils and Askarels, US Dept. of
Interior, Bureau of Reclamation, 1992
47. T. V. Oommen, “Moisture Equilibrium in Paper-Oil Systems”, Proceedings of the
Electrical/Electronics Insulation Conference, Chicago, IL, pp. 162-166, October 3-
6, 1983
48. Tables are taken from IEEE Std. C57.106-2002

381
49. C. M. Bruce, J.D. Christie and P. Griffin, 55th Annual Conference of Doble Clients,
1988 Ed., Section 10-9.1.
50. Table 13, Transformer Maintenance Manual, FIST 3-30, United States
Department of the Interior, Bureau of Reclamation, October 2000
51. H. Moore, “Factors Affecting the Health and Life of Transformers”, Proceedings of
TechCon 2000, page 11.
52. P.M. Balma, R. C. Degeneff, H. R. Moore, L. B. Wagenaar, “The Effect of Long
Term Operation and System Conditions on the Dielectric Capability and Insulation
Coordination of Large Power Transformers”, IEEE Transactions on Power
Delivery, Vol. 14, No. 3, July 1999.
53. T.V. Oommen, “Bubble Generation During Transformer Overload”, EPRI Report
EL-6761, March 1990
54. T.V. Oommen, “Further Experimentation on Bubble Generation During
Transformer Overload”, EPRI Report EL-7391, March 1992
55. T.V. Oommen, E.M. Petrie and S.R. Lindgren, “Bubble Generation in Transformer
Windings Under Overload Conditions”, Paper 8E, Doble Client Conference, March
1995
56. T.V. Oommen and S.R. Lindgren, “Bubble Evolution from Transformer Overload”,
IEEE Transmission and Distribution Conference, October, 2001
57. United States Department of the Interior (Bureau of Reclamation), Transformer
Maintenance Manual, Facilities Instructions, Standards and Techniques, Volume
3-30, October 2000
58. John C. Drotos, John W. Porter, Randy Stebbins, “Dissolved Gas Analysis of
Transformer Oil”, published by the S.D. Myers Co., 1996
59. Martin Heath Cote Associates, “Maintenance of High Voltage Transformers”,
London, England, 1989
60. B. Fallou, “Detection and Research for the Characteristics of an Incipient Fault
From Analysis of Dissolved Gases in the Oil of an Insulation”, Elecktra, No. 42, pp.
31-52, 1975
61. IEEE Standard C57.104-1991 – Guide for the Interpretation of Gases Generated
in Oil-Immersed Transformers.
62. United States Department of the Interior (Bureau of Reclamation), Transformer
Maintenance Manual, Facilities Instructions, Standards and Techniques, Volume
3-30, October 2000
63. R. R. Rogers, “IEEE and IEC Codes to Interpret Incipient Faults in Transformers
Using Gas in Oil Analysis”, CEGB Transmission Division, Guilford, England, 1995
64. International Electrotechnical Commission (IEC 60599 Edition 2), Mineral Oil-
Impregnated Electrical Equipment in Service Interpretation of Dissolved and Free
Gas Analysis, 1999

382
65. T. V. Oommen, “Adjustments to Gas-in-Oil Analysis Data Due to Gas Distribution
Possibilities in Power Transformers”, IEEE Transactions on Power Apparatus and
Systems, Vol. PAS-101, No. 6 June 1982, pp. 1716-1722
66. United States Department of the Interior (Bureau of Reclamation), Transformer
Diagnostics Manual, Facilities Instructions, Standards and Techniques, Volume 3-
31, June 2003
67. T.V. Oommen and E.M. Petrie, “Particle Contamination in Oil-Filled Large Power
Transformers”, IEEE PES Winter Meeting, New York, January 31-February 5,
1982
68. T.V. Oommen, “Particle Counting Of Transformer Oil: Impact Of the New Standard
Based On Medium Test Dust Calibration”, Doble Conference Proceedings, March
2004.
69. ISO 4406-1999, “Hydraulic fluid power – Fluids – Method for coding the level of
contamination by solid particles”
70. ISO 11171, “Hydraulic Fluid Power – Calibration of Automatic Particle Counters
for Liquids”, Available from American National Standards Institute, 11 West 42nd
Street, New York, NY 10036
71. CIGRE Brochure N° 157, – Effect of Particles on Transformer Dielectric Strength,
November 1999
72. P.J. Sinz, Der Einfluss von Feuchte und Partikeln auf die elektrische Festigkeit
von Isolierölen. Ph.D work Technical University, Graz, Austria, 1990.
73. C. Vincent, N.G. Trinh, R. Olivier, J. Aubin, “Behavior of an oil-paper interface in
presence of carbon particle contamination,” IEEE International Symposium on
Electrical Insulation, Pittsburgh, USA, June 1994.
74. NETA Maintenance Testing Specifications for Electric Power Distribution
Equipment and Systems, Page 193
75. H. Horning, J. Kelly, S. Myers, ”Transformer Maintenance Guide”, Second Edition,
2001.
76. A. L. Rickley and R. E. Clark, “Transformer Exciting Current Measured with Doble
Equipment (A Progress Report)”, Doble Client Conference Report, Sec. 6:1001 -
1008, 1971.
77. Paul Gill, “Electrical Power Equipment Maintenance and Testing”, Marcel Dekker,
Inc., 1998.
78. ”A Basic Guide to Thermography”, A LAND Instruments International document
(www.landinst.com), 2004.
79. Z. Azmat and D. Turner, “Infrared Thermography and its Role in Rural Utility
Environment”, Rural Electric Conference, 8-10 May 2005, pp. B2/1-B2/4.
80. IEEE C57.119-2001, “Recommended Practice for performing temperature rise
tests”, page 10.

383
81. Bob Turnbull and Steve McConnell, Southern California Edison, "How Infrared
Thermography helps Southern California Edison Improve Grid Reliability",
InfraMation 2000 Proceedings Volume 1. Available on the internet at
http://www.flirthermography.com/media/027turnbull.pdf.
82. Niancang Houm “The Infrared Thermography Diagnostic Technique of High-
Voltage Electrical Equipments with Internal Faults”, International Conference on
Power Systems Technology (POWERCON), Vol. 1, 18-21 August 1998, pp. 110-
115.
83. Excerpts from ABB Contact, March 1998, Special No. 1.
84. Tord Bengtsson, et. al., “Acoustic Diagnosis of Tap Changers”, CIGRE, Paper 12-
101, 1996
85. Rick Youngblood, et. al., “Application of DGA to Detection of Hot Spots in Load
Tap Changers”, Minutes of the Sixtieth Annual International Conference of Doble
Clients, 1993, Sec. 6-4
86. Rick Youngblood, et. al., “Application of DGA to Detection of Hot Spots in Load
Tap Changers”, Minutes of the Sixtieth Annual International Conference of Doble
Clients, 1993, Sec. 6-4
87. M. Duval, “A Review of Faults Detectable by Gas-In-Oil Analysis in Transformers”,
IEEE Electrical Insulation Magazine, May/June 2002, Vol. 18, No. 3. pp. 8-17
88. F. Jakob, K. Jacob, S. Jones, R. Youngblood, “A Nomograph for Interpretation of
LTC-DGA Data, NETA Conference, San Antonio, March 15, 2004
89. Charles Sweetser and Tony McGrail, “Sweep Frequency Response Analysis
Transformer Applications – A Technical Paper from Doble Engineering”, Ver. 1.0,
January 2003.
90. Various SFRA Case Studies given on Doble Engineering Website
(www.doble.com), used with permission of Doble Engineering.
91. P. Picher et al., Final Report of Cigré WG A2.26, “Mechanical condition
assessment of transformer windings using frequency response analysis (FRA)”,
ELECTRA, No. 237, April 2008, pp. 35-45. CIGRE Technical Brochure No. 342,
April 2008. (67 pages) [ISBN 978-2-85873-030-8].
92. IEEE PC57.149TM/D1. “Draft trial-use guide for the application and interpre-tation
of frequency response analysis for oil immersed transformers”. March 2006.
93. ICS27.100, F24, document No. 15182-2005. “Frequency response analysis on
winding deformation of power transformers”. From the Electric Power In-dustry
Standard of People’s Republic of China, June 2005.
94. IEC Committee Draft 14/626/CD, “IEC 60076-18 Ed.1: Power transformers - Part
18: Measurement of frequency response”, Nov. 2009.
95. Simon Ryder, “Frequency Response Analysis for Diagnostic Testing of Power
Transformers”. Electricity Today Magazine Article, Issue 0601, 2006.

384
96. T. V. Oommen, “Electrical Insulation Materials and Applications”, ABB publication,
November 1989
97. A. M. Elmsley and G. C. Stevens, “ A Reassessment of the Low Temperature
Thermal Degradation of Cellulose”, 6th International Conference of Dielectric
Materials, Measurement Applications, UK, pp. 229-232, 1992
98. L. E. Lundgaard et. al., ”Ageing of oil-impregnated paper in power transformers”,
2002 IEEE Power Delivery Transactions.
99. T. K. Saha, “Review of Modern Diagnostic Techniques for Assessing Insulation
Condition in Aged Transformers”, IEEE Transactions on Dielectrics and Electrical
Insulation, Vol. 19, No. 5, October 2003
100 B. Pahlavanpour et al. , “Experimental Investigation into the Thermal-ageing of
Kraft Paper and Mineral Insulating Oil”, IEEE Int. Symp. On Electrical Insulation,
Boston, MA, pp. 341-345, 2002
101. D. J. T. Hill et. al., “A Study If Degradation Of Cellulosic Insulation In Power
Transformers – Part 2: Tensile Strength Of Paper”, IEE Proceedings On Scientific
Measurement Techniques, Volume 147, pp 285-290, 2000
102. H. P. Moser et. al., Transformerboard II, Rapperswill, Switzerland: H. Weidmann
AG, 1987
103. P. J. Burton et al. “Recent Developments by CEGB to Improve Prediction and
Monitoring of Transformer Performance, CIGRE paper number 12-09, 1984
104. Mineral Insulating Oils – Method for Determining 2-Furfural and Related
Compounds, IEC Document 61198, 1993.
105. T.V. Oommen, E. M. Petrie, R. B. van Breemen and C. A. Heney, “Analysis of
Furanic Compounds from Cellulose Aging by GC-MS and Attempts to Correlate
with Degree of Polymerization”, CIGRE Berlin Symposium, April 1993.
106. S. M. Ortel, “The Colorimetric Analysis of 2-Furfuraldehyde in Transformer Oil”,
Doble Conference Paper 10A-1, 1992.
107. B. Pahlavanpour and A. Wilson, “Analysis of Transformer Oil for Transformer
Condition Monitoring”, IEE Colloquim on An Engineering Review of Liquid
Insulation, Digest No. 1997/003, 1997
108. A. DePablo, “Furanic Compound Analysis: a Toll for Predictive Maintenance of
Oil-Filled Electrical Equipment”, CIGRE Task Force Report 15.01.03
109. X. Chendong et. al., “To Estimate the Ageing Status of Transformers by Furfural
Concentration in the Oil”, CIGRE Committee 33 Colloquium, Leningrad, Moscow,
1991.
110. A. DePablo, “Furfural and Ageing: How They Are Related”, IEE Power Division
Colloquium Insulating Liquids, National Grid Leatherhead, UK, 1999
111. B. Pahlavanpour, “Power Transformer Insulation Ageing”, CIGRE SC 15
Symposium, Sydney, Australia, 1995

385
112. A.B Shkolnik, K. M. Bigin and J.J. Kelly (SD Myers Inc.), “Creating a Preliminary
Model for Estimating Degree of Polymerization of Thermally Upgraded Paper
Based on Furan Concentrations in Transformer Oil”, Doble Conference Paper,
Boston, MA, April 1999
113. A. M. Emsley et. al., “Degradation of Cellulosic Insulation in Power Transformers.
Part 3: Effects of Oxygen and Water on Ageing in Oil”, IEE Proc. Sci.,
Measurement Techniques, Vol. 147, pp. 115-119, 2000
114. A. M. Emsley et. al., “Degradation of Cellulosic Insulation in Power Transformers.
Part 2: Formation of Furan Products in Insulating Oil”, IEE Proc. Sci.,
Measurement Techniques, Vol. 147, pp. 115-119, 2000
115. T.V. Oommen, “Cellulose Insulation in Power Transformers: How Long Shall We
Keep It?”, TechCon Conference Paper, San Diego, Jan, 2005
116. M. Perkins, G. Frimpong, A. Fazlagic, “Dielectric Frequency Response
Measurement as a Tool for Troubleshooting Insulation Power Factor Problems,
IEEE ISEI Conference Paper, Boston, MA, April 2002
117. U. Gafvert, G. Frimpong, J. Fuhr, “Modeling of Dielectric Measurements on Power
Transformers”, CIGRE 1998 Session Paper, 08/31/98
118. G. Frimpong, M. Perkins, A. Fazlagic, U. Gafvert, “Estimation of Moisture in
Cellulose and Oil Quality of Transformer Insulation Using Dielectric Response
Measurements”, Doble Client Conference, Paper 8M, Boston, MA, 2001
119. Derived from Chapter 9 of “Testing of Power Transformers – Routine Tests, Type
Tests and Special Tests”, ABB Publication, October 2003
120. IEC Standard 60270, Partial Discharge Measurements, 2000
121. IEEE Standard C57.113, IEEE Guide for Partial Discharge Measurements in
Liquid Filled Power Transformers and Shunt Reactors
122. From Power Diagnostix website - http://www.power-technology.com/ contractors/
pressure/power_diagnostix/power_diagnostix2.html
123. VDN-Störungs- und Verfügbarkeitsstatistik – Berichtsjahr 2004, Verband der
Netzbetreiber – VDN – e.V. beim VDEW
124. Transformer Lifetime Data Management, CIGRE Technical Brochure No 298 –
2006
125. IEC 60599, IEC Interpretation of the analysis of gases in transformers and other
oil filled electrical equipment in service, 2007
126. IEC 60567, IEC guide for sampling of gases and oil from oil filled electrical
equipment and for the analysis of free and dissolved gases, 2005
127. Duval M., Langdeau F., Gervais P. and Belanger G., “ Acceptable Dissolved Gas-
in-Oil Concentration Levels in Power and Instrument Transformers as a Function
of Age”, DOBLE Client Conference, 1989

386
128. Testing of Power Transformers, ABB Business Area Power Transformers, Pro
Print GmbH, Düsseldorf 2003
129. IEC 60 076-3, Power Transformers – Part 3: Insulation levels, dielectric test and
external clearances
130. C57.12.00, IEEE Standard General Requirements for Liquid-Immersed
Distribution, Power, and Regulating Transformers
131. C57.12.90, IEEE Standard Test Code for Mineral-Oil-Immersed Distribution,
Power, and Regulating Transformers
132. Dörnenburg et.al., “Betriebsüberwachung durch Untersuchung des Isolieröls“,
ETZ vol. 98, 1977 part 3, page 211 – 215
133. Duval M. et al., “Update on Hydro Quebec’s experience in the interpretation of
dissolved-gas analysis in HV Transformers”, CIGRE Symposium Berlin 1993,
Paper 110 – 14
134. Oommen et al., “New mechanism of moderate hydrogen gas generation in oil-
filled transformers”, CIGRE Paris 1998, paper 12-206
135. Recent developments on the interpretation of dissolved gas analysis in
transformers, CIGRE Brochure 296, 2006
136. Mechanical Assessment of Transformer Windings Using Frequency Response
Analysis (FRA), Interim-Report CIGRE WG A2.26, 2006
137. Mechanical condition assessment of Transformer Windings using frequency
response analysis (FRA), CIGRE Technical Brochure 342, 2008
138. A.J. Kachler et al.: „Kritische Anmerkungen zur Feuchtigkeitsbestimmung von
Transformatoren mit der »Recovery – Voltage – Methode“, Elektrizitätswirtschaft,
vol. 95 (1996) part 19, page 1238 - 1245
139. Dielectric response methods for diagnostics of power transformers, CIGRE
Brochure 254, 2004
140. V.D. Houhanesian, W.S. Zaengl: “Time domain measurement of dielectric
response in oil-paper insulation systems”. Presented at the 1996 IEEE Intern.
Symposium on Electrical Insulation, Montreal, June 16-19,1996
141. V.D. Houhanessian: “Measurement and analysis of dielectric response in oil-paper
insulation systems”. Ph.D. dissertation, Swiss Federal Institute of Technology
Zürich, ETHZ, 1998
142. Static electrification in large power Transformers, CIGRE Technical Brochure No.
170, 2000
143. Lars Jonnson, “Advanced transformer control and monitoring with TEC”, ABB
Review, Volume 4, 2004.
144. D.Chu, A. Lux, “On-Line Monitoring of Power Transformers and Components: A
Review of Key Parameters”, Proceedings of Electrical Insulation Conference and
Electrical Manufacturing & Coil Winding Conference, 26-28 Oct 1999, pp.669-675.

387
145. Trevor Lord and Graham Hodge (Lord Consulting, AVO New Zealand), “On-Line
Monitoring as a Strategic Tool to Enhance Supply System Reliability”,
http://www.avo.co.nz/techpapers/on-line-monitor/strategictool.php.
146. Rahmat Shoureshi, Tim Norick, Ryan Swartzendruber, “Intelligent Transformer
Monitoring System Utilizing Neuro-Fuzzy Technique Approach”, PSerc Publication
04-26, July 2004.
147. Tony Pink, Peter Stewart, “Power Transformer Control System Developments”,
TechCon 2004 North America.
148. Yong Liang, “Simulation of Top-Oil temperature for Transformers”, PSerc
Publication 01-21, Feb 2001.
149. P. Köstinger, T.M. Bruarøy: “Drying of power transformers in the field, applying the
LFH-Technology in combination with oil reclamation”, Cigre 2006.
150. V. Sokolov et al ,Moisture Equilibrium and Moisture Migration within Transformer
Insulation Systems, CIGRE WG 12.18, Dublin, 2001
151. B. Fallou R.G.E Tome 79 No8 Sept. 1970
152. H.P. Moser/ V. Dahinden Transformerboard II Weidmann 1987
153. V.W.Montsinger. Loading transformers by temperature“, Trans AIEE, April 1930,
pp 886-192
154. A.M.Emsley,, G.C.Stevens: “Review of chemical indicators of degradation of
cellulosic electrical paper insulation in oil-filled transformers”, IEE Proc. Sci. Meas.
Technol., Vol. 141, No. 5, September 1994, pp. 324-334.
155. J.Fabre, A.Pichon : “ Deteriorating processes and products of paper in oil.
Application to transformers”, CIGRÉ paper 137, (Paris), 1960.
156. H.P.Moser, V.Dahinden et al.: “Application of cellulosic and non cellulosic
materials in power transformers”, Cigre paper 12-12, Paris, 1986.
157. L.E.Lundgaard, W.Hansen, D.Linhjell: “Ageing of Kraft and thermally upgraded oil
impregnated paper “, Nord-IS 05, Trondheim, 2005
158. L.Lundgaard, W.Hansen, D.Linhjell and T.Painter, (2002) Ageing of Oil
Impregnated paper in power transformers”, IEEE Trans Pow.Del., Vol 19, No.1,
2004, pp 230-238.
159. L.E.Lundgaard, W.Hansen; S.Ingebrigtsen, D.Linhjell, M.Dahlund“ Aging of Kraft
paper by acid catalyzed hydrolysis“ ICDL, 2005, Coimbra
160. S.Ingebrigtsen , M.Dahlund; W.Hansen, D.Linhjell, L.E.Lundgaard: “Solubility of
carboxylic acids in Paper (Kraft)-Oil Systems”, CEIDP, Boulder, Colorado, 2004
161. P.Koestinger et al. Practical experience with the drying of power transformers in
the field, applying the LFH Technology, Cigre meeting Paris 2004
162. Ø. Berg et al., Experiences from on-site transformer oil reclaiming, Cigre meeting
Paris 2002

388
163. ABB Transformer Handbook, Rev 02, Sections 9.4-9.4.15, 2004.
164. IEEE C57.125 - IEEE Guide for Failure Investigation, Documentation, and
Analysis fro Power Transformers and Shunt Reactor.
165. ABB Transformer Handbook, Rev 02, Sect. 14, 2004
166. ABB Transformer Handbook, Rev 02, Sect. 3, 2004
167. CIGRE Brochure N° 221, “Improving The Impact of Existing Substations on the
Environment”, 2003
168. J. Vierengel et. al., “Use of Active Noise (ANC) Technology to Quiet Power
Transformers”, CIGRE Report N° 12-301, 1998
169. T. V. Oommen and C. C. Claiborne, “Biodegradable Insulating Fluid from High
Oleic Vegetable Oils”, CIGRE Report 15-302, 1998
170. An International Survey of Failures in Large Power Transformers in Service,
ELECTRA No. 88, pp. 21-48, 1983
171. Forced Outage Performance of Transmission Equipment, Canadian Electricity
Association Report for the Period January 1, 1998 to December 31, 2002,
pp. 80-95.
172. Brochure CIGRE N° 248-2004 – Guide on Economics of Transformer
Management.
173. C. Bengtsson, J-O. Persson, M. Svenson, Replacement- and refurbishment
strategies for transformer populations, CIGRE Colloquium SC 12, Dublin, 2001.
174. P. Boss, T. Horst, P. Lorin, K. Pfammatter, A. Fazlagic, M. Perkins, Life
Assessment of Power Transformers to Prepare a Rehabilitation based on a
Technical –Economical Analysis, Report 12-106, CIGRE SC 12, Session 2002.
175. L. Pettersson, N.L. Fantana and U. Sundermann, “Life Assessment: Ranking of
Power Transformers Using Condition Based Evaluation, A New Approach”,
CIGRE Paris Conference Paper 12-204, 1998 Session
176. L. Pettersson, J.O. Persson, N.L. Fantana and K. Walldén, “Condition Based
Evaluation of Net Transformers, - Experience from a New Ranking
Proceedure”, CIGRE Paris Conference Paper 12-108, 2002 Session
177. T. Grestad & K. Renström, "Experiences From Oil Reclaimed on Energized Power
Transformers", International Conference Power Transformers 6-7 April 2000, New
Delhi, India, Sect. III, p125-134
178. O. Berg, K. Herdlevar, M. Dahlund, K. Renström, A. Danielsen, U. Thiess,
“Experiences from on-site transformer oil reclaiming”, Paper 12-103, CIGRE Paris
Session 2002
179. P. Koestinger, E. Aronsen, P. Boss, “Practical experience with the drying of power
transformers in the field, applying the Low Frequency Heating Technology (LFH)“,
Paper A2-205, CIGRE Paris Session 2004

389
 INDEX

CARBON DIOXIDE ...... 96, 99, 104, 106

A
CARBON MONOXIDE ............... 96, 104
ACCESSORIES ................. 71, 149, 284 CELLULOSE .............................. 85, 204
ACCESSORY FAILURES ............ 64, 73 CHARGE GENERATION ............. 56, 57
ACETYLENE................ 96, 99, 103, 107 CHARGE RELAXATION .................... 56
ACID NUMBER ............................ 81, 82 CHENDONG .................................... 208
ACOUSTIC PARTIAL DISCHARGE. 288 CONDUCTIVITY .............................. 215
ACOUSTIC SIGNAL......................... 267 COOLERS........................ 288, 307, 339
ADVANCED DGA (ADGA) ............... 117 COOLING......... 23, 24, 27, 30, 285, 288
ADVANCED DIAGNOSTICS COPPER .......................................... 121
BROCHURE ................................. 437 Core ................................................. 188
AGING........................ 95, 205, 294, 295 CORE....... 19, 20, 28, 29, 143, 313, 327
AIR CLEANER FINE TEST DUST CORE GROUND . 110, 116, 142, 217,
(ACFTD) ............................... 118, 119 218, 219, 220, 313, 314
APPARENT CHARGE.............. 224, 267 INADVERTENT CORE GROUND 142
ASEA........................................ 320, 322 LAMINATIONS ............................. 142
AXIAL COLLAPSE ........................... 199 CORONA DISCHARGES......... 166, 336
AXIAL WINDING COLLAPSE .......... 199 CREEPAGE ................................. 25, 35
B D
BUBBLE EVOLUTION ................. 92, 93 De-ENERGIZED TAP CHANGER
BUBBLES......................................... 166 (DETC) ................................. 306, 313
BUCHHOLZ RELAY.. 97, 101, 240, 243, DEGREE OF POLYMERIZATION (DP)
245, 246, 247, 275, 310, 311 74, 113, 204, 205, 207, 208, 209, 210,
BUCKLING............................... 200, 201 295, 296, 314
BUSHING DEHALOGENATION
CONDENSER BUSHINGS ............. 32 OF OILS ....................................... 335
CORE ....................................... 32, 75 DEPABLO ................................ 208, 209
DISSOLVED GAS ANALYSIS (DGA) DETC ....................................... 189, 192
.................................................. 156 DIELECTRIC FREQUENCY
MOISTURE................................... 156 RESPONSE (DFR) ....................... 256
OIL SAMPLING ............................ 154 SIGNATURES ..... 216, 217, 218, 219,
POWER FACTOR MEASUREMENT 220, 221
.................................. 158, 161, 164 DISSIPATION FACTOR 64, 90, 91, 175,
TYPE T ......................................... 173 176, 211
TYPE U......... 164, 165, 167, 168, 169 DISSOLVED GAS ANALYSIS (DGA) 64,
VOLTAGE TAP............. 32, 33, 36, 37 82, 96, 97, 102, 106, 107, 108, 113,
C 117, 126, 252, 308, 310, 311
OIL SAMPLING PROCEDURE....... 97
CAPACITANCE... 24, 74, 135, 139, 159,
175, 176

390
 TOTAL DISSOLVED COMBUSTIBLE H
GASES (TDCG) .................. 96, 107 HEALTH ................................... 353, 373
DRYING HIGH VOLTAGE TESTING..... 325, 329,
INSULATION SYSTEM DRYING. 298, 330, 333
305, 328, 333 HIPOT .............................................. 318
E HOT OIL SPRAY HEATING ..... 298, 299
HYDROGEN ................ 74, 96, 102, 107
ECONOMICS
HYDROLYSIS
MANAGEMENT OF TRANSFORMER
OF CELLULOSE............................. 80
FLEETS .................................... 347
ELECTROMAGNETIC FIELDS (EMF) I
.............................................. 335, 336 INDUCTIVELY COUPLED PLASMA
ETHANE............................................. 96 (ICP) ............................................. 120
ETHYLENE ................................ 96, 107 INFRARED................. 74, 149, 150, 153
EXCITATION TEST... 74, 144, 145, 146, INHIBITOR ................... 82, 91, 300, 306
252 INSULATION RESISTANCE ... 131, 132,
F 133, 142, 249, 250, 275, 279, 304,
309, 313, 314
FACTORY REPAIR. 325, 326, 327, 328,
INSULDUR....... 105, 205, 206, 207, 209
329
INTERFACIAL TENSION 80, 81, 82, 90,
FAILURE ANALYSIS.......... 67, 237, 240
91
FAILURE RATE... 59, 60, 66, 68, 77, 78,
IRON ................................ 118, 120, 121
237, 324, 342, 343, 344, 345, 346,
348, 350 K
FANS.................................................. 18 KARL FISCHER ................................. 88
FAULT ANALYSIS ................... 237, 238 KRAFT PAPER ..... 72, 83, 85, 205, 207,
FAULT LOCALIZATION ........... 244, 309 209
FIELD SERVICES BROCHURE....... 429
FLASHOVERS ................. 257, 258, 261 L
Frequency response......................... 184 LEAKAGE CURRENT ........................ 78
FREQUENCY RESPONSE................ 74 LIGHTNING...................................... 329
Frequency response analysis........... 184 LOAD TAP CHANGER.... 40, 41, 42, 43,
FREQUENCY RESPONSE ANALYSIS 44, 45, 46, 47, 48, 49, 52, 53, 69, 79,
(FRA) .............................. 74, 255, 266 113, 117, 180, 285, 288, 307, 317,
FURANS ............................ 82, 207, 284 318, 319, 320, 321, 322, 323, 343,
G 344
ACOUSTIC SIGNAL ..................... 178
GAS CHROMATOGRAPHY............... 74 ARCING TAP SWITCH................... 40
GASKETS ........................ 307, 321, 323 DISSOLVED GAS ANALYSIS (DGA)
GENERATOR STEP-UP .......................................... 179, 181
TRANSFORMER .......................... 242 DIVERTER SWITCH ...................... 40
GROUND RESISTANCE ......... 143, 219 GE TYPE LR300........................... 318
GROUNDED SPECIMEN TEST (GST) GE TYPE LR400........................... 318
..... 134, 136, 137, 138, 140, 159, 163, GE TYPE LR500........................... 318
217 MOTOR CURRENT...................... 178
GROUNDED SPECIMEN TEST WITH MOTOR DRIVE ............................ 307
GUARD (GST/G) . 134, 136, 137, 138, PREVENTIVE AUTO .......... 41, 42, 43
140, 159

391
 REACTANCE TYPE ........... 41, 42, 43 OIL RECLAIMING ... 294, 295, 297, 300,
REINHAUSEN ...................... 320, 322 301, 306
RELAY TIMING ............................ 179 OLTC........................................ 188, 192
RESISTANCE TYPE ................ 44, 45 ON-SITE COIL REBLOCKING
STENESTAM RATIO (DGA)......... 180 BROCHURE ................................. 449
TAP SELECTOR ............................ 53 ON-SITE REPAIR .... 326, 327, 328, 330
WESTINGHOUSE TYPE UTH...... 319 OOMMEN................................... 87, 112
WESTINGHOUSE TYPE UTT ...... 319 OSTWALD COEFFICIENT............... 104
WESTINGHOUSE TYPE UVT...... 319 OVERHEATING
WESTINGHOUSE TYPE UVW..... 319 CELLULOSE ........................ 100, 106
LOAD TAP CHANGER ENGINEERING, Local ............................................... 73
PARTS & SERVICE BROCHURE 445 TERMINAL ........................... 168, 171
LOAD TAP CHANGER RETROFITS OXIDATION ................................. 81, 91
BROCHURE ................................. 443 OXYGEN................ 82, 96, 99, 101, 312
LOW FREQUENCY HEATING. 299, 305 P
M PAPER CONTAMINATION .............. 220
MAINTENANCE .... 36, 66, 92, 294, 302, PARTIAL DISCHARGE ... 33, 46, 74, 75,
303, 304, 305, 306, 307, 315, 317, 83, 100, 111, 112, 116, 157, 223, 224,
318, 319, 320, 322, 323 226, 227, 229, 231, 233, 254, 265,
MATURE TRANSFORMER 267, 284, 329, 330
MANAGEMENT BROCHURE....... 433 ACOUSTIC ................................... 233
MECHANICAL STRESS .................... 77 PARTICLE ANALYSIS ............. 121, 304
MECHANISM ......................... 43, 47, 53 EFFECT ON DIELECTRIC
MEGGER ......... 126, 129, 131, 142, 143 STRENGTH .............................. 122
METHANE.................................. 96, 102 PARTICLE COUNT . 118, 119, 121, 122,
MIGRATING INK .............................. 166 123, 125
MOISTURE ... 74, 77, 83, 84, 86, 87, 88, POLARIZATION....... 133, 249, 256, 270
89, 92, 93, 94, 95, 183, 215, 219, 288, POLARIZATION CURRENTS .......... 256
291, 305, 315, 332, 388 POLARIZATION INDEX ................... 133
MOLONEY ............................... 320, 429 POLYCHLORINATED BIPHENYLS
MONITORING. 284, 285, 286, 288, 290, (PCB).............. 83, 169, 173, 334, 335
291, 292, 293 PORCELAIN .................................... 153
MTD METHOD OF PARTICLE POWER FACTOR33, 82, 140, 165, 304,
COUNTING .......................... 118, 119 314, 315, 316
TAP......................................... 34, 316
N
TIP-UP.......................................... 141
NEUTRALIZATION NUMBER ............ 81 PRESERVATION SYSTEMS ........... 101
NITROGEN ........................................ 96 PRPDA..................................... 227, 266
NOISE
R
CONTROL METHODS ................. 338
RECOVERY VOLTAGE METHOD
O
(RVM) ........................... 256, 266, 270
OIL IMMERSED TRANSFORMERS 126 REFURBISHMENT .......................... 324
OIL LEVEL INDICATOR................... 311 REGENERATION............................. 306
OIL PRESERVATION ...................... 101 REMANENT FLUX ................... 277, 278
OIL QUALITY ....................... 79, 89, 250

392
REPAIR... 105, 150, 305, 311, 324, 327, TAP SELECTOR . 42, 43, 44, 48, 52, 53,
458 258, 318
RISK ASSESSMENT . 60, 355, 370, 371 TEMPERATURE . 18, 79, 88, 90, 93, 95,
ROGERS RATIO.............. 110, 111, 112 103, 132, 175, 178, 288, 299
RUST ............................................... 259 INDICATORS................................ 244
THROUGH-FAULT. 67, 69, 70, 142, 312
S
TRACKING............................... 116, 319
SAFETY .... 39, 300, 302, 303, 318, 334, TRANSFORMER RELOCATION
335, 340, 349, 353, 354, 355, 356, BROCHURE ................................. 453
357, 359, 360, 361, 362, 364, 366, TRANSFORMER
373, 375, 377 REMANUFACTURING BROCHURE
SAMPLING.. 97, 98, 109, 118, 156, 321, ...................................................... 425
322 TRANSFORMER TURNS RATIO (TTR)
SATURATION .............................. 89, 90 41, 128, 129, 130, 249, 250, 251, 275
Short circuit ...................................... 190
U
SHORT CIRCUIT ....... 73, 116, 309, 310
SHORTED TURNS .......................... 202 UNGROUNDED SPECIMEN TEST
Single phase transformer ................. 189 (UST) ..... 36, 134, 136, 137, 139, 140,
SLUDGE ...................... 81, 86, 259, 300 158, 163, 217
SPECIFIC GRAVITY .......................... 80 V
SPRING MECHANISM....................... 35
STANDARDS ............... 32, 33, 158, 329 VACUUM DRYING........................... 298
STATIC CHARGE ............................ 262 W
STREAMING ELECTRIFICATION
ELECTROSTATIC CHARGING WATER ... 79, 83, 84, 85, 86, 89, 90, 91,
TENDENCY (ECT).......... 55, 56, 58 92, 205, 318
SUBSTATION .......................... 338, 374 WINDING RESISTANCE 126, 127, 142,
SURFACE PROTECTION................ 308 249, 251, 275, 279
SURGE ARRESTERS................ 76, 244 Windings .......................... 190, 191, 192
WINDINGS....................... 131, 199, 313
T
X
TANK........................ 136, 140, 284, 317
X-WAX ..................................... 262, 263

393
ABB Transformers Service
General Brochures

395
396
Insert here the 7 pages general brochure

(cover page but not the last back page with ABB contacts)

397
398
399
400
401
402
403
404
ABB Transformers Service
Product Leaflets

405
Insert here a blank page, then the following leaflets

Leaflets: Page Page

- Condition Assessment 407 408

- Spare Parts 409 410
- Drying 411 412
- Oil reclaiming 413 414
- Repair and
upgrade 415 416
- Increase power - Nomex 417 418
- On-Site-Repair 419 420
- On-line monitoring 421 422

406
ABB TRES North America Service Brochures

423
424
Transformer Remanufacturing

ABB Factory Services

Help Protect Your Important
Power Transformer
Investments
When a transformer goes down, one of your
major assets has stopped producing income but is
still accruing costs. ABB knows our clients need
their assets up and working in the shortest
possible time to produce the power that customers
need.
Our facilities in St. Louis, Missouri provide full
capabilities to service power transformers,
including dismantle, design, repair, rebuild, and
comprehensive electrical testing. As the largest
transformer manufacturer in the world, ABB
knows the challenges our clients face and how to
meet them.
Like many businesses, utilities and industry are
asked to do more with less:
• Produce more power without adding
transformers
• Maintain the same physical plant with fewer
resources
• Improve competitive position with lower
O&M and capital cost
Repairing a power transformer, instead of
replacing it, can lower capital and maintenance
costs dramatically and provide a quicker
turnaround than buying a new one. This results
in clients realizing a higher return out of the
investments already made. With top quality
work backed by comprehensive OEM
technology, years of repair experience, and the
largest transformer design database in the
world, ABB offers superior reliability – resulting
in less time, money, and personnel that have to
be devoted to maintenance.

425
Transformer Remanufacturing

Power Transformer
Remanufacturing Resources
at a Glance
Design Documentation and
Drawing Database
ABB maintains the most extensive OEM
database in the transformer industry with
design documents on file from Asea, BBC,
General Electric, Moloney, and
Westinghouse.
Resources include:
• Millions of design drawings for most
transformers installed in the US
• More than 80 design, materials, and
project engineers and technicians
dedicated to customer field support
• Engineering studies
• Core form and shell form capability
Shipping and Receiving
When moving power transformers, ABB
uses only highly qualified riggers and
haulers that meet ABB’s stringent ISO9001
criteria for heavy hauling. We track the
transformer during shipments and provide
risk of loss insurance to protect against any
damage during the transportation process.
A fleet of 14 rail cars including five
special Schnabel cars and six heavy-duty
depressed center cars, as well as off-road
modular platform trailers, 400-ton,
Schnabel trailers are available to meet
these stringent transportation requirements.

426
Major Plant Equipment
and Systems
ABB facilities in St. Louis, Missouri are
ISO 9001 registered to meet the repair
requirements for most large power
transformers.

Major equipment and systems include:

• Multiple Georg lines for core cutting
• Disassembly isolation/remediation area
• 360 tons of overhead crane capacity
• Core stacking tables for disassembly, inspection,
and assembly
• Core form winding areas that include horizontal,
vertical, and GE layer machines
• Insulation fabrication
• Dry-process ovens for newly-manufactured windings
• VapoTherm processing
• Paint shop and specialty shipping department

Testing
The ABB state-of-the-art testing facilities
ensure accurate and repeatable test
results. At the St. Louis plant, ABB has
full dielectric and thermal testing
capability up to 345 kV, 1050 KV BIL,
and 300 MVA.

Testing equipment includes:

• Impulse tests to 1050 kV BIL
• Haefeley oscilloscopes for impulse measurement
• Tettex and Haefeley equipment for resistance and
losses measurements
• Tettex transformer turns ratio (TTR) test set
• Hipotronics 800 kV DC applied test set
• Corona measurement system
• Temperature tests with gas in oil

427
Field Services for Power Transformers

Servicing Power
Transformers On Site
Power transformers are valuable assets,
and companies —utility as well as indus-
trial—have every incentive to keep them
running reliably for as long as possible.
ABB has what it takes to service your
power transformers. ABB TRES
(Transformer Remanufacturing and
Engineering Solutions) is uniquely qualified
to test, analyze, and improve upon any
power transformer, regardless of original
manufacturer. For any requirement, from
a single OEM part, to a maintenance
program, to a long-term Asset Manage-
ment program, ABB’s TRES is the best
choice to keep your power transformers
online and operating efficiently, 24/7.
ABB has the most extensive OEM data-
base in the transformer industry with
design documents on file for Asea, BBC,
ABB, General Electric, Moloney, and
Westinghouse. ABB offers many upgrades
that exceed original operation specifica-
tions. We fabricate and maintain a huge
inventory of OEM parts as well.

429
Power Transformer Field Service
Resources at a Glance

Planning:
Planning is a critical step in a successful
transformer service job. Identifying poten-
tial obstacles, conducting route surveys to
determine transportation requirements,
and doing engineering assessments are
only a few of the many steps taken in the
planning phase. Utilizing the millions of
transformer drawings and years of exper-
tise, the 100+ professionals at ABB ensure
that your transformer arrives on schedule
in a safe and cost effective manner.

Transportation:
ABB provides logistical service and support
to transport, relocate, coordinate, and Installation:
install transformers. When moving your
ABB installs over 100 transformers annually. Our highly qualified
power transformer, ABB uses only our
technicians and top quality equipment provide safe, accurate,
network of highly qualified riggers and
rapid completion of the job. ABB’s fleet of mobile installation
haulers. We track the transformer during
and oil-processing trailers can be dispatched to any job site in
shipments and provide risk of loss insur-
North America.
ance to protect against any damage during
the transportation process. To support this
business, ABB maintains a fleet of 14 rail
cars including five special Schnabel cars
and six heavy-duty depressed center cars
as well as off-road modular platform
trailers and Schnabel trailers.

430
Maintenance and Retrofit:
ABB maintains 15 mobile work trailers with the finest
ABB TRES provides preventative and
installation equipment available, including:
corrective maintenance support for your
installed equipment. Determining the best • Semi-tractor with 15-ton hydraulic crane
maintenance program often begins with a • Motor-generator set
performance and capability study
conducted by an ABB design engineer. • Air compressor and air dryer
Recommendations are made in a formal • De-gassing equipment, 1800-4800 GPM
report to the operator. Many repairs and
• Liquid nitrogen cold traps
retrofits can be successfully performed
onsite, avoiding the cost associated with • Cutting and welding equipment
transformer movement. Retrofit and • OSHA required PPE and safety equipment
upgrade of transformer capabilities are
available in the areas of bushings, tap • Hand tools, rigging, and lifting fixtures
changers, thermal/cooling, controls, • Electrical test equipment
reblocking, and others.
• Oil test equipment

Repair:
ABB’s modern manufacturing and repair MTMProgram – ABB’s Mature Transformer
facilities in St. Louis, Missouri serve the Management Program:
industry when transformers must be
Many transformer fleets in North America are approaching the
removed from the substation for repairs.
end of their design lifetime, increasing the risk of unplanned
These facilities are ISO 9001 registered
outages. These outages can be catastrophic in cost, loss of revenue,
to meet the repair requirements for
and environmental and social impact. ABB TRES can identify and
large power transformers. ABB provides
quantify risk fleet wide. Condition and Life Assessments are used
full capabilities to service power
to extend unit lifetime and optimize transformer performance.
transformers, including dismantle,
With this structured and rigorous asset management program,
r-design, repair, rebuild, and
customers can prioritize maintenance and capital budgets to
comprehensive electrical testing. ABB’s
insure maximum economic benefits.
state-of-the-art testing facilities ensure
accurate and repeatable test results. ABB
has full dielectric and thermal testing
capability up to 765 kV, 2050 BIL, and 1000
MVA.

Testing and Diagnostics:

In addition to standard testing equipment
and capabilities, ABB offers advanced
testing and diagnostic analysis in a variety
of areas, including: frequency response
analysis, dielectric frequency response,
ABB DGA analysis, furanic, and degree of
polymerization testing. More important
than the actual test is the analysis and
interpretation of the results by ABB
transformer design and testing experts.

431
Mature Transformer
Management Program

MTMProgram™

433
Introduction
The MTMProgram™ addresses the emerging issue As the largest OEM in the world with a unique
of power transformer aging. While this subject has history, ABB is often entrusted with diagnosing
been a topic of discussion for over two decades, and assessing the condition of various transformers.
the strength of our transmission infrastructure has ABB has detailed historical transformer design
allowed transformers to operate beyond their knowledge of nearly 75% of the installed base
design life. of large power transformers in North America,
including Westinghouse, GE, Asea, Brown Boveri,
Deregulation and economic controls have changed
and other predecessor technologies. The ABB
the demands of new systems. These transformers
assessment processes and technical tools devel-
that are operating beyond their design life will
oped are the natural result of our distinctive
soon begin to fail in significant numbers.
market position.
This issue can be addressed by identifying weak
units through an economical approach that deals ABB has developed a four-step process (Figure A):
with the dynamics of the aging process.
Transformer Fleet Screening
Transformer Design and Condition Assessment
Life Assessment/Profiling
Implementation of Engineering Solutions

Figure A
Steps of the ABB Mature Transformer Management Program

STEP ONE
Transformer Fleet Screening

STEP TWO
Transformer Design and Condition Assessment
Rigorous, unit-based design evaluation
Risk of Failure Screening and condition assessment.

STEP THREE
Life Assessment/Profiling
Detailed life assessment, including design
and engineering analysis for life extension
of each specific transformer.

STEP FOUR
Implementation of Engineering Solutions

Replacement R efurbishment Cooling Upgrade

434
ABB’s Four-Step Process

STEP ONE STEP TWO

Transformer Fleet Screening Transformer Design and Condition

The fleet risk assessment process is the first or Assessment
precursor step in the transformer life management The transformer design evaluation and condition
process. This process is used to sort through the assessment process is a paramount step in the
readily available analytical and statistical informa- transformer life assessment and asset management
tion about each transformer in the fleet. This way, approach. It requires rigorous, state-of-the-art analysis
accurate and intelligent action plans can be made methods/rules typical to the design, testing, and
regarding the future of each transformer and the quality assurance tools presently used by ABB
fleet as a whole. The goal is to prioritize action designers in the design and manufacture of both
plans for the transformers in the fleet and identify core form and shell form transformers.
transformers that are candidates for detailed
The assessment combines input from the design
design and condition assessments (Figure B). Fleet
assessment, historical loading, operational history,
screening will be performed each year as part of
and also routine and advanced diagnostic data
the program.
(Figure C).

Figure B Figure C
Typical results of the transformer fleet Typical results of the short circuit strength
risk assessment analysis used in a life assessment study.

Little Risk of Failure

Slight Risk of Failure

High Risk of Failure

HV Radial HV Axial LV Radial LV Axial LTC LTC

(hoop) (tipping or (buckling) (tipping or Winding Winding
crushing) crushing) Radial Axial
(buckling) (tipping)

Relative Importance
Design #1 Design #3
Design #2 Design #4
Urgent Action Needed

Preventive Action Needed

Normal Maintenance Action Needed

A B B M A T U R E T R A N S F O R M E R M A N A G E M E N T P R O G R A M

435
ABB’s Four-Step Process (Continued)

STEP THREE STEP FOUR

Life Assessment/Profiling Implementation of

Engineering Solutions
Life assessment/profiling is a unit-based ranking
that typifies the risk of failure based on results Based on the results of this rigorous analysis
of the design and condition assessment. This is a program, engineered solutions are prescribed
more detailed and precise risk of failure estimate to achieve risk reduction, life extension, and in
than a fleet risk assessment since it focuses on general, health improvement of the fleet. Such
specific knowledge of the transformer design and engineering solutions options include:
actual condition assessment. The resulting ranking
provides an accurate estimate of the risk of failure Preventative and Corrective
for the transformer, which can be compared to Maintenance Activities
the expected failure risk of other transformers in Field Repair and Retrofit Solutions
the industry. Relocation and Transportation
Once the status of a unit is known, recommenda- Testing and Advanced Diagnostics
tions for maintaining, improving the unit condition,
or establishing a contingency can be prioritized. Factory Repair Solutions
The result is the elimination of high-risk units Planned Transformer Replacement Solutions
without increasing budget expenditures.

Conclusion
Clearly, capital requirements for wholesale unit higher risk positions. As the health of your fleet
replacement are not economically feasible. However, continues to improve annually, ABB anticipates
neither is the revenue loss resulting from a doubling your operations, and maintenance budgets will
or tripling of system failures. Therefore, the logical begin to decrease, resulting in over-all savings for
approach is to continually identify the “weak system your company.
units” and replace, remanufacture, retrofit, or repo- ABB’s MTMProgram is distinctive in the trans-
sition these critical assets on the system before they former industry. Because we possess nearly 75%
remove themselves from service. of the detailed transformer design database in the
As this system is followed over multiple years, industry, only ABB can provide an analysis based
maintenance and operating budget dollars will not just on statistics, but based on this specific
be directed to the highest risk units. Remaining design knowledge. Consequently, our clients get
expenditures can then be directed to medium- the most accurate, detailed analysis possible. The
risk areas. The result is a dynamic that moves ABB MTMProgram provides a technical, economic,
high-risk units into lower risk areas and prevents optimized solution to improve the reliability and
the migration of low- or medium-risk units into availability of your transformer fleet.

436
ABB Advanced Diagnostic Testing Services

ABB Transformer Repair and Engineering Services

437
Advanced Diagnostic Testing Services
from the World’s Leading Manufacturer
of Power Transformers

ABB leadership begins with our unmatched experience

in the power transformer industry. ABB Transformer and
Repair Services (TRES) experts have a detailed design and
development knowledge of Westinghouse, General Electric,
ABB, Asea, BBC, National Industries, Moloney, and other
transformers built over the years. As the successor OEM,
ABB has the original design information for these units. In
addition to these designs, ABB can draw on broad industry
knowledge gained from years of transformer service and nation of knowledge, diagnostic tools, and field experience
repair experience of all manufacturers’ transformer designs. is what makes ABB the acknowledged leader in the field of
What’s more, ABB utilizes the most up-to-date-design Advanced Diagnostic Services.
programs and design practices in the industry. This combi-
ABB can diagnose your power transformers and provide you
with the information you need to make effective O&M decisions.

Dielectric Frequency Response —

More actionable information than
standard power factor tests.

The Dielectric Frequency Response (DFR) test is used to

assess the integrity of a transformer’s insulation system.
The test determines the moisture level or presence of other
contaminants in the solid insulation and the oil conductivity
simultaneously with oil power factor. Standard power factor
tests alone do not yield this information. This is an extreme-
ly useful tool in an overall condition assessment program.

The DFR test is a measurement of the dielectric properties abnormal power factor results were measured in the field.
(i.e. capacitance, loss, and Power Factor) of the transformer’s
insulation as a function of frequency. This offline test The DFR test has gained popularity in recent years as
utilizes the same type of connections as the standard 60 Hz a diagnostic tool for transformer insulation system testing.
insulation power factor test. However, it covers a frequency One important primary use of the test has been for deter-
range, typically from 1 mHz to 1,000 Hz while the standard mining the moisture content in the cellulose insulation
power factor (Doble) test is done only at 60 Hz. This test structure of power transformers. The analysis of moisture
yields more information with increased sensitivity to insula- in transformers is performed using the results of the DFR
tion issues by utilizing the dielectric response phenomenon. measurement and an analysis tool that models the actual
insulation geometry and the insulating material (oil, paper,
A further application is the Dielectric Frequency Response pressboard, etc.) of the transformer. ABB has used these
Signature method (DFRS), where the signature of the mea- tools for years in the analysis of transformers, both in the
sured response is then compared with a modeled response factory and in the field. The experience gained has shown
of a transformer with a “normal” insulation structure the potential of the DFR test for identifying not only mois-
and a library of signatures of known defects. The method ture problems, but also other defects in the transformer
is demonstrated by utilization in cases where high or insulation structure.

438
Demonstrating the value of DFR: Case Study #1
Better information for client resulted in avoidance of unnecessary maintenance.

Our client provided ABB with a list of seven transformers. Xfrmr# Temp Type Construction Oil Cond Moisture by Oil Moisture by DR
(°C) (pS/m) Sat (%wt) (%wt)
In each case, moisture-in-oil test results indicated the need
1 23 GSU Core 0.381 2.5 0.9
for oil processing and drying. Working with the client,
2 28 GSU Core 0.492 1.8 0.9
ABB performed DFR testing and determined that only two 3 23 GSU Core 0.412 1.4 0.9
units actually required drying, instead of seven. ABB’s 4 23 GSU Core 1.34 2.8 0.7
recommendation to dry two transformers, while carefully 5 13 3-wdg Shell 1.5 * 1.2

monitoring the other five, afforded a significant amount of 6 27 Auto Core 3 3.5 2

O&M savings. This also avoided over-drying and loosening 7 27 Auto Shell 0.3 3.3 1

the windings of several units.

Surface moisture in paper, estimated from moisture in oil,
compared with volume moisture in insulation measured by DFR

Demonstrating the value of DFR: Case Study #2

Carbon tracking of the insulation system identifies a problem
that could have produced catastrophic failure.

This transformer was investigated due to combustible gas

results in the field. The DFR test was done as part of a
condition assessment of the transformer to determine the
cause of the gassing. The power factor results were all
within the normal industry practice (less than 0.5%). The
DFR result showed a noticeable deviation from normal
responses. ABB’S Dielectric Frequency Response Signature
method (DFRS) was used to investigate the cause of the
abnormality. Comparisons were made to library DFRS cases,
and carbon tracking or contamination was identified as
the cause of the deviation. The chart shows a comparison
of the transformer DFRS result to a case of known contami-
nation. The fact that there was much less deviation than
the library case indicated that the extent of the defect
was very limited.

Based on ABB Dissolved Gas Analysis and the suspected

carbon tracking identified by the DFRS method, the
transformer was disassembled for close inspection of the
insulation structure. An area of burning of the winding
was found in the high-low space of one phase of the
transformer. The carbon was probably caused by a partial
discharge. The photograph shows the small area where
carbon was produced. Ultimately, a catastrophic failure
was avoided.

439
ABB’s Advanced Dissolved Gas Analysis Software —
Deeper analysis that comes from years of ABB experience.
ABB has developed an internal software package that
combines DGA raw data, ratios, trending, key indicators,
and ABB’s resident design expertise and transformer
construction knowledge to interpret the results. By
combining ABB’s design and manufacturing knowledge
with the analysis capabilities of the software, we offer
greater analytical depth than what is standard practice
in the industry.

ABB’s design database of Westinghouse, GE, Asea, BBC,

Moloney, National Industries, and other industry-
standard transformers, combined with our modern
transformer design capabilities, put ABB in the unique
position of being able to offer this proprietary software.
The ABB Dissolved Gas Analysis, combined with our
design knowledge and factory and field testing
experience, give us the ability to pinpoint specific
sources and causes of gas generation, rather than simply
identify general categories of gas generation.

440
Frequency Response Analysis —
An important tool for identifying potential winding geometry changes.

Frequency Response Analysis (FRA) is a low-voltage, ABB’s expert interpretation of FRA test results is an
offline measurement of the impedance of the transformer excellent method to check movement or displacement
winding as a function of frequency. The test is performed of windings or other internal winding circuit changes and
by applying a variable frequency AC voltage to each is much more effective than the low-voltage impedance
individual winding of the transformer and measuring test routinely performed on transformers. The value of
the current that flows out of the winding. The plot of the FRA test is to identify potential winding geometry
the current divided by the voltage vs. frequency is changes that may affect the ability of the transformer to
known as the Swept Frequency Response Analysis withstand through- faults, helping to avoid catastrophic
(SFRA) of the winding. transformer failures.

ABB recommends the FRA test be performed in the

factory at the time of original transformer testing to
provide a baseline reading of the windings in an as-
new condition. For installed transformers, a test in the
field can be used to provide the baseline value. FRA
should be performed periodically during the service life
of the transformer or after a specific incident producing
significant through-fault currents in the transformer.
Comparison of such an FRA test to the original baseline
value is very useful in diagnosing the condition of
the windings.

Winding defects shown in coil displacement

441
Load Tap
Changer
Retrofits
Application
ABB load tap changer (LTC) replacements
or retrofits are designed to economically
solve operating problems inherent in existing
tap changers, such as:
LTC mechanisms that have reached the
end of their useful mechanical life
The rapid diminishing of parts availability Before Retrofit
and qualified service personnel on
vintage equipment
Forced reliance on expensive and
sometimes questionable quality parts due
to obsolete design or being built by
former manufacturers
Cost of unplanned outages has risen
dramatically

Features
All mechanical parts are designed and
tested for demanding service
Single source for all parts and services
Proven, reliable vacuum or resistive
technology
After LTC Retrofit
Up-to-date tap changer designs
Backed by over 80 years of transformer
engineering and design experience
All engineering, drafting, field
modification drawings, new controls,
wiring diagrams, parts, materials, and
even installation can be provided

Benefits
Increased transformer life up to 25
years in some cases
Reduced maintenance costs Troublesome LTC Removal Field Installation of Replacement LTC

Improved reliability of operation,

particularly under demanding conditions
Increased contact live available from
modern, proven tap changers
Field retrofit is an economical alternative
to shipping large power transformers
Same downtime as same-model, active
mechanism replacement

443
Load Tap Changer Retrofits
Because of the wide variety of flange sizes, lead
configurations and lengths, operating
mechanisms, and control differences, field
retrofitting has generally been considered to be
impractical. ABB has proven that it is both
possible and economical to retrofit an existing
transformer with a new production model load
tap changer. The transformer outage typically
required for a retrofit is only two weeks, and all
transformer work is completed at the site.

ABB has designed and installed over 35

LTC retrofits on different makes of
transformers, demonstrating the ability to
perform custom engineering and service
work with proven, predictable results.
Projects include:
Installation of Westinghouse UVT to
non-LTC transformer
Westinghouse URT UVT
Federal Pacific TC-25 UVT
W URT-HC Reinhausen RMV-I
GE LRT Westinghouse UVT
Ferranti Packard RT UVT
Moloney MC Reinhausen RMV-I
FP TC-25 Reinhausen RMT-I
FP LR-525 Reinhausen RMT-I
GE LRT-500 Reinhausen M (in-tank)
Westinghouse URH Reinhausen M
Westinghouse UNR RMV-II
GE LR-83 Reinhausen RMV-II
Westinghouse URT RMV-II Load Tap Changer Retrofit Assembly is Shipped for Field Installation

Many customers have found retrofits to be a technically sound as well as an economical alternative to the purchase
of a new transformer. Contact ABB or your local representative for more information.

444
Load Tap Changer (LTC)
Engineering, Parts, & Service

ABB Transformer Remanufacturing & Engineering Solutions (TRES)

445
 How ABB Can Help
Proper Load Tap Changer ABB offers a comprehensive and
systematic approach to effectively
(LTC) maintenance has maintain many manufacturers’
become increasingly LTCs, including ABB, Westinghouse,
Reinhausen, Moloney, GE, and
complex even for the
others. Our LTC program can
best utility and industrial supplement existing in-house
programs or work independently
managers. Factors resulting
to satisfy the total maintenance
from interconnections and need. ABB’s LTC team provides
power networks to aging support 24/7, and all eld services
are furnished by the ABB quali ed
equipment and the declining technicians and engineers.
pool of technical experts For services ranging from
add to the challenge. When documentation and design
engineering, to the latest in testing
you evaluate the liability and oil processing, to providing
of a single LTC failure at 100% OEM grade parts and life
extension kits, ABB is the best
today’s costs, establishing
resource for all LTC service
an effective maintenance requirements. Our LTC services
Documentation
increase transformer life, reliability,
strategy becomes a matter and Engineering
and lower overall maintenance
no one can ignore. and operations costs. The ABB design engineers and
technicians specialize in LTCs and
the LTC interface with control
circuits and power transformers.
With access to an original
OEM grade renewal engineering documentation and
parts for LTCs are procedures database containing
available from ABB more than seven million drawings
on an immediate for most current LTC designs, our
basis anywhere in dedicated service professionals are
North America. When ready to help.
necessary, obsolete
Services provided are:
parts can be made
Establishing proper test
by ABB because we
procedures
maintain over
Identifying parts
seven million
Setting proper operating
drawings to assure
parameters
that LTC services
Performing studies to engineer
conform to the
improvements
original design
Troubleshooting
specifications.
Failure analysis

446
 Original Retrofit and Upgrade Kits
Replacement Parts ABB is continuously monitoring
Whenever feasible, ABB and searching for LTC retr t and
upgrade opportunities where our
procurement specialists restore
an LTC to original speci cations designs can be applied towards
Factory-trained
newer technologies and materials.
engineers with years when replacement parts are
In fact, a large percentage of
of LTC experience needed. Our procurement
the units ABB services involve
supervise every specialists have the ability to
modi cations to incorporate
field project ABB immediately source and deliver
currently produced parts and/or
undertakes. Our goal OEM-grade parts to customers
upgrade units.
is always a quality anywhere in North America.
job, on time, on Retr ts and upgrades often involve:
ABB provides:
budget, with total Replacing obsolete LTCs with
Oil seats
customer satisfaction. current-production LTCs
Insulating panel boards
Assemblies (contact assemblies, Addition of a LTC to a non-LTC
brake assemblies, cam switch transformer
assemblies, and mechanism Conversion from a reactance-type
assemblies) LTC to a resistance-type LTC (or
Control cabinets the reverse procedure)
Kits (vacuum interrupter kits, Cam switch assembly and brake
Field Service assembly upgrading
remote indication kits, voltage
ABB performs eld maintenance, regulating relay kits) Control modi cation
testing, troubleshooting, and failure Arc chutes Addition of paralleling equipment
analysis. We match our employees Rotary position switches and/or data acquisition
to the appropriate make and model equipment
Inspection door gaskets
of LTC, equip them with accurate
diagnostic tools, and ensure that
they follow stringent procedures
Project Management ABB’s LTC service capability
to minimize outage time and A project engineer manages includes all the necessary electrical,
maximize equipment integrity every LTC service project that mechanical, oil testing, and
and job site safety. ABB undertakes. The engineer processing equipment in a self-
assures that all engineering, contained trailer to perform onsite
Typical equipment includes: maintenance repairs or retrofits.
drafting, purchasing, eld
Timing xtures for many
resource requirements, terms of
LTC models
service contracts, and warranty
Testers (hi potential testers,
considerations always meet or
turns ratio testers, and insulation
exceed ABB’s corporate standards
power factor testers)
and customer expectations.
Filter presses
Vacuum pumps Each ABB service project includes:
Boroscope equipment Schedule coordination
Meggers 24-hour-per-day project support
Oil test sets Project review and
Custom tools documentation of changes
High-current test sets

447
ABB provides OEM-grade parts, maintenance, and modernization services for the following Load Tap Changers (LTCs).

GENERAL FEDERAL ALLIS McGRAW

ABB WESTINGHOUSE REINHAUSEN MOLONEY
ELECTRIC PACIFIC CHALMERS EDISON
UZE UVT UTS LR9 LR68 RMS-1 MA-1 & 2 TC-15 TLB 220
UZF URT URS LR10 LR69 RMT-1 MB TC-25 TLH 397
UCG UTH UNR LR15 LR72 RMV-1 (-II) MB-1 TC-515 TLS 996
UCC UTN UT LR17 LR79 TYPE M MB-2 TC-525 500
UCD UTR UR LR19 LR81 TYPE T MC TC-546 V2A
UBB UB UC LR21 LR83 TYPE G MH
UCL URF URH LR27 LR85 TYPE R
UZD LR URV LR29 LR89 TYPE F
VR SDR LR31 LR91 TYPE MS
PDR URT-HC LR38 LR92 M Delta
URT-DTS UT-ATS LR40 LR95 V
URT-ATS LR41 LR96 VV
UTT (-A,-A70, -B) LR45R LR300 MA4
URT-46, -69 LR47 LR400 MA7/MA8
UVW (UVW-A) LR48 LR500 MA9
URL-4 (-8,-16) LR59 LR700
LR67

448
Power Transformer
On-Site Coil Reblocking Service
Certain large power transformers are prone to coil
degradation due to various design, operating and
environmental factors. Particularly, large power
substation and generator step-up transformers
that were built by General Electric in Pittsfield,
Mass. prior to 1972. Because of aging, loading and
the use of low density pressboard in their original
construction, it’s probable that their coil blocking
is loose. Also, these units have experienced oil box
problems, with resultant cooling deficiencies. As a
result:

Coil clamping pressure could be less than

needed to withstand a through fault
GE power transformers are highly susceptible to
Conductor and turn insulation could be coil movement. Regardless of the manufacturer, if
abraded through excessive movement of the it is feasible from a design standpoint, ABB’s
windings field reblocking services can restore transformer
The oil box could be ruptured, starving the integrity.
windings of necessary coolant. This creates
hot spots in the windings and deteriorates Identify High Risk Transformers
insulation even further ABB evaluates a number of indicators to
The transformers could fall catastrophically, determine if reblocking is needed, including:
which leads to the following situations: safety Design and construction. Essentially all
and ecological hazards of large oil spills, the GE Mark II designs (1965-1971), Mark I designs
high costs of cleanup, and the even higher (1959-1964), and pre-Mark I designs (pre 1959)
costs, both financial and social, of interrupted are at risk. These units have shown a tendency for
service shrinkage of the low density pressboard used in
their construction, resulting in loose coil blocking.
ABB Power Transformer Technical Services Operating history. The number of faults
engineers have been called on by utilities for experienced and their severity, as well as the
OEM support on power transformers built by loading history, are factors that can contribute to
ABB, Westinghouse and General Electric. ABB the loosening of the transformer coils.
has a staff of former GE transformer designers Gas-in-oil analysis, degree of
and specialists, along with all the drawings and polymerization. A high CO2/CO ratio in power
specifications, for all GE power transformers over transformers could indicate aging of the
40 MVA. insulation. A more definitive test which can
determine degradation of insulation is the degree
of polymerization test. ABB may recommend
these tests and is capable of providing the analysis.

449
 Internal inspection. The only certain way to Define the Scope of Work
evaluate tightness of blocking and loosening of the By evaluating the general condition of the
coils is to drain the oil from the transformer then transformer, including bushings, tap changers
visually and physically inspect the Internals of the and their drives, coolers, tank shielding, leads
transformer. Some of the items inspected are: and lead supports.

Tightness of window wedges and wedge blocks Determine Where the Work Can be
Bearing of spacers on end rings Completed
Any crushing of end rings at Either in the field or in an ABB repair
the bearing surface between facility. Cornpleting work in the field is
them and spacers preferable if there is sufficient room to position
Clearance to the winding jacks, replace channel fill if necessary, and in
cylinders general maneuver inside the unit.
Alignment and tightness of
radial spacers and
associated channel till
Oil box condition
Areas that are inaccessible to
visual inspection by the
human eye, are viewed through
a portable
for a better evaluation boroscope

ABB engineers have the advantage

of referring to original design
information arid drawings to
determine the correct clamping
force, the proper reblocking and fill
materials and the need for any
additional components that may
be necessary to ensure the future
reliability of the transformer
Any additional components that
may be needed can be supplied
by ABB.

A typical General Electric Mark II power transformer with flanged ends and
layers tapered. These transformers are very likely to experience coil movement
problems which can be corrected by the ABB coil reblocking process.

450
In rare cases where on-site field reblocking is not Implement and Guarantee the Work
possible, the unit can be shipped to either the ABB will review design drawings for the specific
ABB transformer repair facility in St. Louis transformer. We will supply materials that meet
present day design requirements. All material will
Estimate the Costs and Benefits be properly vapotherm processed and coordinated
Utilities cannot afford unreliable transmission to arrive at the site immediately prior to the start of
systems. Power transformers must be capable of the workscope.
surviving in a real world environment where ABB will support the job by supplying
through faults can occur at any moment. supervision, labor, tools and material for
A unit with very little or no compression force reblocking only, or can also provide complete
on the windings has virtually no short circuit turnkey operation including all oil handling and
integrity. The result of this could be a failed testing as well. We will guarantee all materials and
transformer, which would need to be rebuilt or workmanship in accordance with our standard
replaced by a new unit. The new equipment costs terms and conditions.
can be minor compared to the unexpected service Each ABB service project, regardless of size or
interruption. workscope, includes complete:

• Schedule coordination with the customer

including crews and outages
• Complete schedule coordination between
service crews, equipment and parts delivery
• 24-hour-per-day project support
• Complete project review arid documentation
which includes changes to the original shop
order information, this ensures that any
changes or modifications made are documented
for future reference

Every field reblocking project ABB undertakes is

supervised by factory trained engineers that
average 30 years of experience. Our goal is always
a quality job, delivered on time and budget, with
total customer satisfaction.

Before setting the jacks in place and

applying the final clamping force to
the coils, all necessary spacing over
the end rings is performed.

451
Railcar Rental Program for
Power Transformer Relocation
Moving large power transformers can be a Railcar Capacities and Daily
challenge. Their large dimensions and Rates/Fees
extreme weight pose unique requirements to
ensure safe and efficient transportation. The ABB Railcar Rental Program
ABB’s Railcar Rental Program helps includes the type and class of railcars
customers meet those unique requirements listed below.
with a fleet of specialty rail cars.
Car # Type Load Daily Fees
(lbs.) Rate
Railcar Types and Uses PTDX7060* Dep Center 512,000 $2,000 $5,000
PTDX7061* Dep Center 620,000 $3,000 $10,000
ABB owns, operates and rents a fleet of
PTDX-201 Schnabel 750,000 $3,500 $2,000
railcars designed specifically for the PTDX-202** Schnabel 1,000,000 $4,500 $2,500
transportation of large, heavy loads. Three PTDX-203 Schnabel 750,000 $3,500 $2,000
different types of cars are utilized based on PTDX-204 Schnabel 750,000 $3,500 $2,000
the specific load requirements: depressed PTDX-205* Side Rail 733,000 $2,000 $2,000

center, Schnabel and side rail. ABB’s rail * One time loading fee – no daily charge
specialists are available to assist you in ** A special train is required to move this car under load.
determining the best choice. Special train rates are charged on a per-mile basis
In addition to normal freight charges.

Railcar Rental Arrangements and

Terms
ABB will help you make the choice in
all areas of railcar selection, scheduling,
insurance, and contracting. Rental terms
are negotiated individually for each
contract however; typical terms include,
but are not limited to, the following:
Depressed Center. The 12-axle depressed center •Daily rental rates are applicable from
flatcars are used primarily with larger load
requirements. These cars are equipped with a the day the car departs until the day it
cushioning system to provide a smooth ride along the returns.
rails.
•All freight charges are the customer’s
responsibility.
•Any damage to the railcar, whether
caused by customer or railroad, is the
responsibility of the customer.
•A rental contract must be executed
prior to the release of any railcar to
Schnabel. Certain transformers are designed to be an the customer.
integral part of the “Schnabel-type” car. The
transformer attaches to railcar frames utilizing a •ABB technical assistance is required
pinning system located near the base. This allows the on-site during the loading and
transformer to ride approximately six inches above the
rail. These cars range in mechanical design from 12 to unloading of Schnabel cars utilizing
20 axles. Load requirements determine car capacity ABB price list 48-519.
requirements.

453
 Contact list for main ABB Service Centers
Below is a contact list with the main ABB Service centers. Also updated contacts for more than hundred
countries are available on:
www.abb.com/transformers
___________________________________________________________________

Head Quarters
ABB Management Services
Business Unit Transformers
Product Group Service
P.O. Box 8131
Affolternstrasse 44
8050 Zürich
Switzerland
www.abb.com/transformers

North Asia
China
ABB Zhongshan Transformer Co., Ltd
No.1 Haicheng North Road
Hengmen, Nanlang Twon
Zhongshan, Guangdong
Post code: 528449
P. R. China
Tel: +86 760 3392288-3802
Fax: +86 760 3392898

________________________________________________________________

Europe

Denmark
ABB A/S
Meterbuen 33
DK- 2740 Skovlunde
Denmark
Tel: +45 4450 4450
Fax:+45 4450 4700

Finland
ABB Oy, Transformers
Strömbergin Puistotie 15D
Vaasa
Finland
Tel: +358 50 3342235
Fax: +358 10 22 41021

454
Germany
ABB AG
GBU Transformatoren
Delitzscher Straße 74
06112 Halle
Germany
Tel: +49 345 5686 247
Fax: +49 345 5686 120

Italy
ABB Power Technologies SpA
Unità Operativa Trasformatori
Viale dell’Industria, 18
20010 Vittuone (MI)
Italy
Tel: +39 02 9034 71 60
Fax: +39 02 9034 7493
Norway
ABB AS
Power Products Division
P.O.Box 470 Brakeroya
N-3002 Drammen
Norway
Tel: +47 2416 5547
Fax: +47 3224 7934
Email : [email protected]

Poland
ABB Sp. z o.o.
67/93 Aleksandrowska str
91-205 Lodz
Poland
Tel: + 48 601079417
Fax: +48 42 6526096

Spain
Asea Brown Boveri S.A.
San Romualdo 13
28037 Madrid
Spain
Phone: +34 91 5819393
Fax: +34 91 5810733

Sweden
ABB AB
Power Transformers
771 80 Ludvika
Sweden
Tel: +46 240 784322
Fax: +46 240 13091

455
Switzerland
ABB Sécheron SA
Rue des Sablières 4-6. P.O. Box 2095
1211 Geneva 2
Switzerland
Tel: +41 58 586 22 11
Fax: +41 58 586 23 08

Turkey
ABB Elektrik San A.S.
Esentepe Mah. Milangaz Cd. No.52
Kartal, Istanbul
Turkey
Tel: +90 216 528 2526
Fax: +90 216 387 1890

United Kingdom
ABB Ltd - Transformer Service
Oulton Road
Stone
Staffordshire, ST15 0RS
United Kingdom
Tel: +44 1785 825 050
Fax: +44 1785 819 019

Middle East and Africa

Egypt
7, Dr.Mohamed Kamel Hussein St.,
El Nozha El Gadida-Heliopolis ,
Cairo, Egypt
Tel: +20 (2) 26251555
Fax: +20 (2)26222620

Jordan
ABB Ltd- Jordan
P.O Box 926098
Amman 11190
Jordan
Tel: +962 6 416 3122
Fax: +962 6 416 4906

Saudi Arabia – Middle East

ABB Service Company Limited
P.O. Box 2873 Al-Khobar 31952
Kingdom of Saudi Arabia
Tel: +966 3 882 9394
Fax: +966 3 882 4603

456
South Africa
ABB Campus
2 Lake Road, Longmeadow Business Estate (North)
1609, Modderfontein, Gauteng,
South Africa
Tel: +27 10 202 5356
Fax: +27 11 579 8571
________________________________________________________
North America

Canada
TRES Canada
201 Westcreek Blvd,
Brampton, Ontario
Canada, L6T 5S6
Tel: +1 905 460 3250
Fax: +1 905 460 3018

Mexico
TRES Mexico
Blvd. Centro Industrial No.12
Los Reyes Zona Industrial
54073 Tlalnepantla, Edo. De México
Apartado Postal M-2434
06000 México D.F.
Mexico
Tel: +52 (55) 5328 1400 Ext. 3435
Fax: +52 (55) 55 65 65 28

United States of America

TRES US
4350 Semple Avenue
St. Louis, MO 63120-2241
USA
Tel: +1 877 839 7877
Fax: +1 314 679 4595

South Asia

Australia
ABB Australia Pty. Limited
Power Technologies Service – Transformer service
Locked Bag 7315, Liverpool BC NSW 1751
Australia
Tel: +61 2 9821 0228
Fax: + 61 2 98210919

New Zealand
ABB Limited
Private Bag 92609 Symonds Street
Auckland
New Zealand
Tel: +64 3 3383787
Fax: +64 3 3380110

457
India
ABB Ltd, Transformers
Vadodara-390013
India
Tel: +91 80 8395181
Fax: +91 265 638 921

Thailand
ABB Ltd
Transformer Repair & Service
322 Moo 4 Bangpoo Industrial Estate Soi 6
Sukhumvit Rd. Prakesa, Muang, Samutprakarn 10280
Bangkok
Thailand
Tel: +66 2 762 2033
Fax: +66 2 709-3368

___________________________________________________________________

South America

Brazil
ABB Ltda.
Av. Monteiro Lobato
3411 – Guarulhos – SP
Brazil
Tel: +55 11 6464 8690
Fax: +55 11 6464-8399

Perú
ABB S.A
Avenida Argentina 3120
Lima
Perú
Tel : +51 1 415 5100
Fax: +51 1 561 3040

458