merger of ieee 56 & 432 draft - iranelectrical...

P56 Rev 16 Nashville .doc of 63

Working Group Copy IEEE P56 Rev. 16 Draft

IEEE P56 Draft Guide for Insulation Maintenance of

Electric Machines

Prepared by the P56 Working Group of the Materials Subcommittee of the Electric

Machinery Committee

This is an unapproved draft of a proposed IEEE Standard, subject to change. Permission is hereby granted for IEEE Standards Committee participants to reproduce this document for purposes of IEEE standardization activities. Permission is also granted for member bodies and technical committees of ISO and IEC to reproduce this document for purposes of developing a national position. Other entities seeking permission to reproduce this document for standardization or other activities, or to reproduce portions of this document for these or other uses, must contact the IEEE Standards Department for the appropriate license. Use of information contained in this unapproved draft is at your own risk.

IEEE Standards Department Copyright and Permissions 445 Hoes Lane, P.O. Box 1331 Piscataway, NJ 0885-1331, USA

Revision Printed: 10/18/2007 7:07:00 PM


Working Group Copy IEEE P56 Rev. 16 Draft INTRODUCTION

(This introduction is not part of P56, Guide for Insulation Maintenance for Electric Machines 5 hp and Higher).

This revision represents the merger of:

IEEE Std 432-1976, Guide for Insulation Maintenance for Rotating Electrical Machinery (5 HP to less than 10 000 HP) and

IEEE Std. 56-1977, IEEE Guide for Insulation Maintenance of Large Alternating-Current Rotating Machinery 10 000 kVA and Larger

Participants

This document was originally developed by a working group of the Insulation Subcommittee of the IEEE Rotating Machinery Committee. The members of this working group were:

[Add members]

The 1977 revision was prepared by a working group of the Insulation Subcommittee. The members were:

[Add members]

This revision was prepared by a working group of the Materials Subcommittee of the Electric Machinery Committee. The members were:

Douglas Conley, Chairman

Meredith Stranges, Secretary

[Add members]



Working Group Copy IEEE P56 Rev. 16 Draft CONTENTS

CLAUSE Page 1. INTRODUCTION..............................................................................................................................................9

1.1 OVERVIEW ..................................................................................................................................................9 1.1.1 Scope......................................................................................................................................................9 1.1.2 Purpose..................................................................................................................................................9

2. REFERENCES...................................................................................................................................................9

2.1 IEEE STANDARDS .......................................................................................................................................9

NOTE: THE STANDARDS DOCUMENTS WILL BE CORRECTLY REFERENCED AT THE TIME OF SUBMISSION FOR BALLOT. THIS WILL BE CHECKED AGAINST THE SA WEB SITE AT THAT TIME..........................................................................................9

2.2 ASTM1 STANDARDS .................................................................................................................................10

2.3 IEC STANDARDS .......................................................................................................................................11 3. DEFINITIONS .................................................................................................................................................11 4. SAFETY [L. RUX – CLAUSE HEAD]............................................................................................................12

4.1 GENERAL...................................................................................................................................................12

4.2 MACHINE ROTATION.................................................................................................................................12

4.3 SOLVENTS .................................................................................................................................................12

4.4 ASBESTOS, LEAD, AND OTHER HAZARDOUS MATERIALS..........................................................................13 5. THE SIGNIFICANCE OF MAINTENANCE...............................................................................................13 6. INSULATION SYSTEMS IN GENERAL USE [BILL MCDERMID – CLAUSE HEAD] ........................13

6.1 INSULATING MATERIALS...........................................................................................................................14

6.2 ARMATURE WINDING INSULATION............................................................................................................14 6.2.1 Strand Insulation .................................................................................................................................15 6.2.2 Turn Insulation ....................................................................................................................................15 6.2.3 Groundwall Insulation.........................................................................................................................15 6.2.4 Semi-conductive Slot Coating..............................................................................................................17 6.2.5 Stress Control Coating ........................................................................................................................17 6.2.6 Commutator Insulation........................................................................................................................18 6.2.7 Support Insulation ...............................................................................................................................18

6.3 WOUND ROTOR WINDINGS (3 PHASE INDUCTION MACHINES) ..................................................................18 6.3.1 Partially closed slots - Strap Windings (Generally used on high-speed machines) ............................18 6.3.2 Open slots ............................................................................................................................................18 6.3.3 Strand (wire) Insulation.......................................................................................................................19 6.3.4 Ground wall Insulation........................................................................................................................19 6.3.5 Collector Rings ....................................................................................................................................19

6.4 FIELD WINDING INSULATION ....................................................................................................................19 6.4.1 Field Windings.....................................................................................................................................19



Working Group Copy IEEE P56 Rev. 16 Draft 6.4.2 Turn (Conductor) Insulation................................................................................................................19 6.4.3 Ground Insulation................................................................................................................................19 6.4.4 Collector Insulation .............................................................................................................................20 6.4.5 Brush-Rigging Insulation ....................................................................................................................20

6.5 CORE AND FRAME-ASSEMBLY INSULATION ..............................................................................................20 6.5.1 Stator Core Interlaminar Insulation (Core Plate Insulation)..............................................................20 6.5.2 Insulation Punchings ...........................................................................................................................20 6.5.3 Core Tightening Through-Bolt Insulation ...........................................................................................20

6.6 OTHER INSULATING PARTS .......................................................................................................................20 7. SERVICE CONDITIONS AFFECTING INSULATION LIFE [IAN CULBERT – CLAUSE HEAD] ..21

7.1 AGING MECHANISMS.................................................................................................................................21

7.2 AC STATOR WINDING AGING MECHANISMS.............................................................................................22 7.2.1 Thermal Deterioration.........................................................................................................................22 7.2.2 Thermal Cycling ..................................................................................................................................22 7.2.3 Poor Impregnation ..............................................................................................................................23 7.2.4 Internal Water Leaks ...........................................................................................................................23 7.2.5 Loose Coils, or Bars in the Slot ...........................................................................................................24 7.2.6 Semi-conductive Coating Degradation................................................................................................24 7.2.7 Electrical / Mechanical (Contact) Erosion..........................................................................................24 7.2.8 Semi-conductive/Grading Coating Interface Failure ..........................................................................25 7.2.9 Electrical Stresses................................................................................................................................25 7.2.10 Electrical Tracking due to Contamination ..........................................................................................26 7.2.11 Voltage Surges.....................................................................................................................................26 7.2.12 Environmental Factors ........................................................................................................................27 7.2.13 Endwinding Vibration..........................................................................................................................30

7.3 CYLINDRICAL (ROUND) ROTOR WINDING AGING MECHANISMS...............................................................31 7.3.1 Thermal Aging .....................................................................................................................................31 7.3.2 Thermal Cycling ..................................................................................................................................32 7.3.3 Abrasion Due to Imbalance or Turning Gear Operation ....................................................................32 7.3.4 Electrical Tracking form Contamination.............................................................................................33 7.3.5 Repetitive Voltage Surges ....................................................................................................................33 7.3.6 Rotational Force..................................................................................................................................34

7.4 SALIENT POLE ROTOR WINDING AGING MECHANISMS .............................................................................34 7.4.1 Thermal Aging .....................................................................................................................................34 7.4.2 Thermal Cycling ..................................................................................................................................35 7.4.3 Pollution (Tracking and Moisture Absorption) ...................................................................................36 7.4.4 Abrasive Particles................................................................................................................................36 7.4.5 Rotational Force..................................................................................................................................36 7.4.6 Repetitive Voltage Surges ....................................................................................................................37

7.5 WOUND ROTOR WINDING AGING MECHANISMS.......................................................................................37 7.5.1 Thermal Aging .....................................................................................................................................37 The thermal aging and its effects in this type of winding is similar to that discussed under Section 7.3.1 for round rotor windings. This is so since the materials used are similar and both are subjected to significant mechanical stresses resulting from Rotational forces. Therefore no discussion on this topic is included under wound rotor windings. .......................................................................................................................................37 7.5.2 Transient Overvoltages........................................................................................................................37 7.5.3 Unbalanced Stator Voltages ................................................................................................................38



Working Group Copy IEEE P56 Rev. 16 Draft 7.5.4 High-Resistance Connections ..............................................................................................................38 7.5.5 End-Winding Banding Failures...........................................................................................................38 7.5.6 Slip Ring Insulation Shorting and Grounding .....................................................................................38 7.5.7 Pollution (Tracking and Moisture Absorption) ...................................................................................39

7.6 DC MOTOR AND GENERATOR FIELD WINDINGS .......................................................................................39 7.6.1 Thermal Aging .....................................................................................................................................39 7.6.2 Thermal Cycling ..................................................................................................................................40 7.6.3 Abrasive Particles................................................................................................................................40 7.6.4 Pollution (Tracking and Moisture Absorption) ...................................................................................40

7.7 DC MOTOR AND GENERATOR ARMATURE WINDINGS AND COMMUTATORS .............................................41 7.7.1 Thermal Aging .....................................................................................................................................41 7.7.2 High Resistance Connections ..............................................................................................................41 7.7.3 End-Winding Banding Failures...........................................................................................................41 7.7.4 Pollution (Tracking and Moisture Absorption) ...................................................................................41

7.8 DC MOTOR AND GENERATOR COMMUTATORS .........................................................................................41 7.8.1 Glass Band Contamination..................................................................................................................41 7.8.2 Electrical Tracking ..............................................................................................................................42 7.8.3 Commutator Wear ...............................................................................................................................42 7.8.4 Commutator Eccentricity.....................................................................................................................42 7.8.5 Commutator Brush Wear.....................................................................................................................42

7.9 STATOR CORE INSULATION AGING MECHANISMS .....................................................................................43 7.9.1 Thermal Aging .....................................................................................................................................43 7.9.2 Electrical Aging ...................................................................................................................................44 7.9.3 Mechanical Aging................................................................................................................................46 References 49

8. VISUAL INSPECTION METHODS [NANCY FROST – CLAUSE HEAD]...............................................49

8.1 VISUAL INSPECTION SAFETY .....................................................................................................................50

8.2 ARMATURE WINDING................................................................................................................................50 8.2.1 Thermal Aging .....................................................................................................................................50 8.2.2 Cracking ..............................................................................................................................................50 8.2.3 Girth Cracking.....................................................................................................................................50 8.2.4 Contamination .....................................................................................................................................51 8.2.5 Carbon Deposits ..................................................................................................................................51 8.2.6 Abrasion...............................................................................................................................................51 8.2.7 Loose Slot Wedges or Slot Fillers........................................................................................................51 8.2.8 Erosion.................................................................................................................................................51 8.2.9 Corrosion / Chemical Attack ...............................................................................................................51 8.2.10 Corona.................................................................................................................................................51 8.2.11 Rotational Forces ................................................................................................................................51 8.2.12 Commutator Condition ........................................................................................................................52

8.3 FIELD WINDINGS.......................................................................................................................................52 8.3.1 Coil Distortion .....................................................................................................................................52 8.3.2 Loose Coils or Poles............................................................................................................................52 8.3.3 Rotor Coil Tightness ............................................................................................................................52 8.3.4 Brush Rigging ......................................................................................................................................53

8.4 CORE AND FRAME ASSEMBLY...................................................................................................................53



Working Group Copy IEEE P56 Rev. 16 Draft 8.4.1 Stator (Armature) Core........................................................................................................................53 8.4.2 Core Insulated Through-Bolts .............................................................................................................53 8.4.3 Bearing, Hydrogen-Seal, and Other Insulation...................................................................................54

9. INSULATION MAINTENANCE TESTING [RICHARD – CLAUSE HEAD] ..........................................54

9.1 PRINCIPLES OF MAINTENANCE TESTING....................................................................................................54

9.2 TESTS CONDUCTED ON THE FIELD (ROTOR) .............................................................................................54 9.2.1 Insulation Resistance. (numbering?) ..................................................................................................55 9.2.2 Winding Resistance..............................................................................................................................55 9.2.3 Field Winding Voltage Drop Test........................................................................................................55 9.2.4 Impedance Test ....................................................................................................................................55 9.2.5 Flux Distribution Tests ........................................................................................................................56 9.2.6 . 56

9.3 TESTS CONDUCTED ON THE ARMATURE (STATOR)...................................................................................57 9.3.1 Insulation Resistance Test at Low Voltage. .........................................................................................57 9.3.2 Dielectric Absorption Test. ..................................................................................................................57 9.3.3 Over Voltage Tests...............................................................................................................................58 9.3.4 Controlled Over Voltage Test (DC).....................................................................................................58 9.3.5 Alternative Method of Controlled Over Voltage Test. .........................................................................59 9.3.6 Other Over Voltage Methods...............................................................................................................59 9.3.7 Insulation Power-Factor Test or Dissipation Factor Test ..................................................................59 9.3.8 Insulation Power-Factor Tip-up Test or Dissipation Factor Tip-up Test. ..........................................60 9.3.9 Slot Discharge and Corona Probe Tests. ............................................................................................60 9.3.10 Corona-Probe Test. .............................................................................................................................60 9.3.11 Partial-Discharge Tests.......................................................................................................................61 9.3.12 Slot-Discharge Test. ............................................................................................................................64 9.3.13 Turn-To-Turn Insulation Test. .............................................................................................................65 9.3.14 Insulation-Resistance Test of Embedded Temperature Detectors. ......................................................65 9.3.15 Coil-to-Core Contact Resistance. ........................................................................................................65 9.3.16 Stator Core Interlaminar Resistance Insulation Test ..........................................................................66 9.3.17 Slot Discharge and Corona Probe Tests. ............................................................................................66 9.3.18 Corona-Probe Test. .............................................................................................................................66 9.3.19 Partial-Discharge Tests.......................................................................................................................67 9.3.20 Slot-Discharge Test. ............................................................................................................................68 9.3.21 Resistance Temperature Detectors (RTDs) .........................................................................................68 9.3.22 Insulation-Resistance Test of Embedded Temperature Detectors. ......................................................69 9.3.23 Insulation Resistance Test of Insulated Stator-Through-Bolts. ...........................................................70 9.3.24 Coil-to-Core Contact Resistance. ........................................................................................................70 9.3.25 Winding Resistance..............................................................................................................................70 9.3.26 Test of Interlaminar Insulation of Stator Core. ...................................................................................70 9.3.26.1 Safety Considerations. .........................................................................................................................71 9.3.26.2 Test Procedure.....................................................................................................................................71 9.3.27 Stator Core Interlaminar Resistance Insulation Test ..........................................................................71

10. CLEANING [LORI RUX – CLAUSE HEAD] ...........................................................................................72

10.1 GENERAL...................................................................................................................................................72

10.2 CLEANING TECHNIQUES ............................................................................................................................73 10.2.1 Vacuum Cleaning ................................................................................................................................73 10.2.2 Air-Lance Cleaning .............................................................................................................................73 10.2.3 Solvent Cleaning..................................................................................................................................73



Working Group Copy IEEE P56 Rev. 16 Draft 10.2.4 ABRASIVE BLASTING............................................................................................................................74

10.2.5 CO2 BLASTING (CRYOGENESIS)............................................................................................................75

10.2.6 STEAM CLEANING ................................................................................................................................75

10.2.7 CLEANING BY WATER IMMERSION OR WATER HOSE ...........................................................................76 10.2.4 Cleaning Instructions...........................................................................................................................76 10.2.5 Field Service Cleaning of Assembled Machines..................................................................................76 10.2.6 Service Shop Cleaning of Disassembled Machines. ............................................................................77

10.3 GENERAL...................................................................................................................................................78

10.4 CLEANING TECHNIQUES ............................................................................................................................78

10.5 VACUUM CLEANING..................................................................................................................................78

10.6 AIR-LANCE CLEANING. .............................................................................................................................79

10.7 SOLVENT CLEANING..................................................................................................................................79

10.8 TYPES OF SOLVENTS .................................................................................................................................79

10.9 RISK OF DAMAGE ......................................................................................................................................79

10.10 ABRASIVE BLASTING.................................................................................................................................80

10.11 CLEANING WITH SOLID CO2 ......................................................................................................................80

10.12 STEAM CLEANING .....................................................................................................................................80

10.13 CLEANING BY WATER IMMERSION OR WATER HOSE................................................................................81 11. BIBLIOGRAPHY [DOUG CONLEY– CLAUSE HEAD] ........................................................................81 ANNEX A (INFORMATIVE) (FORMALLY FROM 56) ....................................................................................87

A.1. DESIGN OF MAGNETIZING COIL.................................................................................................................87

A.2. SEARCH COIL ............................................................................................................................................87

A.3. CALCULATIONS .........................................................................................................................................87

A.4. TEMPERATURE MEASUREMENTS ...............................................................................................................90 ANNEX B (NORMATIVE) ......................................................................................................................................92 [FORMALLY FROM IEEE 432…??] ......................................................................................................................92

B.1. DISCUSSION OF TESTS DESCRIBED IN THE GUIDE ......................................................................................92 B.1.2 Dielectric Absorption Test ...................................................................................................................92 B.1.1.3 Slot Discharge and Corona Probe Tests .............................................................................................93 B.1.4 Stator Core Interlaminar Insulation Test (Loop Test).........................................................................94 Induction Requirements .....................................................................................................................................95 Induction 95 Kilolines per Square Inch ..................................................................................................................................95



Working Group Copy IEEE P56 Rev. 16 Draft Tesla 95 Ampere Turns, per Inch of Core Mean Periphery .............................................................................................95 Oersteds, per Centimeter of Core Mean Periphery ...........................................................................................95

ANNEX C (INFORMATIVE) ..................................................................................................................................97 Rotating Electric Machinery Insulation Condition Visual Inspection Appraisal – Visual Inspection Checklist 100

NEW ANNEX - THERMOSETTING RESINS USED IN INSULATION SYSTEMS.....................................103 NEW ANNEX - STATOR CORE LOW ENERGY TEST ..................................................................................104 FIGURES

Figure 1. RTD (3 wire) and Wheatstone Bridge Circuit............................................................... 69



Working Group Copy IEEE P56 Rev. 16 Draft 1. Introduction

1.1 Overview

1.1.1 Scope

This insulation maintenance guide is applicable to rotating electric machines rated from 35 KVA and higher. The procedures detailed herein may also be useful for insulation maintenance of other types of machines.

1.1.2 Purpose

The purpose of this guide is to present information necessary to permit an effective evaluation of the insulation systems of rotating electrical machines. Such an evaluation can serve as a guide to the degree of maintenance or replacement as might be deemed necessary, and also offer some indication of the future service reliability of the equipment under consideration.

2. References

This guide should be used in conjunction with the following publications. When the standards are superseded by an approved revision, the revision should apply.

2.1 IEEE Standards

Note: The standards documents will be correctly referenced at the time of submission for ballot. This will be checked against the SA web site at that time.

IEEE Std 4-1995, IEEE Standard Techniques for High Voltage Testing (ANSI C68.1-1968).i

IEEE Std 43-2000, IEEE Recommended Practice for Testing Insulation Resistance of Rotating Machinery (ANSI).

IEEE Std 51-1955, Guiding Principles for Dielectric Tests. [Withdrawn]

IEEE Std 62.2- 2004, IEEE Guide for Field Testing of Electric Power Apparatus Insulation – Electrical Machinery.

IEEE Std 67-1990, IEEE Guide for Operation and Maintenance of Turbine Generators (ANSI).

IEEE Std 95-2002, IEEE Recommended Practice for Insulation Testing of Large AC Rotating Machinery with High Direct Voltage (ANSI).

iIEEE publications are available from the Institute of Electrical and Electronics Engineers, Service Center, 445 Hoes Lane, P. O. Box 1331, Piscataway, NJ 08855-1331, USA.



Working Group Copy IEEE P56 Rev. 16 Draft IEEE Std 100-1996, IEEE Standard Dictionary of Electrical and Electronics Terms 4th ed. (ANSI).

IEEE Std 112 - 1996, IEEE Standard Test Procedure for Polyphase Induction Motors and Generators

IEEE Std 115-1965, Test Procedure for Synchronous Machines. (under revision)

IEEE Std 118-1949, Master Test Code for Resistance Measurement.

IEEE Std 119-1974, Recommended Practice for General Principles of Temperature Measurement as Applied to Electrical Apparatus. [Withdrawn]

IEEE Std 286-2000 (Reaff. 2006), Recommended Practice for Measurement of Power-Factor Tip-Up of Rotating Machinery Stator Coil Insulation.

IEEE Std 433-1974 (Reaff. 1991), IEEE Recommended Practice for Insulation Testing of Large AC Rotating Machinery with High Voltage at Very Low Frequency (ANSI)

IEEE Std 454-1973, Recommended Practice for the Detection and Measurement of Partial Discharges (Corona) During Dielectric Tests.

IEEE Std 492-1999, IEEE Guide for Operation and Maintenance of Hydro-Generators (ANSI).

IEEE Std 510 – 1983 “Recommended Practice for Safety in High Voltage and High Power Testing. [Withdrawn]

IEEE Std 522-1992, (Reaff 1998) IEEE Guide for Testing Turn-to-Turn Insulation on Form-Wound Stator Coils for Alternating-Current Rotating Electric Machines.

IEEE Std 1434 – 2000 Recommended Practice for Partial Discharge Measurements of Electrical Machinery.

2.2 ASTM1 Standards

Std. D3382-95(2001)e1 Standard Test Methods for Measurement of Energy and Integrated Charge Transfer Due to Partial Discharges (Corona) Using Bridge Techniques

Std. D1868-93(1998) Standard Test Method for Detection and Measurement of Partial Discharge (Corona) Pulses in Evaluation of Insulation Systems

Std. D2275-01 Standard Test Method for Voltage Endurance of Solid Electrical Insulating Materials Subjected to Partial Discharges (Corona) on the Surface

Std. F855-04 Standard Specifications for Temporary Protective Grounds to Be Used on De-energized Electric Power Lines and Equipment



Working Group Copy IEEE P56 Rev. 16 Draft 2.3 IEC Standards

Std. 60034-27(2006): Off-line partial discharge measurements on the stator winding insulation of rotating electrical machines

3. Definitions

Discharge Detector (ionization or corona detector) – An instrument that can be connected in or across an energized insulation circuit to detect current or voltage pulses produced by electric discharges within the circuit.

Power Factor (PF) – The cosine of the dielectric phase angle or the sine of the dielectric loss angle when tested under a sinusoidal voltage. The ratio of the power dissipated in the insulation, in watts, to the product of the effective voltage and current in voltamperes, when tested under a sinusoidal voltage and prescribed conditions. The insulation power factor is equal to the cosine of the phase angle between the voltage and the resulting current when both the voltage and current are sinusoidal.

Resistance Temperature Detector (RTD) (resistance thermometer resistor) (resistance thermometer detector) – A resistor made of some material for which the electrical resistivity is a known function of the temperature and that is intended for use with a resistance thermometer. It is usually in such a form that it can be placed in the region where the temperature is to be determined.

Stator – The portion that includes and supports the stationary active parts. The stator includes the stationary portions of the magnetic circuit and the associated winding and leads. It may, depending on the design, include a frame or shell, winding supports, ventilation circuits, coolers, and temperature detectors. A base, if provided, is not ordinarily considered to be part of the stator.

Stator Bar – A unit of winding on the stator of a machine.

Stator Coil – A unit of a winding on the stator of a machine.

Stator Core – The stationary magnetic-circuit of an electric machine. It is commonly an assembly of laminations of magnetic steel, ready for winding.

Semiconductive Slot Coating – The partially conductive paint or tape layer in intimate contact with the groundwall insulation in the slot portion of the stator core. This coating ensures that there is little voltage between the surface of the coil or bar and the grounded stator core.

[Note Section Heads!!! Please add words as needed. ]



Working Group Copy IEEE P56 Rev. 16 Draft 4. Safety [L. Rux – Clause head]

WARNING - Testing and maintenance activities should be conducted and supervised by qualified personnel, and adequate safety precautions should be taken to avoid injury to personnel and damage to property.

It is not the intent of this document to prescribe safety and health requirements. Before starting any work, the relevant laws, regulatory standards, manufacturer instructions, and company policies should be consulted.

4.1 General

Prior to performing any test, inspection, or servicing of an electric machine where the unexpected energizing, start up, or release of kinetic or stored energy could occur and cause injury or damage, the apparatus shall be de-energized, isolated, blocked, and secured to control hazardous energy. Personnel and equipment shall not be considered protected until the appropriate safety procedures have been implemented. Safety procedures may involve the use of locking devices, warning tags, physical barriers, safety tape, caution signs, and/or observers as necessary to restrict access to the equipment being maintained. There should be a meeting of all personnel involved in or affected by the maintenance activities. Test and inspection procedures should be discussed so there is a clear understanding of all aspects of the work to be performed. Particular emphasis should be placed on personnel and equipment hazards, and the safety precautions associated with these hazards. In addition, details of the maintenance activities should be discussed to assure the successful completion of the planned tasks. Responsibilities for the various duties involved in performing the work should be assigned and documented. (Refer to IEEE Std. 510.)

4.2 Machine Rotation

Some test and inspection procedures are performed with the machine rotating slowly and with cover plates, guards, and end-shields removed. In the case of hydro-generators under test, the machine may be operated at rated speed or at runaway speed with its covers removed. These tests present mechanical and electrical hazards, and appropriate procedures are required to prevent injury to personnel and damage to equipment.

4.3 Solvents

Persons who carry out cleaning using solvents shall be instructed on the safe storage, use, and emergency actions related to the solvents used. Manufacturer’s recommendations, local procedures, and safety regulations should be followed to assure the proper use of personal protective equipment (PPE) and the correct handling of waste materials..



Working Group Copy IEEE P56 Rev. 16 Draft 4.4 Asbestos, Lead, and Other Hazardous Materials

Operators should be aware that older machines may contain asbestos and/or lead products that, if disturbed, could pose a threat to worker health and safety. In addition, electrical apparatus that has been subjected to extreme overheating or other unusual conditions may be contaminated with dangerous or toxic substances. Chemical sampling and testing may be warranted to determine the presence of hazardous materials. In some cases, clean up or abatement activities may be necessary and workers must take special precautions, including the use of PPE.

5. The Significance of Maintenance

The experience and data obtained from regular maintenance inspection and testing programs can, in addition to providing an evaluation of the present condition of the equipment, give some indication of long-term trends and probable need for future repair or replacement.

Rotating electric machines are complex structures that are subjected to mechanical, electrical, thermal, and environmental stresses of varying magnitude. Of the various components, the insulation systems are the most susceptible to aging or damage due to these stresses. The service life of an electric machine will, therefore, largely depend on the serviceability of the insulation systems.

The extent to which a maintenance program is pursued will depend largely on the operator's own experience and policy, but should also take into account the importance of service reliability for the equipment. Where high service reliability is required, a regular maintenance program involving periodic disassembly and knowledgeable visual examination of the equipment, together with the application of electrical tests of proven significance, is strongly recommended.

Where reliability is of concern, adequate inspection and testing programs are advocated to assure that the equipment is maintained in satisfactory condition to minimize the possibility of in-service failure. The size and age of the asset in question will also factor into the applicability of the individual tests and inspection methodologies employed.

It should be recognized that over-potential tests can damage insulation that is contaminated or in marginal condition. Where there is uncertainty of the condition of the insulation system, refer to clause 11 for additional information. Consultation with the manufacturer is recommended for setting up an appropriate maintenance-testing program.

6. Insulation Systems in General Use [Bill McDermid – Clause head]

Insulation is present in various machine components, but the complexity of the subject is such that only a general description can be given here.



Working Group Copy IEEE P56 Rev. 16 Draft 6.1 Insulating Materials

Some common insulating materials applied to electric machines have the following thermal classifications:

– Class 105 (formerly Class A): impregnated cotton, silk, cellulose based paper, linen (cambric)

– Class 130 (formerly Class B): mica, glass fiber, asbestos, etc. Typical bonding materials are shellac, asphalt varnish and some polyester resins.

– Class 155 (formerly Class F): mica, glass fiber, asbestos, etc. Bonding materials are usually epoxy resins.

– Class 180 (formerly Class H): silicone elastomer, mica, glass fiber, asbestos, etc. Bonding materials may consist of silicone resins.

Mica is a vital component in most insulation for high voltage electric machines because it has good dielectric strength at high temperatures and is resistant to partial discharges. Muscovite mica is the best choice for electrical applications because of its tensile strength and dielectric strength.

Originally mica was used in the form of large flakes, or splittings. When applied in this way it is difficult to exclude all voids, and some problems with delaminations may occur. More recently mica has been used in the form of mica paper where small flakelettes of mica are deposited on and bonded to a backing tape. With mica paper it is easier to obtain a void free structure. However mica paper is less resistant to partial discharges than is large flake mica.

Materials bonded with shellac or with asphalt varnish are termed “thermoplastic”. Materials bonded with polyester or epoxy resin are termed “thermosetting”. Further information on thermosetting resins will be found in the Annex.

When mica and glass fibers are bonded together with varnish or resin they form a composite insulation system. Such a system should have good thermal, electrical and mechanical properties.

Caution: Asbestos may be present in slot packing materials, armor tape and strand insulation of many older windings. Some slot packing materials and varnished cambric may contain polychlorinated biphenyls (PCBs). Appropriate workplace safety and environmental regulations should be followed when examining, disturbing or disposing of these materials.

6.2 Armature Winding Insulation

The armature windings of ac machines are stationary and are known as stator windings. The armature windings of dc machines are rotating. In addition, dc machines have commutator and banding insulation.

The armature winding with its associated leads is the main current carrying winding of the machine. The coils of the armature winding have strands, ground insulation and may have turn



Working Group Copy IEEE P56 Rev. 16 Draft insulation. Wedges, blocks, and other insulated mechanical supports are a part of the armature winding assembly.

The following insulation systems have been used for armature windings:

6.2.1 Strand Insulation

The individual strands of armature coil conductors are usually insulated. Strand insulation can be made up of organic resin enamels, polymeric films, resin bonded fibers (such as paper, cotton, asbestos, glass, polyester, or combinations thereof) or resin bonded mica.

In the 1920s cotton was commonly used as the insulation on individual copper strands. By the 1950s asbestos was in use as strand insulation because of its higher temperature classification. By the 1970s polyester glass fiber (Dacron® Glass) was commonly used as strand insulation. So as to maximize the copper some manufacturers alternate polyester glass fiber insulated strands with strands having an enamel coating.

6.2.2 Turn Insulation

In a coil with more than one turn, groups of strands forming a single-turn (conductor) may be held together and insulated. Individual strand insulation, as described above, may also serve as turn insulation.

Where dedicated turn insulation is provided for multi-turn coils it usually involves similar materials to those in the groundwall insulation in the slot section.

6.2.3 Groundwall Insulation

Groundwall insulation is the material intended to insulate the current-carrying components (i.e. the coils, the circuit rings, and connections thereto) from one another and from the non-current-carrying components, which are usually considered to be grounded (such as the core iron, the frame, and other structural members).

Groundwall insulation takes on many different forms depending on the type of machine and the manufacturer's practices. Groundwall insulation is generally a dry-type, multi-layered system comprised of various bonded and filled insulating materials. Mica based products are generally preferred in high voltage machines for at least a part of the groundwall insulation system. Typical insulation systems in use are: – Varnished Cambric (Class 105 Insulation): Due to the absence of mica, this insulation

system was usually restricted to windings rated 2300 V and below. Heat transfer is relatively poor as is resistance to ingress of moisture and oil. A typical temperature rise rating for such windings is 50°C.

– Shellac Micafolium (Class 130 Insulation): In this system mica flakes are bonded together by shellac to form sheets. These sheets are wrapped and hot pressed around the slot section of the coil. The end windings are insulated with tape, such as asphalt-mica, or sometimes only with varnished cambric. Due to the evaporation of the volatiles in the



Working Group Copy IEEE P56 Rev. 16 Draft shellac this system may have a high void content and is thus susceptible to partial discharge damage as well as to reduced heat transfer.

– Asphalt Micafolium (Class 130 Insulation): As above, except that asphalt is substituted for shellac in the slot section.

– Asphalt Bonded Mica Tape (Class 130 Insulation): The entire coil is insulated with asphalt bonded half lapped mica tape. The mica flakes are bonded together with asphalt and attached to a paper tape. By the 1960s some manufacturers added a polyester terephthalate film (Mylar™) tape to allow greater tension to be used during the taping operation. It was common to apply asphalt varnish as the coil was being taped. Some manufacturers used an autoclave in which vacuum was drawn to remove volatiles, followed by flooding of the tank with hot asphalt and application of pressure in order to consolidate the layers of asphalt mica tape. Prior to installation in the stator core it was common to heat the coils to render them flexible and thus achieve a better fit in the slot. Lift coils had to be heated to facilitate bending at the knuckle. Asphalt mica stator coils can be susceptible to delamination or “puffing” as a result of overload, poor ventilation, or the use of unsuitable asphalt varnish. The asphalt bonded mica tape insulation system is also vulnerable to tape separation near the end of the stator core as a result of thermal cycling. This is especially true of windings in long cores such as turbo alternators.

– Polyester Bonded Mica Tape [VPI] (Class 130 Insulation): This system was first introduced in North America in the early 1950s. At that time it involved exclusively the use of large flake mica. With the passage of time it was found that lower power-factor tip-up values could be obtained if mica paper tape was used immediately adjacent to the copper. In the vacuum/pressure impregnation (VPI) process the coils or bars are placed in an autoclave and subjected to a high vacuum for drying purposes. The tank is then flooded with polyester resin and pressure is applied to achieve the desired impregnation. Following removal from the tank the coils are placed in mold angles. Heated presses are used to cure the resin. In the case of the end arms, splints are applied with shrink tape, and strap-on heaters are used to cure the resin.

– Epoxy Bonded Mica Tape [VPI] (Class 155 Insulation): In recent years it has become common to use epoxy in place of polyester resin in VPI operations in order to obtain improved bonding characteristics and a higher temperature classification. One impregnating epoxy resin that has been used in a VPI process is a catalyzed blend of Bisphenol-A resin. Once the catalyst has been added the resin must be refrigerated in order to reduce reactivity. A loss of cooling incident is documented in the 1981 EEIC Proceedings, pp 51-55.

– Global VPI (Class 155 Insulation): The coils or bars are insulated with mica tape over which are applied the semi-conducting slot armor tape and the stress control tape on the end arms. The coils or bars are installed in an unbonded state in the stator core. The entire core and winding is then placed in an autoclave for impregnation with epoxy resin following which the resin is cured. Critical to the success of this process is the maintenance of a suitable slip plane to allow for the longitudinal expansion and contraction of the copper with respect to the core.

– Epoxy Bonded Mica Tape [Resin Rich] (Class 155 Insulation): This system involves a mica paper formulation in which the mica flakelettes are deposited on a glass fiber backing tape. An uncured (B stage) epoxy resin is applied during the manufacture of the



Working Group Copy IEEE P56 Rev. 16 Draft tape. A polyester film (Dacron) layer may be included to make it easier to handle the tape during its application to coils or bars. Once the bars or coils have been insulated, the B stage resin is cured under elevated temperature and pressure in one of two ways. In the heated press method, mold angles are applied to the slot section of the coil or bar prior to insertion in the heated press. The rate of temperature and pressure increase is critical in the bonding process. Additional heaters are used with splints and shrink tape to cure the tape on the end arms. An autoclave process is used by some manufacturers that involves three major steps: 1) high vacuum to remove moisture, volatiles and trapped air; 2) elevated temperature at reduced pressure to allow the B stage resin to flow to a limited extent; and 3) high temperature and high pressure while the resin cures.

– Silicone Rubber (Class 180 Insulation): Silicone rubber is a material that is suitable for use at high temperatures. When used as the groundwall insulation of a stator coil it usually has a fiberglass backing. In the absence of mica it is commonly restricted to lower voltages, e.g. 4160 V or lower. A common temperature rise for such windings is 85°C. A disadvantage of silicone rubber is its vulnerability to mechanical damage.

– Varnish Dip & Bake (Class 130 Insulation or Class 155 Insulation): Especially at lower voltages (2300 V and below) and for small sized stators, such as some motors, the varnish dip/oven bake process is common. In this case the stator coils are insulated with fully cured mica tape, following which the coils are installed in the stator core and connected, together with all applicable lashings and bracing. The completed stator is then dipped in a vat of suitable insulating varnish, followed by baking in an oven to drive off the volatiles.

6.2.4 Semi-conductive Slot Coating

The surface of slot portions of stator coils and bars, including several centimeters of the coil beyond the core, is normally semi-conducting. The semi-conducting characteristic is accomplished by the application of semi-conducting varnish over the armor tape (if any) or by the use of semi-conducting armor tapes. These treatments are often referred to as conductive and are generally applied to machines with rated voltage of 4 kV and above. Thermosetting stator coils and bars usually require some form of side packing to achieve an interference fit with the slot.

6.2.5 Stress Control Coating

Due to the semi-conductive slot coating, the surface of the stator coils and bars just outside of the core is at ground potential. The surface of the end turn insulation, however, is typically not at ground potential. In order to linearize the electric field distribution along the coil or bar end turn a stress control coating is applied. This coating can be made from varnishes or tapes having a non-linear resistance characteristic through the use of silicon carbide. These stress control coatings are often referred to as grading paint or tape and are generally applied to stator coils and bars insulated with thermosetting materials in machines with rated voltage of 4 kV and above.



Working Group Copy IEEE P56 Rev. 16 Draft Such coatings were less necessary for thermoplastic insulation systems which operate at a lower electric stress.

6.2.6 Commutator Insulation

Installed on the connection side of a rotating armature of a dc electric machine, a commutator is a cylindrical assembly of wedge-shaped copper segments separated from each other and ground by insulation that is usually mica based. This structure is mechanically locked together by various techniques including V-grooves, cones and support rings at commutator bar ends, steel shrink rings over ring insulation on the commutator surface, and high-tension fiberglass bands applied into grooves in the commutator surface. Small-size units are often compression molded with a high-strength molding compound.

6.2.7 Support Insulation

Supports may be nonmetallic or metallic in design. Nonmetallic supports include blocks, spacers, ties, slot wedges, slot fillers, etc. Nonmetallic supports are made of various insulating materials including wood, molded parts, compressed laminates of cotton, asbestos, glass or synthetic fibers, and felt pads impregnated with various types of bonding agents including phenolic, polyester and epoxy resins. These materials have a range of thermal classifications as well as different physical and electrical properties. Metallic supports such as the surge (bull) ring and its supports are insulated where necessary.

6.3 Wound Rotor Windings (3 Phase Induction Machines)

The rotating secondary windings of wound rotor ac induction machines are similar to armature windings. The three phase winding, with its associated leads and collector rings is the secondary current carrying winding of the machine, but the current may be higher than the stator current. The voltage is usually less than 3000 volts, but there is no real limit. The coils have ground insulation and may have turn insulation. Wedges, blocks, and other insulated mechanical supports are a part of the wound rotor winding assembly. They also have banding or retaining rings because of the centripetal forces. The coils of the winding may be constructed of insulated copper straps mounted vertically (radially) in partially closed slots or may be constructed with strands similar to an armature winding in an open slot.

6.3.1 Partially closed slots - Strap Windings (Generally used on high-speed machines)

The straps may be half straps or open ended full coils. There are usually two to four straps inserted radially in each partially closed slot. Each strap is insulated with insulation appropriate to the secondary voltage, usually mica. This insulation also serves as the ground insulation.

6.3.2 Open slots

Coils made of copper wire(s) wound in loops (Generally low speed)



Working Group Copy IEEE P56 Rev. 16 Draft 6.3.3 Strand (wire) Insulation

The individual strands of wound rotor coil conductors are usually wound in loops as full coils. Strand insulation can be made up of organic resin enamels, polymeric films, resin bonded fibers (such as paper, cotton, asbestos, glass, polyester, or combinations thereof) or resin bonded mica. Nomex™ can be used in high temperature machines. The turn voltages are generally not high enough to require mica insulation or dedicated turn insulation.

6.3.4 Ground wall Insulation

See 6.2.3

6.3.5 Collector Rings

There are three collector rings mounted on a steel hub with insulation to separate the rings from the hub. This insulation may be mica or other materials applied oversize, then turned to allow for a shrink fit of the rings. The three rings are separated from one another by a distance dependant on the secondary 3 phase voltage. A track resistant compound is usually applied over all insulating surfaces of the collector.

6.4 Field Winding Insulation

6.4.1 Field Windings

The field windings of ac machines are rotating and can be either salient pole or cylindrical type. The field coils of dc machines are stationary and are constructed in a similar fashion to ac rotating field coils, except that they need not be built to withstand the effect of rotational forces. dc field coils are usually a complex assembly of exciting and commutating coils, each fitted over and insulated from a pole piece. Some coils contain multiple windings.

In all cases, field windings have turn and ground insulation, insulated mechanical supports, and lead insulation. In addition, ac machines may have collector ring insulation, retaining ring insulation, and banding insulation.

6.4.2 Turn (Conductor) Insulation

Turn insulation on wire-wound coils usually incorporates a thin insulating layer on the strand itself. Various materials have been used such as asbestos, cotton, fiberglass, papers, micas, and synthetic materials.

Turn insulation on strap-wound coils usually incorporates various forms of tape or strip material with resin bonding.

6.4.3 Ground Insulation

Various types of ground insulation are used on the field coils of rotating machines. A variety of organic and inorganic materials are used.



Working Group Copy IEEE P56 Rev. 16 Draft 6.4.4 Collector Insulation

The insulation used on collector rings and leads must be adequate both for support and creepage to the grounded shaft. The insulation usually consists of laminated fibers or mica suitably bonded or impregnated.

6.4.5 Brush-Rigging Insulation

The insulated components on brush riggings are generally made from molding compounds, laminated boards, or tubes made from paper, cotton, or glass fibers suitably bonded and impregnated. Moisture-resistant surfaces are very important for these components.

6.5 Core and Frame-Assembly Insulation

6.5.1 Stator Core Interlaminar Insulation (Core Plate Insulation)

Stator cores are built up with thin laminations insulated from each other to reduce core losses. A variety of thin insulating coatings, such as varnish, water-glass, and other chemical deposits, are used. On very small machines or machines where the volts per turn in the stator winding is low, iron oxides produced by appropriate annealing processes are often used.

6.5.2 Insulation Punchings

In some designs, where deemed necessary, insulation punchings are used to supplement the interlaminar insulation. These layers are also sometimes used as backing for ventilation layers where there are spot welded or riveted vent fingers.

6.5.3 Core Tightening Through-Bolt Insulation

The core tightening through bolts are insulated from ground along their length with insulating materials suitably bonded. Bolt-end hardware, such as nuts and washers, must also be insulated from ground. Key bars or bolts, used at the outer diameter of the core for tightening, usually require no insulation.

6.6 Other Insulating Parts

Insulation is sometimes used on bearings or bearing brackets to eliminate shaft currents. Insulation is also used to isolate temperature-measuring devices such as thermocouples (TCs), resistance temperature detectors (RTDs) and thermistors.

NOTE: Annex has been added.




7. Service Conditions Affecting Insulation Life [Ian Culbert – Clause head]

Electric machines and their insulation systems are subjected to mechanical, electrical, thermal and environmental stresses that give rise to many deteriorating influences. The degradation rate of electrical insulation systems is substantially greater when these stresses act simultaneously than when they are sequentially applied.[Ref.: R. Bartnikas and R. Morin, " Multi-stress aging of stator bars with electrical, thermal and mechanical stresses as simultanious acceleration factors", IEEE Trans. on Energy Conversion, Vol.19, pp.702 - 714,2004.] A thermal stress applied independently prior to application of electrical stress is appreciably less deleterious than when the two stresses act simultaneously. [R. Bartnikas and R. Morin, "Analysis of multi-stress accelerated aged stator bars using a three phase test arrangement", IEEE Trans. on Energy Conversion, Vol.21, pp.162 - 171, 2006.] The synergistic effects of environmental stresses are more difficult to ascertain. An insulation system operating in the presence of nuclear radiation may, in some cases, have a prolonged life expectancy. For example, polymeric based insulating systems operating in the presence of ionizing radiation may exhibit initially a prolonged life due to radiation induced crosslinking of the polymers. [Ref. F. J. Campbell in Engineering Dielectrics , Vol. II A, Electrical Properties of Solid Insulating Materials: Molecular Structure and Electrical Behavior, R. Bartnikas and R. M. Eichhorn, Editors, STP 783, ASTM, Philadelphia / West Conshohocken PA,1983.]

7.1 Aging Mechanisms

No machine insulation system that is economically produced is expected to last forever. The thermal, mechanical, electrical, and environmental stresses will gradually reduce the electrical and mechanical strength of the insulating materials. At some point, the materials will have aged significantly. In such a case, the insulation breaks down or cracks under the normal operating voltages or as a result of a transient electrical (e.g., lightning or switching voltage surges) or mechanical (from motor switch-on in-rush current or current transients from faults in the power system, which cause large magnetic field impulses) situation. If the insulation breakdown occurs in the stator ground-wall or turn insulation, this will rapidly lead to high-power-frequency fault currents and circuit-breaker operation. Failure of the strand insulation in stators or the turn (and to a limited degree the ground) insulation in rotors will not result in motor or generator failure. However, performance will be adversely affected since the magnetic field intensities will be weaker and non-symmetrical, leading to vibration, or the efficiency of the machine will be reduced due to circulating currents. The circulating currents will cause additional heating, which will accelerate insulation aging processes.

Additional failure processes can occur due to on-off cycling of motors or load cycling of generators. Such cycling leads to large and sometimes rapid swings in winding temperatures. Such temperature swings can lead to different thermally induced growth among the different winding components, developing shear stresses between the components. For example, when a large generator goes from no load to full load in a few minutes, the stator winding copper



Working Group Copy IEEE P56 Rev. 16 Draft temperature goes from a low temperature to a high temperature, and the copper grows axially along the slot. Immediately after the load increase, the insulation temperature remains relatviely low. The result is that the groundwall insulation experiences a much smaller axial growth. Since the copper expands more than the groundwall, a shear stress develops between the conductor and the insulation. With a sufficient number of load cycles, the groundwall may debond away from the conductors, creating an air gap, leading to failure from partial discharges.

The single and multistress interactions, together with load cycling, yield about 20 different identifiable failure processes in stator windings, and about 10 mechanisms in rotor windings. See Tables 1 and 2 for a summary of the main failure processes. Which process will occur in a specific machine and how quickly the failure will occur will depend on: – The design stress levels (i.e., operating temperatures, mechanical stress, etc.) the machine

designer employed, and how close these levels are to the insulation material capabilities. – How well the windings were manufactured and assembled. – The operating environment the user provides, that is, is the machine run at constant load

or cycled, is it over-loaded; are oil, moisture, or abrasive particles present? – How well the user maintains the windings, that is, keeping them clean, keeping them

tight to prevent vibration, etc.

Knowing which deterioration processes are occurring is important, since any winding maintenance to extend winding life should directly address the processes.

7.2 AC Stator Winding Aging Mechanisms

7.2.1 Thermal Deterioration

Thermal stresses induce thermal aging of the insulation system whose rate is increased with the service temperature to which it is subjected. [Ref 2 from Ray] The deterioration rate doubles approximately for every 10 degrees Celsius rise in temperature. The aging rate is influenced by unusually high temperatures of operation caused by conditions such as overload, high ambient temperature, loss of or restricted ventilation, loss of cooling liquid or air and foreign materials deposited on windings. Long-term operation of a winding at high temperature leads to embrittlement of the insulation bonding resins and delamination.

7.2.2 Thermal Cycling

Additional failure processes can occur due to on-off, cycling of motors or load cycling of generators. Such cycling leads to large and sometimes rapid swings in winding and to a lesser extent core temperatures with resulting accelerated insulation aging. Such temperature swings can lead to different thermally induced growth among the different winding components, developing shear stresses between them. For example, when a large generator goes from no load to full load in a few minutes, the stator winding copper temperature goes from a low temperature to a high temperature, and the copper grows axially along the slot. Immediately after the load increase, the insulation temperature remains relatively low. The result is that the groundwall insulation experiences a much smaller axial growth than the copper expands more than the



Working Group Copy IEEE P56 Rev. 16 Draft groundwall to create a shear stress between the two. With a sufficient number of load cycles, the groundwall may de-separate from the conductors, creating an air gap between the two that can cause a failure from partial discharges.

7.2.3 Poor Impregnation

Conventional form windings (coils and bars) are impregnated by the resin-rich or VPI process. This is done to seal them against moisture and other contaminants, improve heat transfer to the core, consolidate the conductor stack to prevent abrasion of strand and conductor insulation from relative movement and minimize void size and content to control the level of partial discharge (PD) activity. In addition if the winding is globally VPI’d the impregnating resin ensures that the windings remain tight in the core slots and helps to consolidate the endwinding bracing system. Consequently, if the winding is poorly impregnated the following aging mechanisms can result;

– Insulation degradation from PD attack – Abrasion of Turn, or strand insulation from relative movement – Abrasion of slot ground insulation from radial movement in the slot – Failure from contaminant ingress – Failure from high endwinding vibration.

Inadequate impregnation of the groundwall insulation near the conductor stack is more likely in global VPI stator windings. For voltage ratings of 4 kV and above failures from the above aging mechanisms can occur in as short a period as 2 years

7.2.4 Internal Water Leaks

This problem relates to large hydro and steam turbine generators with direct water- cooled stator windings. Such machines generally have ratings greater than 200 MVA. If the generator is a pressurized hydrogen design the possibility of water leaking into the stator winding insulation is reduced, but not eliminated. The most likely causes of water leaks are improper re-assembly of bar-to-teflon hose connections during maintenance and crevice corrosion cracking of brazed joint between the bar nozzle and bar copper strands. Other possible causes are porosity of the brazing between bar nozzles and strands and bar strand cracks.

Small water leaks can have the following three effects on the stator winding insulation; – Reduction of the ground insulation dielectric strength which will make the winding more

prone to failure if an over-voltage occurs due to switchyard faults, or AC/DC hipot testing

– Slot PD, if groundwall insulation delamination, due to water ingress, progresses to the slot regions of line end bars this can lead to winding failure.

– Reduction of the mechanical strength of the groundwall insulation. This makes the winding susceptible to failure from high electromechanical forces induced by current surges from the power system, or out-of-phase synchronization.



Working Group Copy IEEE P56 Rev. 16 Draft 7.2.5 Loose Coils, or Bars in the Slot

This problem normally occurs in coils, or bars with, thermo-set, resin-rich, and individual VPI insulation systems, which have cured and consolidated slot sections, if the slot wedging and packing system is poorly designed.

If a coil or bar is not tightly held in its core slot it will vibrate primarily in the radial direction under the influence of magnetically induced mechanical forces with a frequency of at twice the power supply frequency, i.e., 120 Hz for 60 Hz power. In addition the magnitude of these forces is proportional to the square of the current passing through the conductors and so they vary with machine output. The resulting relative movement between the winding ground insulation outer surface and the core, which is somewhat serrated, abrades the semi-conductive coating (if fitted) and then the groundwall insulation. If not detected soon enough this insulation degradation mechanism can result in a winding ground fault.

7.2.6 Semi-conductive Coating Degradation

This aging process deals with deterioration of the semi-conductive coating on the slot section coil or bar in the absence of vibration due to looseness. Normally form wound stator windings, rated 4 kV and above, have such coatings in the form of carbon-loaded tapes, or paints.

Poor quality materials and thermal aging with service are the most likely causes of semi-conductive coating degradation. Both of these can lead to the material becoming non-conductive, in localized areas, by oxidation of the carbon particles. If this occurs on high voltage coils, PD activity between the bar/coil and core will start to develop in the affected areas. Air-cooled stator windings are more susceptible to this problem since the pressurized gas in hydrogen-cooled machines suppresses the resulting PD activity. If slot PD from this aging mechanism develops in an air-cooled machine it will create ozone, which is a very chemically reactive gas, which combines with other gasses in the air to create nitric acid. This nitric acid can attack insulation and packing materials in the slot. This can lead to coil/bar looseness, which if not detected early enough, can cause degradation from insulation from abrasion as described in 7.2.7. Experience indicates that coils/bars with paint-based coatings are more susceptible to this problem. This aging mechanism will take many decades to cause a failure if the winding is kept tight in the slot since mica based insulation has a high PD resistance.

7.2.7 Electrical / Mechanical (Contact) Erosion

This occurs at the neutral end of the winding, where the voltage stresses are not as prominent and occurs as a result of poor electrical contact of the coil to the grounded slot. The primary reason for this is due to high contact resistance between the semi-conductive layer and the ground plane combined with the coils being loose in the slot, which leads to contact arcing and erosion of the semi-conductive layer and erosion of the main wall insulation due to sparking.

Here a distinction must be made between contact arcing and contact sparking. The latter effect is caused by an electrostatic mechanism, which arises as two surfaces with different work functions contact each other intermittently. The two surfaces acquire surface charges that create potential differences, which initiate spark type discharges when they exceed or become equal to the



Working Group Copy IEEE P56 Rev. 16 Draft breakdown voltage of the changing gap separation between the two approaching surfaces, [B] Unlike the arcing mechanism, the electrostatically induced sparking process requires no externally applied voltage in order to sustain itself. Such sparking generally results from the slot semi-conductive coating being too conductive and the bars not being securely held within the slot/or the global VPI process has isolated the bars.

7.2.8 Semi-conductive/Grading Coating Interface Failure

This problem is associated with the semi-conductive coating deterioration process described in 7.2.6 above and only occurs in form windings that have both slot semi-conductive and an overlapping stress control tape or paint coating (normally silicone carbide) just outside the slot. The stress grading coating has a characteristic of having a resistance that is high in areas of low electrical stress and low in areas of high electrical stress. Since the end winding surfaces are at the same potential as the conductors its purpose is to make the electrical stress at the end of the semi-conductive coating more uniform. Again this problem is normally confined to windings rated 6 kV and above.

If the overlapping electrical connection between the semi-conductive and grading coatings degrades to become non-conductive the grading coating “floats’ and rises to the voltage of the copper conductors, due to capacitive coupling. If this happens a very high voltage, separated by a small gap, develops between the grading and semi-conductive coatings on bars/coils near the line ends. This air gap breaks down, resulting in discharges over the surface of the coil/bar between the two different coatings. Since this discharge mechanism is parallel to the insulation surface degradation from it is a very slow process. However, if it is present it may be an indication that the more serious problem of slot semi-conductive coating degradation is also occurring.

7.2.9 Electrical Stresses

The service life of an insulation system varies inversely with the voltage, V-n, where the exponent n is dependent upon the specific voltage-dependent ageing mechanism. Abnormal voltages, exceeding the specified service rating, such as those caused by switching, lightning surges, or certain power supplies further accelerates the ageing process. Partial discharges, at higher operating voltages may produce several undesirable effects, such as chemical degradation, localized heating, ionic bombardment and physical erosion. These effects tend to increase the aging rate

Electrical insulation that operates in a high-stress field is subject to deteriorating influences not present at lower stress levels. Partial discharge (corona) can cause degradation of insulation in two ways, by chemical effects and by ion bombardment. This occurs when the voltage gradient on gas molecules in void spaces in the insulation exceeds a certain value, depending on the nature of the gas and its pressure and temperature. Ozone and nitrogen oxides that can attack organic materials in the insulation may be formed. Ozone, nitrogen oxides and oxalic acid crystals may be formed when polymeric materials are exposed to discharges; the symptoms of this type of degradation are mostly whitish deposits. The effects of corona are not manifested to a noticeable degree in hydrogen cooled machines, since the oxygen content is small and degradation is principally physical pitting and erosion.



Working Group Copy IEEE P56 Rev. 16 Draft Positive and negative ions, as well as electrons, generated during each partial discharge event, are accelerated in the direction of the electrical field. This subjects the boundaries of the occluded cavities to regularly occurring bombardment by the charged particles in each half cycle of voltage. The resulting deterioration may manifest itself in the formation of electrical trees propagating into the dielectric from the void boundaries and tracking along the inside walls.

7.2.10 Electrical Tracking due to Contamination

If electrical tracking due to conductive coating contamination does occur current will flow across the surfaces of the winding, especially in the end winding regions. Such contamination can result from the ingress of oil from bearings, or hydrogen seals in combination with moisture, or carbon particles (such as from carbon brushes) from the atmosphere. Open-enclosure, air cooled machines are most susceptible to this degradation mechanism.

If the insulation surfaces and blocking between adjacent high voltage sections of end winding, or circuit ring buses in different phases become contaminated with conductive materials currents will flow between them. This is so because the surfaces of these sections of winding will rise to the potential of their conductors, between which voltages approaching the phase-to-phase value are present. If the contamination resistance was uniform little deteriorating would likely result from these currents. However, dry areas where the resistance is much higher are commonly present and as a result the whole voltage can appear across these high resistances to cause electrical breakdown of the adjacent air or hydrogen. This discharge degrades and may carbonize the underlying organic resin and tape. This area then be comes very conductive and the high electrical stress then transfers to another dry area. In the longer term an electrical tack can develop between phases and can start eating into the groundwall insulation and this can eventually lead to a phase-to-phase failure. This mechanism is usually very slow and can take more than 10 years after initiation to cause a failure.

7.2.11 Voltage Surges

Voltages surges in stator winding insulation systems in motors and generators are transient bursts of relatively high voltage that increase the electrical stress beyond that which normally occurs in service. Such voltages can occur from:

– Lightening strikes – Power system ground faults – Out-of-phase generator breaker closing – Motor circuit breaker closing and opening – Voltage source motor converter drives

It is unlikely that the first four surge sources will age the winding insulation, but will cause it to fail if the dielectric strength is inadequate, or has been reduced by some other aging mechanism. This is not the case with converter drives since they can continuously impose high voltage fast rise time that can cause insulation aging.

Voltage surges from the first four sources are most likely to affect stator windings in motors or generators with multi-turn stator coils. When a high frequency voltage is applied the voltage



Working Group Copy IEEE P56 Rev. 16 Draft distribution is non-linear with a much greater percentage of the voltage appearing across the first coil connected to the phase terminal. This non-uniform voltage distribution occurs because the series inductive impedance of the winding is relatively large compared to the capacitive impedance to ground at this high frequency. In addition it has been shown that this voltage does not evenly distribute across the first coil. The consequence of this is that in the line end coils inter-turn voltages that can be orders of magnitude higher than the 50/60 Hz steady state value can be induced. These surges can cause line end coil turn insulation failures if the turn insulation dielectric strength is too low or has been weakened by insulation aging, e.g. thermal aging or PD activity at ground insulation-to-conductor interface. Such failures can also occur in motors fed from voltage source inverters, but most manufacturers appreciate the need to strengthen turn insulation by using mica to avoid failures from these repetitive voltages.

The voltage waveform from a pulse width modulated (PWM) voltage source converter can lead to increased groundwall insulation heating, which can increase the winding temperature and thus accelerate the normal thermal aging processes described above.

But the PD may be larger and more frequent with a converter, because the peak voltages are usually higher than the peak voltage from a sinusoidal supply. The peak-to-peak voltage can be higher than from a 50/60 Hz supply due to the transmission line effects that may cause the step voltage changes that occur with converters to possibly double.

The partly conductive coatings that normally cover the coil insulation in the stator slot and the silicon carbide material at the slot exits are intended to suppress the probability of PD occurring on the coil surface in the slot and just outside of it. Several studies have shown that under PWM voltage, these coatings will operate at higher temperatures and thus increase the rate of thermal aging, if they are not properly designed [5,8,9]. Since PWM voltage waveforms contain voltages at high frequencies (from the rise time of the voltage steps and the PWM switching rate), higher capacitive currents flow through the groundwall and then through the PD suppression coatings. These higher currents create higher I2R losses in the coatings than would occur under 60 Hz operation increasing the operating temperature of the coatings. The effect is exacerbated because the higher frequencies also cause the silicon carbide materials to be less effective in linearizing the voltage along the surface of the coils – which tends to concentrate the heating to a shorter area.

7.2.12 Environmental Factors

In addition to the Electrical Tracking due to Contamination discussed in Section 7.2.9 there are other environmental factors that can cause winding insulation to age and fail. These include: – Chemical attack – Abrasive particles – Ionizing radiation

These are discussed as follows;

Chemical Attack



Working Group Copy IEEE P56 Rev. 16 Draft Most types of older insulation systems are prone to chemically induced degradation due to the presence of solvents, oil, water, or other chemicals and gasses. For example magnet wire insulation materials such as polyester can soften and swell from exposure to moisture. Groundwall insulation using asphalt, varnish and some of the earlier polyester bonding agents are prone to softening and swelling and loss of mechanical strength when exposed to moisture, or chemicals.

Softening of insulation makes it susceptible to cold flow, i.e., the insulation gradually becomes thinner in areas where mechanical pressure is applied. If the thickness of the insulation is reduced to such an extent that it can no longer normal or higher transient voltages then it will fail.

Modern stator winding insulation systems are more resistant to most types of chemical attack, but can still be affected by some chemicals and gasses. However, if epoxy is exposed to oil and water for many years it will eventually start to degrade. Also, if it is known that a winding will be exposed to a humid environment it can be designed to be sealed against moisture, e.g., nuclear plant motors that have to survive under main steam line break conditions.

Abrasive Particles

Abrasive particles in the cooling gas stream can grind away stator winding insulation to the extent that it fails under normal or transient voltage conditions. This is most likely to occur in machines with open-air ventilation I operating environments where sand, iron ore dust, etc. are present in the operating environment.

If abrasive particles are enter the machine enclosure they will be blown though the stator windings at high velocity suspended in the cooling air stream. If there is sufficient quantity of these materials at a high enough velocity they will abrade the insulation surfaces parallel to their flow to reduce their thickness. Such abrasion can eventually expose the winding copper, which leads to failures. The most likely areas to be affected by abrasion are the endwindings and the sections of winding bridging radial core cooling ducts.

Ionizing Radiation

[Input from Ray Bartnikas]

Polymers are particularly susceptible to ionizing radiation, which is found to lead to the production of electronically excited states of molecules and the formation of radicals. [C] The extent of the resulting polymer degration depends upon the absorbed radiation dose and dose rate. [D, E] The radiation induced degradation rate of the insulating material is furthermore influenced by the electrical, thermal and mechanical stresses to which the insulation may be simultaneously subjected. As the mechanical properties of the insulating material deteriorate, electrical breakdown ensues when the material reaches its brittle fracture state.

In cases where the insulating material is only exposed to short intermittent periods of ionizing radiation, its electrical properties such as conductivity and dissipation factor will exhibit an increase during the irradiation period as well as for a short period following the removal of the



Working Group Copy IEEE P56 Rev. 16 Draft radiation source. The increase in the conductivity and dielectric losses is a direct result of the electrons being excited into the conduction band. The finite conductivity, remaining after the removal of radiation of radiation, is caused by the still mobile electrons which had fallen into shallow traps. However, eventually these conduction electrons fall into deep traps, thus becoming immobilized from which they can be only emitted when the next radiation exposure takes place. The copious supply of free electrons during the irradiation periods will also lower the partial discharge inception and extinction voltages; the abundant availability of free electrons may facilitate the occurrence of pseudo-glow discharges.

[C] Ref. F. J. Campbell in Engineering Dielectrics , Vol. II A, Electrical Properties of Solid Insulating Materials: Molecular Structure and Electrical Behavior, R. Bartnikas and R. M. Eichhorn, Editors, STP 783, ASTM, Philadelphia / West Conshohocken PA,1983.] [D] ASTM Method D2953, Classification System for Polymeric Materials for Service in Ionizing Radiation. [ E ] ASTM Method D1672 Recommended Practice for Exposure of Polymeric Materials to High Energy Radiation.

There are a number of high-voltage motor applications, such as reactor coolant pump motors within the reactor containment area of a nuclear generating station, where radiation levels can be high. If acceptable insulation life is to be achieved, the insulation systems in these machines must contain materials with a high radiation resistance to prevent rapid deterioration of mechanical properties of the binders and backing materials.

Materials such as ceramics, mica, glass and epoxy resins are known to be only slightly affected by the radiation levels seen in these applications. Organic backing, bracing, and bonding materials, on the other hand, are strongly affected by ionizing radiation while polymers with aromatic rings (see tables B-2 and B-3) will tolerate larger doses without deterioration. Because of the susceptibility of organic insulations to radiation damage, therefore, great care is required in selecting the proper backing, insulation bonding, and bracing materials for the winding of machines to be used in a radiation environment.

The types of radiation from nuclear reactors are primarily alpha and beta particles, gamma rays, and neutrons. Since alpha and beta particles do not penetrate the reactor shield, they are not a significant factor. On the other hand, gamma rays and neutrons do penetrate this shield, reacting with insulating materials to produce electrons that can be responsible for radiation damage.

The two molecular changes that may be produced by radiation in an organic insulation are cross-linking of molecular chains and bond scission, or cutting of polymer chains. Cross-linking builds up the molecular structure, initially increasing tensile strength, but then reduces elongation and eventually results in the loss of impact strength. This changes rubbery or plastic materials into hard, brittle solids. Scission breaks down molecular size, reduces tensile strength, and usually adversely affects other properties.



Working Group Copy IEEE P56 Rev. 16 Draft Most modern high-voltage insulation systems contain mica with backing materials such as glass or Dacron TM glass fibres, all of which have fairly high radiation resistance. The susceptible part of the insulation systems is the organic impregnating resin used to bind these materials together. Resins used must therefore be carefully selected to assure good insulation system radiation resistance.

As already indicated, materials susceptible to radiation aging lose their mechanical properties and as a consequence become brittle or weak. This makes them susceptible to mechanical failure under the stresses imposed, e.g., if the bonding resin becomes brittle it will crack and delamination may occur, and the failure mechanism associated with this will cause electrical breakdown of the insulation. It is for this reason that insulation systems used in motors, that operate in high radiation areas, are environmentally qualified by simulated aging and testing. IEEE 334 gives guidelines on how such qualification testing, which includes radiation aging, should be performed. However, the information in these tables may not apply if the radiation environment is very harsh.

7.2.13 Endwinding Vibration

Normal 50 or 60 Hz current flowing through the stator coils and bars creates large (100 or 120 Hz) magnetic forces. If the endwinding is not adequately braced, the coils/bars vibrate relative to there bracing, gradually abrading the insulation. The problem is most likely to occur on form-wound two- and four-pole machines, since such machines have long end-windings, which may have resonant frequencies close to the frequency of the magnetic forces. End-winding vibration is one of the most common failure mechanisms of large steam turbine generators rated at several hundred megawatts and above. Any form-wound stator can fail due to this problem if the end-windings are not adequately supported.

If the endwinding support is inadequate, the bars or coils will begin to vibrate. In form-wound machines with long end-windings, this vibration in the radial and circumferential directions will pivot at the stator slot exit, assuming the coils/bars are tight in their slots. . Thus, the coil/bar insulation may eventually fatigue crack just outside of the slot. The fatigue cracking can occur even if the coils themselves are tightly held together and will eventually lead to a phase-to-ground fault.

If the blocking and bracing is loose, then coils and bars can also vibrate relative to one another in the endwinding. The bars and coils will then rub against each other, the blocks, surge rings, support cone, and/or other end-winding support structures. The rubbing will abrade through the insulation. Fiberglass roving is especially hard and very effective in cutting through the groundwall insulation. If not corrected, sufficient groundwall insulation can be abraded so that a phase-to-ground failure occurs.

There are additional consequences of end-winding vibration in directly cooled stator windings. In water-cooled stators, the 100 or 120 Hz end-winding vibration can fatigue crack the brazed connections of one bar to another and/or the water nozzles. This can allow water to leak into the insulation (see Section 7.2.4). Also, some failures have occurred because hydrogen becomes entrained in the stator cooling water. The hydrogen bubbles combined with excessive vibration



Working Group Copy IEEE P56 Rev. 16 Draft lead to cavitation of the copper conductors. The thinning copper tubes eventually crack, allowing large amounts of hydrogen into the cooling water.

In direct hydrogen-cooled windings, end-winding vibration has led to broken copper strands from copper high cycle fatigue. The strand arcs at the break, causing localized overheating of the insulation. The insulation eventually melts, precipitating a ground fault. Arcing has also occurred at broken resistor connections (due to vibration). The resistors are installed to give a fixed potential to the hollow hydrogen tubes within the bar. Direct hydrogen inner cooled stator windings require a longer end winding than other designs (everything else being equal), since the exposed metallic tubes require longer creepage distances. This makes such windings more susceptible to end-winding vibration problems.

7.3 Cylindrical (Round) Rotor Winding Aging Mechanisms

This section deals with the common aging mechanisms found in cylindrical rotor windings that are also called round rotor winding that are used mainly in medium to large turbine generators, but are also starting to appear in large 2-pole motors supplied power from variable speed converter drives.

7.3.1 Thermal Aging

In older windings of this type contained organic components, which thermally age and shrunk in service and allowed the inorganic components to be displaced by cyclic mechanical forces experienced in operation. This lead to cracks and gaps in the ground and turn insulation and electrical failure.

Starting in the 1950’s, modern thermosetting resins and glass fabrics began to replace the older organic materials. Endwinding blocking was changed from phenolic-bonded asbestos cloth laminates to glass cloth with either polyester or epoxy laminating resins. Similar changes were made in the slot cell insulation. These changes raised the temperature class of the insulation to Class 155 (F) and significantly reduced the incidence of thermal aging. Smaller fields are insulated with no-nwoven aramid sheets and tapes, whereas larger fields often use non-woven glass laminates for insulation and blocking.

Glass laminates bonded with epoxy or polyester resins are commonly used for both the turn and the ground insulation in direct-gas-cooled rotor windings. Thermal degradation of these materials may be treated as a chemical rate phenomenon (described by the Arrhenius relationship) and includes loss of volatiles, oxidation, depolymerization, shrinkage, surface cracking, and embrittlement.

Modern windings use glass laminate insulation systems are made from Class 155 (F) insulation materials for operation at Class 130 (B) temperatures. Since the average rotor winding operating temperature is in the range of 60°C to 90°C, there would appear to be an adequate temperature margin. However, the margin is reduced at hot spots in the winding, which cannot be measured directly since some machines have brushless exciters directly connected to the rotor field winding and for those slip rings only an average rotor temperature can be derived from rotor



Working Group Copy IEEE P56 Rev. 16 Draft amps and volts. Depending on the type of the cooling-gas flow system, the estimated hot spot temperature could exceed 130°C. If this is the case thermal aging becomes a factor, particularly where some Class 130 materials have been used. The thermal degradation is less likely on hydrogen-cooled rotors because of the lack of oxygen, which accelerates chemical aging, and because of the operating temperature margin usually available. The higher the temperature, the faster will be the chemical reaction, resulting in shortened life of the insulation under thermal degradation.

7.3.2 Thermal Cycling Operating temperatures have a direct aging effect on the insulation materials, as seen in the previous section. However, when the operating temperature varies due to load changes, start-stops, etc., additional stresses are set up that accelerate the thermal aging process. The temperature changes cause the expansion and contraction of the winding copper relative to the insulation, giving rise to insulation aging from abrasion. Unless the rotor winding design can accommodate this movement of the copper, additional built-up stresses can damage not only the insulation, but also the winding copper.

During operation, copper losses from the winding, and to a lesser extent windage and stray losses from the rotor forging, cause an increase in temperature of the rotor components. The physical location of the components in the rotor is altered due to the axial thermal expansion caused by the increase in temperature. When the unit is shut down, the rotor cools down and the components contract to their original position, provided that the copper has not been stretched beyond its elastic limit and there is no restriction to their movement. Depending on the duty, this expansion-contraction cycle may be repeated hundreds of times during the life of the unit. The axial movement of the copper tends to abrade the ground insulation, especially toward the end of the rotor slots. Wear imposed on the winding insulation and other components, due to this repeated back and forth movement, results in mechanical aging due to thermal cycling.

Peaking and two-shifted machines are more susceptible to thermal cycling damage compared to base loaded units because of the higher number of start-stop cycles. Longer rotors are likely to experience a higher level of damage due to the larger amount of expansion and relative movement. Similarly, modern air-cooled units, which generally operate at a higher temperature, will be affected more than hydrogen-cooled units.

7.3.3 Abrasion Due to Imbalance or Turning Gear Operation

Two and four-pole rotors in large generators can weigh as much as 100 tons. Moreover, they have components such as the windings that can move independently in the axial, radial, and transverse directions. To ensure long term reliable performance the rotor must run smoothly within acceptable vibration limits under all operating conditions. High vibration can lead to relative movement of the rotor winding components, which, in turn, can lead to insulation and copper abrasion. Considerable effort is required to ensure a mechanically balanced rotor system, starting with the design stage, to factory assembly and testing, site assembly, and setup, to operating practices and monitoring. Nevertheless, this balance is upset at times due to a number of factors, resulting in an increase in rotor vibration, which can cause further insulation damage and even lead to shutdown of the generator.



Working Group Copy IEEE P56 Rev. 16 Draft Although rotor vibration is a mechanical phenomenon, its origin can be electrical or thermal in nature. Because of electrical and thermal stresses that build up during operation, additional forces are superimposed on the distributed weight of the rotor. Initially, these forces may be small; however, the underlying mechanisms being progressive in nature, the forces can become large enough over time to affect the weight distribution of the rotor. Rotor dynamic performance is sensitive to changes in the weight distribution, particularly for two-pole designs in which the longer and thinner rotor forging is more susceptible to bending forces. The increased vibration due to rotor unbalance can cause damage to the rotor components, leading to further imbalance. Ultimately, if relative movement occurs between the copper and the insulation, or the insulation and the rotor body, causes abrasion that can lead to turn shorts or ground faults.

Large turbine generator rotors have to be operated on turning gear at very low speeds (a few rpm) to prevent catenary "sets" in the shaft during unit shutdown. Operation at turning gear speed with low radial Rotational forces on the windings can cause copper-dusting abrasion, particularly when there are two or more conductor sub-strands that are not insulated from one another and the slot side packing is not tight enough to prevent sideways strand movement. The copper particles can lead to turn or ground shorts.

7.3.4 Electrical Tracking form Contamination

Insulation is used to electrically separate the copper winding which operates at voltages up to about 500 volts relative to grounded rotor body. The insulation takes many forms and shapes due to the complex nature of the rotor winding, which must allow for movement due to expansion, and the geometry of the rotor forging, end rings, balance rings, support rings, slip rings, and connection hardware. During winding assembly, care is taken to ensure that the isolation between the live copper winding turns and between copper and ground is maintained at these numerous interfaces. Normally, the creepage paths between turns and to ground are more than adequate. However, during service, contamination at these critical interfaces can reduce the creepage path to such an extent that turn-to-turn and ground shorts could develop.

Hundreds of interfaces exist in a rotor where insulation separates the live parts from the grounded components such as the forging, wedges, retaining rings, and balance rings. Intermittent surface discharge between turns or from the winding components to grounded parts occurs when these insulation interfaces are compromised due to surface contamination. The discharge results in a chemical reaction of the components involved, producing carbon and other chemicals. The products of the reaction lodge themselves across the interface, creating a path of reduced resistance along which subsequent discharges occur.

7.3.5 Repetitive Voltage Surges In generator rotors of this type, the applied voltage is around 500 V DC which is well with the rating of insulating materials used, particularly in modern designs. This voltage further divides between the turns to result in as little as 10 V between them, which is a well within the turn insulation voltage rating. The turn insulation would last indefinitely under these very low



Working Group Copy IEEE P56 Rev. 16 Draft electrical stresses. However, transient over voltages from the excitation supply or system transients can be orders of magnitude higher, and may lead to insulation degradation.

Events internal or external to the excitation system can induce large transient voltages in the rotor windings. The occasional spike may not be harmful, but continuous repetitive spikes from an excitation system can cause gradual deterioration from partial discharges. This aging mechanism is similar to the stator winding aging process caused by converter drives Insulation that has been already weakened by other aging mechanisms is particularly vulnerable to repetitive voltage surges. Voltage surges are most likely to cause turn-to-turn faults since the insulation between turns is the thinnest and is subject to high levels of mechanical stresses.

7.3.6 Rotational Force

At operating speed, rotor winding components are subjected to high mechanical compressive stresses from Rotational forces. In large turbine generators these forces can exceed 1,500 tons at the wedges and 15,000 tons at each retaining ring. Significant tangential forces are also present, particularly during startup and shutdown of the generator. The contribution of the copper conductors to the total stress on the insulation materials is an important factor. Insulation materials made from quality materials and with adequate design margins can endure these compressive forces over long-term operation. However, where the materials are weakened due to inadequate quality control or other aging mechanisms such as thermal aging, the insulation can bend, buckle, and crack under the influence of the large Rotational forces. This can lead to turn-to-turn shorts or ground faults. The insulation materials involved include slot liners, turn insulation, slot packing and pads, bracing materials, and connection insulation. The effects of Rotational forces are a function of the design of the winding slot wedging and endwinding bracing systems, the properties of the materials used and the frequency of start-stop cycles.

7.4 Salient Pole Rotor Winding Aging Mechanisms

There are two types of windings used in salient pole rotors. These are the "strip-on-edge" and "multi-layer wire wound" types. The type used depends on the rating and speed of the machine. The coil insulating materials and their arrangement dependent on the winding construction while the insulation between the coils and poles can be the same for both types the turn insulation is quite different.

7.4.1 Thermal Aging

All insulating and nonmetallic bracing materials deteriorate with time due to the heat from the windings. The rate at which component materials deteriorate is a function of their thermal properties and the temperatures to which they are subjected. If the thermal ratings of component materials have been properly selected, the thermal aging and associated deterioration will occur gradually over an acceptable service life.

In older Class 130 winding material such as asbestos backed mica splittings were used for pole ground insulation and turn insulation on strip-on-edge windings and these were bonded with organic varnishes such as shellac, asbestos board, or phenolic bonded glass fibres were used for pole washers. Also in early insulation systems organic materials were used to insulate conductors



Working Group Copy IEEE P56 Rev. 16 Draft in "multi-layer wire wound "types. These materials are more susceptible to shrinking and cracking under the influence of thermal aging.

Modern salient pole winding designs typically use Nomex™, or resin bonded fiberglass round and Nomex™ turn insulation in strip-on-edge windings, glass laminate pole washers, Dacron-glass-covered high-temperature enamel turn insulation in wire wound poles, and thermosetting bonding resins to provide insulation systems that have a thermal rating of at least Class 155 (F). If these materials are operated at Class 130 (B) temperatures, they should have a more than adequate thermal life. One problem is that many of the machines with this type of winding have brushless exciters and so even average winding temperatures are not monitored. The materials most susceptible to thermal degradation are organic bonding and backing materials, whereas inorganic components such as mica, glass, and asbestos are unaffected at the normal operating temperatures of electrical machines.

The thermal life of insulation at hot spots in windings is significantly reduced since the margin between operating temperature and thermal rating is much less. This effect is more critical in older Class 130 insulation systems and the presence of such hot spots is very difficult to detect.

The following are the most common causes of thermal aging in salient pole windings: – Overloading or high air temperatures leading to operating temperatures well above

design values. – Inadequate cooling, which can be general, e.g., insufficient cooling air or cooling water,

or local dead spots in the cooling circuit due to poor design, manufacturing, or maintenance procedures

– The use of materials that have inadequate thermal properties and consequently deteriorate at an unacceptable rate when operated within design temperature limits

– Overexcitation of rotor windings for long periods of time – Negative sequence currents due to system voltage imbalance, etc., which leads to

circulating currents on the rotor winding


Insulation aging from thermal cycling occurs mainly in synchronous motors and hydro-generators that are started and stopped frequently.

There are two heat sources within a rotor when a synchronous motor is started. One mainly applies to motors that are started directly-on-line, causing heating due to currents flowing in the pole tips of solid pole rotors, or the damper winding in those with laminated poles. The other is the I2R heat generated in the windings once excitation is applied. Frequent starts and stops cause winding expansion and contraction as a result of the presence or loss of these winding heat sources. Relative movement of winding and insulation due to the different coefficients of thermal expansion in the various components leads to insulation abrasion.

The thermal cycling resulting from frequent starts and stops leads to the cracking of the resin or varnish bonding the insulation system components together. This causes loosening and relative movement between these components, which increases looseness and abrasion. Also if the



Working Group Copy IEEE P56 Rev. 16 Draft windings are restrained from returning to their cold position, they may become distorted. Poor design or too-rapid or too-frequent load cycles for the design are the root causes.

7.4.3 Pollution (Tracking and Moisture Absorption) Salient pole rotor windings, especially strip-on-edge types, are generally susceptible to failure from contamination by conducting materials because they rely on adequate creepage distances between bare copper conductors to prevent shorts. Such problems are not confined to machines with open-type enclosures since oil leaking from bearings, moisture from condensation, leaking air coolers and dust from hydro-generator bakes can contaminate windings. Such problems can be avoided in wire-wound types by encapsulating the pole windings and connections to keep contaminants out.

When contaminants such as moisture, coal dust, and oil-dust mixtures cover the surfaces of salient pole windings they can produce conducting paths between turns and to ground. This can lead to turn-to-turn failures (especially in strip-on-edge types) and ground faults. Certain chemicals can also attack insulating materials to cause them to degrade.

Earlier insulation systems containing materials such as asbestos, cotton fibers, paper, etc. bonded by organic varnishes are much more susceptible to failure from moisture absorption.

7.4.4 Abrasive Particles

As with stator windings (Section 7.2.10b), rotor windings operated in environments containing abrasive dusts can also experience insulation failures from dust impingement.

Abrasive dust from the surrounding atmosphere carried into the interior of a motor or generator by cooling air will abrade the rotor winding insulation surfaces. This may eventually expose the conductors in multilayer wire-wound poles, resulting in turn shorts. Also, the ground insulation in both types of salient pole windings and their interconnections may be eroded to cause ground faults.

7.4.5 Rotational Force

Among the most common causes of failure in salient pole rotor windings are the continuous Rotational forces imposed on them by rotation and the cyclic Rotational forces induced by starting and stopping.

The radial and tangential Rotational forces imposed on rotor winding insulation system components tend to distort the coil conductors and inter-coil connections and crack the coil insulation if they are not adequately braced. If the pole winding bracing is inadequate or becomes loose, the resulting coil vibration and movement of the coils on the poles will cause abrasion of the conductor and ground insulation. Inadequate inter-pole bracing in large, high-speed machines will lead to coil distortion, whereas erosion from loose windings will occur mainly during starts and stops. Winding looseness can also lead to pole washer and inter-coil connection cracking from fatigue. Mechanical winding stresses will become excessive and cause serious winding damage if the rotor is made to run over speed. Inadvertent overspeed of the rotor



Working Group Copy IEEE P56 Rev. 16 Draft can result from slow response of the wicket gates in a hydro-generator after opening its breaker and faulty valves in pumps that allow a head of water to drive the pumpset at high speed in the reverse direction.

7.4.6 Repetitive Voltage Surges The normal DC voltage applied to rotor windings does not cause rotor insulation aging. Also, normal voltage levels in a rotor winding are usually so low that they will not induce insulation aging even in weakened materials. Hence, normal operating electrical stress is not an important cause of aging. However transient overvoltages induced by fault conditions on the stator side or faulty synchronization can cause aged rotor winding insulation to puncture.

High transient overvoltages may be induced into rotor windings by phase-to-phase stator winding short circuits, faulty synchronization, asynchronous operation, or static excitation systems (see Section 7.3.5). Such transient voltages, in conjunction with weak insulation or insulation that has been degraded by thermal or mechanical aging, can cause failures, which are predominantly turn-to-turn. These overvoltages are most severe in salient pole windings due to their design configuration.

7.5 Wound Rotor Winding Aging Mechanisms

On the basis of this standard’s the minimum machine rating of 35 KVA it is assumed that only three-phase bar-lap/wave-wound rotor windings need be considered. Operating voltages for this type of winding are generally limited to about 2000 V, or less so there is no need for the use of insulation materials with a high dielectric strength. As a result of the low operating voltage, continuous electrical aging is not a factor. Modern windings of this type typically have resin bonded glass fibre tape applied as conductor insulation and Nomex™, or Dacron/Mylar/Dacron slot liners. The following are electrical and mechanical failure mechanisms that are specific to wound-rotor windings. Other features of this type of winding are brazed joints to make connections between bars, endwinding banding to control mechanical stresses from Rotational force and sliprings to connect each phase to an external source for the control of rotor current.

7.5.1 Thermal Aging

The thermal aging and its effects in this type of winding is similar to that discussed under Section 7.3.1 for round rotor windings. This is so since the materials used are similar and both are subjected to significant mechanical stresses resulting from Rotational forces. Therefore no discussion on this topic is included under wound rotor windings.

7.5.2 Transient Overvoltages

In a wound-rotor induction motor, there is a transformer effect between the stator and rotor windings. Consequently, power-system surge voltages imposed on the stator winding will induce overvoltages in the rotor winding. This overvoltage may puncture the turn or ground insulation.

Providing there is adequate turn and ground insulation on the rotor winding, such voltages should not cause electrical aging; that is, partial discharge is unlikely. : Transients will, however,



Working Group Copy IEEE P56 Rev. 16 Draft accelerate the failure of insulation that is initially weak, or that has been degraded by thermal or mechanical aging.

7.5.3 Unbalanced Stator Voltages

Unbalanced stator winding power supply voltages will induce negative sequence voltages and currents in the rotor winding. These negative sequence currents increase rotor winding heating in all phases and, therefore, induce accelerated thermal aging of both the turn and ground insulation.

7.5.4 High-Resistance Connections

If a joint between two conductors has been poorly soldered or brazed, it will present a high resistance to the current flowing through it under load and this will produce overheating of the joint insulation. The excessive amount of heat produced by high-resistance bar-to-bar connections induces rapid thermal aging of the insulation around the connection and on adjacent connections until a turn-to-turn, phase-to-phase, or ground fault develops. In many cases, the heat generated is sufficient to melt the solder or brazing material in the joint. A secondary effect could be thermal damage and failure of the endwinding banding discussed in Section 7.5.5.

7.5.5 End-Winding Banding Failures

Application of banding over the rotor end-windings is required to brace them against the high Rotational forces imposed on them during operation. Up until the early 1950s, end-winding banding consisted of a number of turns of round steel wire applied tightly over an insulating layer, which was required to give mechanical and electrical separation from the conductor insulation. The round wires were bonded together with a low-melting-point solder. The development of pre-stressed resin-coated fiber material then prompted motor manufacturers to start using this material because of its superior mechanical and thermal capabilities, as well as its elasticity.

If the steel wire or resin-coated fiber materials fails from overheating, overstressing, or poor manufacture, the end-windings fly outward under the influence of Rotational forces and pieces break off. This results in a rotor winding ground fault and often, a consequential stator winding failure.

7.5.6 Slip Ring Insulation Shorting and Grounding

The three slip rings in a wound-rotor motor must be separated from the shaft by a layer of insulation applied between the two. The spacing between the rings must be sufficient to provide an adequate electrical creepage distance and sometimes barriers are used to increase this. Also, the two outer rings are usually connected to the winding leads via studs that pass through the other rings. These studs must be electrically isolated from the slip rings and this is normally done by fitting insulating tubes over them. The failure mechanisms in this section also apply to round rotor and salient pole machines with slip ring connections to their field windings.



Working Group Copy IEEE P56 Rev. 16 Draft If the slip ring enclosure is contaminated with oil, dust from brushes, moisture, or a combination of these, then shorting between rings and the shaft and/or between the rings can occur. If this happens, serious damage can occur to the shaft, the rotor windings, and the slip rings. Also, if the shaft or stud insulation fails due to thermal aging or mechanical stresses, these types of failures will also occur.

7.5.7 Pollution (Tracking and Moisture Absorption)

All windings are susceptible to aging and failure from this cause, especially if they are not well sealed. Even though the operating voltages of wound rotors are much lower than those of stator windings the absorption of moisture and surface contamination can lead to ground faults if the winding is not sealed. Cracked insulation, or impregnating resin is more likely to occur in this type of winding since it is subjected to high mechanical stresses.

7.6 DC Motor and Generator Field Windings

The construction of DC machine, series, shunt and inter-pole field windings is similar to that of the “multi-layer wire wound” salient pole rotor type discussed in Section 7.4. As such many of the aging mechanisms are similar. However, from the descriptions below it can be seen that there are some differences in the failure mechanisms.

7.6.1 Thermal Aging

All insulating and nonmetallic bracing materials deteriorate with time due to the heat from the windings. The rate at which component materials deteriorate is a function of their thermal properties and the temperatures to which they are subjected. If the thermal ratings of component materials have been properly selected, the thermal aging and associated deterioration will occur gradually over an acceptable service life.

In older Class 130 winding material such as asbestos backed mica splittings were used for pole ground insulation were bonded with organic varnishes such as shellac, asbestos board, or phenolic bonded glass fibers were used for pole washers. Also in early insulation systems organic materials were used to insulate conductors. These materials are more susceptible to shrinking and cracking under the influence of thermal aging.

Modern DC field pole winding designs typically use Nomex™, or resin bonded fiberglass ground and Nomex™ turn insulation in strip-on-edge windings, glass laminate pole washers, Dacron-glass-covered high-temperature enamel turn insulation and thermosetting bonding resins to provide insulation systems that have a thermal rating of at least Class 155 (F). If these materials are operated at Class 130 (B) temperatures, they should have a more than adequate thermal life. The materials most susceptible to thermal degradation are organic bonding and backing materials, whereas inorganic components such as mica, glass, and asbestos are unaffected at the normal operating temperatures of electrical machines.



Working Group Copy IEEE P56 Rev. 16 Draft The thermal life of insulation at hot spots in windings is significantly reduced since the margin between operating temperature and thermal rating is much less. This effect is more critical in older Class 130 insulation systems and the presence of such hot spots is very difficult to detect.

The following are the most common causes of thermal aging in salient pole windings: – Overloading or high air temperatures leading to operating temperatures well above

design values. – Inadequate cooling, which can be general, e.g., insufficient cooling air or cooling water,

or local dead spots in the cooling circuit due to poor design, manufacturing, or maintenance procedures

– The use of materials that have inadequate thermal properties and consequently deteriorate at an unacceptable rate when operated within design temperature limits

Weakening of both turn and ground insulation from this type of aging can cause winding failures from both ground and inter-turn shorts both of which will make the machine inoperable.


Insulation aging from thermal cycling occurs in motors and generators that are load cycled.

I2R heat generated in the windings causes the field winding conductors to expand. Frequent variations of the field current as a result of the variations in I2R losses cause expansion and contraction. As a result of different coefficients of thermal expansion in the various components, relative movement of winding and insulation occur.

Thermal cycling leads to the cracking of the resin or varnish bonding the insulation system components together. This causes loosening and relative movement between these components, which increases looseness and abrasion. Also if the windings are restrained from returning to their cold position, they may become distorted. Poor design or too rapid or too-frequent load cycles for the design are the root causes, as with thermal aging thermal cycling can lead to winding ground faults and inter-turn shorts.

7.6.3 Abrasive Particles

As with stator windings (Section 7.2.10b), field windings operated in environments containing abrasive dusts can also experience insulation failures from dust impingement.

Abrasive dust from the surrounding atmosphere carried into the interior of a motor or generator by cooling air will abrade the rotor winding insulation surfaces. This may eventually expose the conductors in the multi-layer, wire-wound poles, resulting in turn shorts. Also, the ground insulation in field pole windings and their interconnections may be eroded to cause ground faults.


All windings are susceptible to aging and failure from this cause, especially if they are not well sealed. Even though the operating voltages of dc machine field windings are much lower than



Working Group Copy IEEE P56 Rev. 16 Draft those of stator windings the absorption of moisture and surface contamination can lead to ground faults if the winding is not sealed.

7.7 DC Motor and Generator Armature Windings and Commutators

The following descriptions are based on the assumption that only bar-type armature windings are used for the size of machine covered by this standard. It should also be noted that this type of winding is similar to that for the wound rotor type discussed in Section 7.5 and so the many of the aging mechanisms will be the same.

7.7.1 Thermal Aging

The thermal aging and its effects in this type of winding is similar to that discussed under Section 7.3.1 for round rotor windings. This is so since the materials used are similar and both are subjected to significant mechanical stresses resulting from rotational forces. Therefore no discussion on this topic is included under dc machine armature windings.

7.7.2 High Resistance Connections

If a joint between two conductors, or between a conductor and the commutator riser has been poorly soldered or brazed, it will present a high resistance to the current flowing through it under load and this will produce overheating of the joint insulation. The excessive amount of heat produced by high-resistance bar-to-bar connections induces rapid thermal aging of the insulation around the connection and on adjacent connections until a turn-to-turn, phase-to-phase, or ground fault develops. In many cases, the heat generated is sufficient to melt the solder or brazing material in the joint. A secondary effect could be thermal damage and failure of the endwinding banding discussed in Section 7.7.3.

7.7.3 End-Winding Banding Failures

The causes of these are essentially the same as for wound rotor windings discussed in Section 7.5.5 so no detailed descriptions are provided in this section.


The aging and failure mechanisms relating to this are essentially the same as those discussed for wound rotors windings in Section 7.5.7 and so no detailed description is given in this section.

7.8 DC Motor and Generator Commutators

The following section covers common aging mechanisms found in DC machine commutators.

7.8.1 Glass Band Contamination

The glass band tape on glass-banded commutators can fail if contamination, such as carbon dust, gets underneath the band. This banding is usually protected against contamination by covering it with a material such as Viton™. Solvents should not be used to clean commutators because



Working Group Copy IEEE P56 Rev. 16 Draft contamination can be washed under the bands. The purpose of the glass band tape is to continually apply tension on the copper and insulated segment pack of the commutator during service to maintain stability at high speeds.

7.8.2 Electrical Tracking

A buildup of carbon dust behind the commutator risers on the steel shell causes problems with arcing. If carbon dust builds up on the commutator and the mica plate, which insulates the bars is bridged over, a flashover can occur.

7.8.3 Commutator Wear

A proper film on the commutator service is necessary in order to provide proper commutation. Temperature, atmosphere and brush grade affect the film and if the film is changed electrically or mechanically, commutator wear will be accelerated. Chemical contamination, abrasive dust, and oil vapor will wear away or change the commutator film to a nonconductive film eventually causing threading. Threading is circumferential grooves on the commutator caused by abrasive or an electro-chemical action of the brushes, light electrical loading, light brush pressure, or porous brushes.

Commutator bar surface etching caused by arcing between the brushes and commutator typically looks like pitting, eroded or burnt bars. If etching is not corrected, flat spots will develop on the commutator service. Flat spots can also be caused by vibration.

Another problem that can occur is “copper drag”. Copper is dragged over the trailing edges of the commutator bars, and it looks like small flakes or feathers. It is caused by a contaminated atmosphere, excessive vibration, low current density of the brush or the wrong brush grade, or copper imbedded into the brush.

As the commutator surface gradually wears, the undercut portion of the commutator (mica insulation between copper bars) will eventually protrude above the copper surface. If this is not corrected, excessive commutator wear and brush wear will develop.

If the problems with commutator wear are not corrected, eventually a flashover will occur.

7.8.4 Commutator Eccentricity

If the commutator runs off center, the brushes will ride up and down within their holders on every rotation. As the speed increases, the brushes loose contact with the surface causing burning on the commutator. This eccentricity can be caused by distortion due to wide temperature changes and high speed, a bent shaft, bearings that are not running true, an offset coupling, and if the commutator was machined on bad centers.

7.8.5 Commutator Brush Wear

The wear of commutator brushes may be accelerated by the rough surface of a commutator. The commutator with a runout greater than .003 inches is considered damaging to the brushes.



Working Group Copy IEEE P56 Rev. 16 Draft Silicone vapor or abrasive dust contaminants will accelerate brush wear. Excessive sparking and an incorrect bush grade will also cause accelerate brush wear. If the brush pressure is too light or too heavy, this can cause sparking.

(Reference: GE Industrial & Power Systems, DC Motor Wind Repair Course)

7.9 Stator Core Insulation Aging Mechanisms

This section discusses the most common causes of stator core insulation failures in both induction and synchronous machines. Core lamination insulation shorting and mechanical damage can occur from a variety of aging and failure mechanisms that can be thermal, electrical, mechanical, design, or manufacturing related. Stator cores used in large turbine generators, hydro-generators, and motors, which have segmented stator cores, are most susceptible to failures from these causes. Some of these failure mechanisms will only occur in specific types of machines, whereas others are applicable to all types.

7.9.1 Thermal Aging

Degradation of the core condition due to the effects of thermal aging can occur in all rotating machine laminated cores. Core overheating will cause accelerated aging of the core insulation if its thermal rating is exceeded for an extended period of time. This is most likely to occur if an organic varnish is used since this will dry out due to loss of solvents by evaporation of low-molecular-weight components. Once this occurs, the varnish becomes brittle, cracks, and eventually breaks down. As a consequence, interlamination shorts will develop, eddy currents will increase, and this will eventually lead to core melting due to even higher temperature operation. In large hydrogen-cooled turbo-generators, condition monitors may be installed to provide early detection of core insulation overheating. These monitors detect the presence of materials driven off the core by excessive heating in the hydrogen cooling gas.

(a) General Core Overheating

The most common causes of are:

– Loss of cooling water for hydrogen or air-to-water coolers in totally enclosed machines. – High ambient air temperatures for open-ventilated, air-cooled machines – Blockage of air inlets in open-ventilated air-cooled machines due to pollution or debris. – Complete or partial blockage of cooling-air passageways due to the accumulation of oil,

dirt, etc. – Turbine generator operation at reduced hydrogen pressure

General overheating of cores in large machines, especially turbine generators, may also cause core slackness that results from thermal expansion of the core steel. When core temperatures exceed design limits, the radial and axial forces exerted on the core support structure may be high enough to cause permanent deformation of these components. As a consequence, the core will loosen when the condition causing the overheating is removed. This core looseness will lead



Working Group Copy IEEE P56 Rev. 16 Draft to insulation failure from abrasion. Core looseness has sometimes occurred as a result of extended full-flux core testing. General thermal aging of core insulation also causes loosening of the core due to shrinkage and weight loss of the organic components.

(b) Local Core Overheating.

The most common causes of this are:

– Inadequate cooling of certain areas of the core due to poor design or blockage by debris

(e.g., cooling air or hydrogen flow is too low or nonexistent in these areas) – Manufacturing errors (e.g., some cooling-medium, passages have been blocked due to

missing holes or cut-outs in the core support structure)

Overheating can, of course, also result from core insulation degradation initiated by electrical or mechanical aging mechanisms, as described in Sections 7.8.2 and 7.8.3 below.

7.9.2 Electrical Aging

Electrical aging occurs when the voltage across the lamination insulation induced by magnetic fluxes, electromagnetic forces, or high ground-fault currents causes deterioration. Although DC and transient voltages may cause aging, it is normally AC-voltage-induced effects that cause the most severe damage. Degradation from electrical aging can occur in all types of stator and rotor laminated cores.

The root causes of electrically induced core insulation degradation can be subdivided into the categories of overheating due to over- or underexcitation, winding ground faults, and stator-to-rotor rubs due to unbalanced magnetic pull effects. These are discussed separately below:

(a) Stator Core End Overheating Due to Underexcitation.

The main air gap flux in synchronous machines is in the radial direction. This flux is responsible for generating the voltage in the stator winding. In addition, synchronous machines have significant leakage fluxes in the end region, especially when the rotor winding is under excited. These fringing fields are produced by currents in the stator and rotor end-windings and by the discontinuities at the stator and rotor core surfaces. The axial component of this field generates circulating currents within the segments of the end region stator laminations, generating some additional electrical losses and, thus, heat. The eddy currents due to the axial magnetic field cause stray losses in the end regions. The axial magnetic field is sensitive to changes in load and power factor. During leading power factor (under excitation) operation, this field can be quite high in large machines, especially if they have direct-water-cooled windings. This can degrade the interlaminar insulation as follows:

– Higher temperatures occur, which may reduce the dielectric strength of the interlaminar

insulation over time and also give rise to other stresses due to expansion and relative motion between components. Modern inorganic core insulation such as aluminums



Working Group Copy IEEE P56 Rev. 16 Draft orthophosphate are capable of withstanding temperatures as high as 500°C. However, in some machines, especially if they were built before 1970, this limit is much lower. However, the long-term effect of operating near the core insulation rated temperature is that this could reduce its life.

– The circulating currents in the laminations can result in relatively high voltages being developed between adjacent core laminations. Under extreme conditions, this voltage may be an order of magnitude higher than normal. It has been shown that minor defects in the interlaminar insulation may provide a path for circulating currents, causing further, perhaps serious, local deterioration.

The combined effect of the above two mechanisms in conjunction with other existing stresses can damage interlaminar insulation in the stator core end regions near the bore. This increases circulating currents between laminations, causing temperature rise, local weakening, and tooth vibration and breakage from high cycle fatigue. Some manufacturers install thermocouples in this area of the core, during manufacture, to monitor for insulation degradation from such aging mechanisms. A rising trend or a sudden increase in temperature recorded by these sensors can provide an early warning of the problem.

(b) Overheating of Back-of-Stator Core Due to Overexcitation.

In order to keep the physical size of large two- and four-pole synchronous machines within reasonable limits, it is necessary to excite the stator core at a fairly high magnetic flux density. Laminated steel, as well as the lamination insulation, is, therefore, selected to avoid high core losses. The lamination insulation is selected for its low dielectric permitivity and good insulation properties under high stress. High temperatures due to increased core loss can result from overexciting the field winding, thus producing higher than normal magnetic flux densities

The large volume of core behind the slot is more prone to overheating due to increased flux compared to the tooth area which, in high-speed machines, operates at much lower flux densities. The core behind the slot has relatively less ventilation than the teeth. Consequently, the core losses can quickly raise the temperature of the back of core and the temperature increase is particularly steep as the iron begins to saturate. Once the temperature has been elevated, the chances lamination insulation breakdown, are increased. Such a breakdown would give rise to interlaminar shorts and increased eddy currents, which can cause even higher temperatures. The higher temperature can also cause mechanical stresses, resulting in distortion and vibration. When combined, these effects can eventually lead to fusing of laminations, melting of iron, and core failure.

(c) Stator Winding Ground Faults in Core Slots.

The energy and heat produced by stator ground faults in or just outside the slot region are often sufficient to melt and fuse the core laminations at the core surface. If this core damage is not repaired when the failed coil or bar is replaced or electrically cut out and bypassed, the new coil and adjacent coils could also fail to ground as a result of the heat generated by the shorted laminations. It is, therefore, important to perform tests to check the condition of core insulation



Working Group Copy IEEE P56 Rev. 16 Draft in the vicinity of ground fault damage before installing a new bar or coil or making a temporary repair on the winding.

(d) Stator Core Faults from Through-Bolt Insulation Damage.

In some medium to large motor and generator designs, the stator core pressure is maintained by bolts that pass though axial holes in the stator core laminations and endplates and have nuts, steel washers and insulating washers installed on either end. Core pressure is maintained by keeping these nuts tight. These through bolts have to be electrically isolated from the core with tube insulation to prevent core insulation shorting. If retaining nuts become loose or the bolts stretch, the bolt insulation can fail from insulation abrasion resulting from core lamination movement. If this happens, core lamination shorting and core burning may occur.

7.9.3 Mechanical Aging

The most common causes of mechanical degradation in cores are inadequate core pressure applied in manufacture, core pressure reduction in service due to relaxation of the core support structure, core vibration, back-of-core looseness, and mechanical damage causing smearing of the core surface at the bore. Degradation due to core looseness is predominantly found in large generators and motors with a segmented core construction. Core insulation damage due to vibration is most commonly found in large two-pole turbine generators. Mechanical damage to the core bore due to foreign body impact can occur in any type of machine.

a) Core Looseness

When the laminations in a stator core become loose they can move relative to one another under the influence of mechanical vibration and/or electromagnetic forces, and the insulation on them degrades due to abrasion. If not detected in time, all of the lamination insulation in the areas of core looseness is removed and lamination shorting occurs. The eddy currents that flow as a result of this shorting create excessive heat that can eventually lead to core melting and lamination fracture. Core looseness can also result from vibration caused by the natural frequency of the core and frame being too close to the main twice-power frequency (100 or 120 Hz), electromagnetic core excitation frequency that occurs in all AC machines.

The stator cores in large machines with segmented laminations are built on axial keybars (two per lamination) welded to the frame. A dovetail fit between each core lamination segment and the keybar provides radial support for the core. If the fit between the core laminations and keybars is, or becomes loose, then arcing between the two and shorting of the core laminations at the back of the stator core will occur.

b) Stator Core Relaxation, Fretting, and Failure- Turbine Generators.

Core design and manufacturing problems contribute to this type of core deterioration. The following gives some general information, which may not be applicable to particular situations.



Working Group Copy IEEE P56 Rev. 16 Draft – Excessive use of resilient materials during manufacture may contribute to relaxation

during service. – Core pressure is another important factor to consider during manufacture. This becomes

more critical as the length of the core increases with the rating. If the core support structure relaxes in service, then the core laminations become loose. The most common location of such looseness is at the stator bore since this is farthest from where the core pressure is applied.

If the laminations at the core bore are loose at the end of the core, the following sequence of degradation occurs if this problem is not detected and quickly corrected:

– The core insulation is abraded due to lamination relative movement under the influence

of axial electromagnetic forces from end leakage fluxes. – Core lamination shorting and overheating starts to occur when the core insulation is re-

moved by abrasion. – Eventually, pieces of lamination teeth will break off due to fatigue failures, and vent

spacers may also break free. Such debris can cause core insulation damage in other locations.

(b) Stator Core Vibration-Turbine Generators.

Some of the causes of high stator vibration in service are: – Inadequate support of the core in the stator frame, creating an assembly resonant

frequency close to twice the power supply frequency (100 or 120 Hz) – Unbalanced phase loading – Inadequate stator end-winding support (causing vibrations that are reflected back to the

core).

Even without these factors, a certain amount of vibration exists due to the 100 or 120 Hz "ovalizing" force caused by the magnetic field. The displacement of a loose core caused by these forces may result in relative motion between laminations and fretting of the lamination insulation to the point of breakdown.

(c) Stator Core Fretting, Relaxation, and Failure-Hydro-Generators.

Since a hydro-generator core has a large diameter and short depth behind the winding slot, it is relatively flexible and the frame is the main support. The magnetic forces between the rotor poles and stator core will, therefore, tend to produce displacements in the core.

In general, the displacements and resulting vibration are small in multi-pole generators. The traveling wave produced by the magnetic force has a number of nodes equal to twice the number of poles. This results in smaller displacement. The only exception is in the case of fractional slot windings (that is, windings having a non-integer number of slots per pole per phase), where larger displacement is possible due to wavelengths that can be longer than the pole pitch. These mechanisms can cause high temperature and melting anywhere in the core.



Working Group Copy IEEE P56 Rev. 16 Draft However, the bore section is particularly susceptible. It is physically weaker than the section behind the slot and carries a higher flux density. As these mechanisms weaken the bond between the laminations, the teeth are likely to chatter and break off. Secondary damage to the stator and the rotor can occur if the debris finds its way into the air gap.

(d) Back-of-Stator Core Overheating and Burning.

Overheating and burning of the back of a stator core can be caused by a loose connection between the core laminations and the stator frame. Axial key bars welded to the frame are used for piling the core laminations. A dovetail fit between each core plate segment and the key bar provides radial support for the core. To allow assembly or stacking on the key bars, a clearance is necessary in the “dovetail” fit. If excessive, this clearance will permit relative motion, intermittent contact, circulating currents and overheating during operation.

Leakage flux at the back of the core induces currents in the key bars. These currents flow to ground through the stator frame without causing any harm provided there is no path for them to flow through the core. This is the case if the punchings are in good contact (positive grounding) with the key bars or if they are completely isolated electrically (insulated keybars). However, should the clearance become excessive at some point, any core vibration could cause intermittent contact at that point. This can cause cracking of key bar insulation, if used, or core-to-key bar arcing if un-insulated key bars are used. In either case arcing between cores and key bars can result leading to overheating, and core melting in the local area. Small amounts of melting are difficult to detect due to inaccessibility and the impracticability of monitoring a large area with thermocouples. An early indication of this problem is an upward trend in frame vibration due to the increase in core-to-key bar clearances.

Should a number of intermittent contacts develop on the key bars, the possibility of current transfer between key bars would increase. The currents would begin to circulate through the low-resistance path offered by the laminations. The resultant overheating could escalate into failure of interlaminar insulation and an increase in circulating currents and temperature, leading to fusing of laminations, melting, and, ultimately, core failure due to widespread melting of the laminations.

This problem can be greatly reduced by interconnecting all the key bars at each end of the core by means of welded copper straps to form a ring for carrying the circulating currents. This is better done during manufacture since retrofitting can be a major task.

(e) Stator-to-Rotor Rubs Due to Bearing Failures and Unbalanced Magnetic Pull.

These types of failures mostly occur in induction motors, which have relatively small air gaps, or salient pole synchronous motor or generators. Bearing failures or excessive unbalanced magnetic pull due to poor air gap eccentricity or design can allow the rotor to move toward the stator and rub on its core bore surface. If this occurs, smearing of both the stator core at its bore and the rotor core outside diameter will occur. In addition, the rotor-to-stator rubbing may generate sufficient heat to cause rapid thermal aging of the stator winding insulation and a consequent ground fault. Such rubs will cause shorting of the core insulation on both the stator and rotor.



Working Group Copy IEEE P56 Rev. 16 Draft (f) Loose Metal Components Entering the Air Gap.

Bolts and other metallic components that break free from machine internal components can enter the air gap. Such objects are "sucked" into the air gap by cooling air or gas flow and magnetic attraction, if they are made from magnetic steel. If this occurs, stator core gouging or smearing and consequent insulation surface shorting will occur. Since synchronous machines have much larger air gaps than induction types, they are more susceptible to such damage.

[B] [R. Bartnikas and R. Morin, "Analysis of multi-stress accelerated aged stator bars using a three phase test arrangement", IEEE Trans. on Energy Conversion, Vol.21, pp.162 - 171, 2006.]

References

1. Greg C. Stone, Edward A. Boulter, Ian Culbert, Husssein Dhirani: Electrical Insulation for

Rotating Machines – Design, Evaluation, Aging and Repair, IEEE Press Series on Power Engineering published by Wiley Interscience.

2. I M. Culbert, H. Dhirani, G.C. Stone: EPRI Power Plant Electrical Series, Volume 16 – Handbook to Assess the Insulation Condition of Large Rotating Machines.

8. Visual Inspection Methods [Nancy Frost – Clause head]

[Need to Revise to include what to look for and remove aging or why and put above.]

Visual inspections of stator windings are usually made at convenient intervals in the range of 1 to 5 years. Machine availability and maintenance history should be considered when selecting this frequency of inspections. Depending on the machine design and physical size a limited visual inspection can be conducted with the rotor in place. Stator bar looseness in the endwindings or at the core edge may be visible. Step iron deterioration may also be detected and the general level of winding contamination can be determined with a rotor in place. The bushing box area should also be inspected to evaluate the circuit rings and tightness of the endwinding structure.

Robotic inspection techniques are available from several sources. These can provide more details along the air-gap than visual inspection from the stator ends. Both stator and rotor surface can be inspected and a visual record is produce for future reference.

A rotor out inspection should be considered every 5 to 10 years based on the results of past limited inspections and operating information. A rotor out inspection permits detailed rotor and stator condition assessment. Some testing such corona probe and core iron testing must be performed with the rotor or several poles of a salient pole machine removed. Major stator repairs such as rewedging can only be performed with the rotor removed. This is a major activity and should not be scheduled with out good reason.



Working Group Copy IEEE P56 Rev. 16 Draft To achieve maximum effectiveness, a visual inspection program should be directed initially to those areas that have been shown by previous experience to be most prone to the forms of damage or degradation caused by the influences listed in this guide.

A suggested condition appraisal, summarizing areas suspect for deterioration or damage, is shown in Annex B.

8.1 Visual Inspection Safety

Generator inspections must be conducted with the proper safety precautions. Some of the requirements may be site specific since a generator is often designated a “confined space”. The machine stator windings must be grounded with a visible open in each phase circuit to prevent energizing the machine during an inspection. Spool pieces in the co2 line and hydrogen line are to be removed to prevent gas from entering during the inspection. Oxygen levels are to be monitored during the inspection. The proper turbine generator or penstock clearances must be signed prior to the generator inspection.

8.2 Armature Winding

[Nancy Frost volunteered to revise this section]

8.2.1 Thermal Aging

Examination of coils may reveal general puffiness, swelling into ventilation ducts, or a lack of firmness of the insulation. These symptoms suggest a loss of bond and separation between insulation layers or between the insulation and the conductors (The winding insulation will have a hollow sound when tapped.). The insulation may also be brittle.

Shrinkage of the insulation may lead to loosening of the coil in the slot or of the bracing. The consequent vibration can result in ground wall abrasion and loss of coil semiconducting slot stress control.

8.2.2 Cracking

Cracking of the insulation or in the surface paint may result from prolonged or abnormal mechanical stresses. A common cause of cracking in armature windings is looseness of the bracing structure.

8.2.3 Girth Cracking

Circumferential cracking of the ground wall identifies girth cracking. Girth cracking can occur on asphaltic windings, particularly in machines with core lengths greater than about 12 ft (4 m). Particular attention should be paid to the areas immediately adjacent to the ends of the slots. Where considerable cracking is observed, it is recommended that the wedges at the ends of the slot be removed for inspection, as dangerous cracks may also have occurred just within the slots.



Working Group Copy IEEE P56 Rev. 16 Draft 8.2.4 Contamination

Surface contamination adversely affects insulation strength. The most common contaminants are carbon and other conducting substances, oil, and moisture.

Particularly damaging are magnetic particles (iron termites) that vibrate with the effects of the magnetic field in the machine.

8.2.5 Carbon Deposits

Carbon accumulation over insulation surfaces can provide paths for leakage currents. For example, the risers (the connection straps between commutator bars and coils on dc armature winding) may collect carbon deposits that can initiate electrical tracking, with resultant burning, and subsequent failure.

8.2.6 Abrasion

Insulation surfaces may be damaged by contact with abrasive substances. Abrasion-resistant coatings are often used to extend the life of windings operating in abrasive environments.

8.2.7 Loose Slot Wedges or Slot Fillers

This condition may result in abrasion of the insulation. It can also reduce the effectiveness of coil bracing against short circuit and other abnormal mechanical forces. The semiconducting stress control system can be damaged, resulting in slot discharge.

8.2.8 Erosion

Foreign substances impinging against insulation surfaces may cause erosion. Even the high-velocity cooling air in salient pole machines over time can wear down the winding protective paint coating.

8.2.9 Corrosion / Chemical Attack

Corrosive atmospheric conditions commonly found in chemical plants, rubber mills and paper manufacturing facilities can chemically attack some insulation materials. Solvents can swell and reduce the mechanical properties of insulating materials.

8.2.10 Corona

Insulation of higher-voltage rated windings can be deteriorated by corona discharges in the slot section and in the end windings. Corona discharge is evident by white, gray, or red deposits in areas where the insulation is subject to high electrical stresses. Some experience is required to distinguish these effects from powdering, which can occur as a result of movement between surfaces such as in loose end-winding structures.

8.2.11 Rotational Forces

The effects of over speed may be observed on dc armatures by distortion of the windings, commutator risers, looseness or cracking of the banding, or movement of slot wedges.



Working Group Copy IEEE P56 Rev. 16 Draft 8.2.12 Commutator Condition

Commutators should be checked for uneven discoloration, which can result from short-circuiting due to breakdown of insulation between bars. They should also be checked for pinholes and burrs caused by flashover.

Commutator groove bands should be carefully inspected for (1) dryness or darkening, which may be an indication of loss in strength due to over-temperature, and (2) circumferential cracks within the band or at the groove walls, which may admit conductive contaminants. Separation of the band at the groove wall may be indicative of internal damage.

Restraining bands ("string bands"), which secure the exposed surface of the commutator cones, should be inspected for separation from the segment surface. Separation at this point is an area for the entry of contaminants.

The area immediately behind the commutator can also be a repository for carbon.

8.3 Field Windings

[Chuck Wilson revised this section]

In addition to insulation degradation from causes similar to those listed in 9.1, close attention should be directed to the following in field windings:

8.3.1 Coil Distortion

Distortion of field coils may be caused by abnormal mechanical, electrical, or thermal forces. Such distortion may cause looseness which can lead to failure of turn or ground insulation.

8.3.2 Loose Coils or Poles

Shrinkage and looseness of field coil washers or supports permits coil/pole movement during periods of acceleration and deceleration with the probability of abrading ground or turn insulation and breaking or loosening of connections between coils. This is a slightly more intense issue on horizontal units.

8.3.3 Rotor Coil Tightness

In cylindrical rotors (defined as "round-rotor" in IEEE Std 100-1988 [7]), evidence of heating of wedges at their contact with the retaining ring body, and "half-mooning" or cracks on the retaining rings, can be caused by high circulating currents. These currents may be due to unbalanced operation, excessive loads, or sustained single-phase faults close to the generator, such as in the leads or generator bus.

The condition and tightness of end-winding blocking, signs of deterioration or movement of the retaining ring insulating liner due to the above effects, and any other looseness should be noted.

Powdered insulation on surfaces or in air ducts is evidence of coil movement. Red oxide at metallic joints is evidence of fretting (relative movement of metal parts).



Working Group Copy IEEE P56 Rev. 16 Draft The integrity of field lead connections and condition of collector and collector-lead insulation should be checked on a regular basis, since failure of a joint during operation will lead to very serious consequences.

8.3.4 Brush Rigging

Insulation supporting the brush rigging should be checked for evidence of flashover or carbonized leakage paths and cleaned on a regular basis. Spring pressures and brush alignment should be inspected.

8.4 Core and Frame Assembly

The following items are considered to be the most significant in inspecting the core and frame assembly.

8.4.1 Stator (Armature) Core

A close examination should be made at the core surface for evidence of damage. Failure of interlaminar insulation may occur and is usually precipitated by external causes. Among these causes are operation in the far under excited region (round rotor machine), over-excitation, mechanical damage due to foreign objects, vibration, excessive heating due to power arcs created by winding failure, and excessive losses in the finger plates of large machines. Failure may also occur from re-wind processing when excessive heat, such as burnout, is applied for coil removal at the service shop to cores with laminations insulated with organic coatings. Such damage can initiate winding insulation faults and equipment failure. Careful inspection of the core condition is therefore mandatory whenever the machine in question is out of service for maintenance purposes. If distress is observed, a loop test is recommended. The loop test is described in Annex A.

Inspect for looseness of core laminations. Loose core laminations at the air-gap side of the core (teeth), especially at core ends, will vibrate, abrade interlaminar insulation (and ground insulation), short circuit laminations, and cause heating. Also, vibrating laminations may fatigue, crack, break off, and contaminate the machine with iron particles. Iron oxide powder deposits (evidence of fretting) are an indication of loose core iron or loose wedges.

Inspect ventilation ducts for loose or broken ventilation duct separators (fingers). These can cause core looseness or they can break off, resulting in mechanical damage to coil insulation and to the core interlaminar insulation.

Overheating of the end finger plates is evident by discoloration of the paint or components in the areas affected. Abnormal overheating can lead to thermal degradation of the interlaminar insulation.

End flux shield overheating (when present) is evidenced by discoloration of the paint or components in the areas affected. When these shields are insulated, abnormal heating can lead to thermal degradation of the insulation.

8.4.2 Core Insulated Through-Bolts

Where these are fitted, examination of the insulated washer and associated pieces is indicated together with verification of tightness and locking of the nuts. The insulation resistance of through bolts to the core should also be checked.



Working Group Copy IEEE P56 Rev. 16 Draft 8.4.3 Bearing, Hydrogen-Seal, and Other Insulation

Whenever bearings and other mechanical parts are disassembled, inspect their insulation for signs of deterioration. Pitting in the bearing material may be evidence of bearing-insulation failure. Refer to IEEE Std 115-1965, Test Procedure for Synchronous Machines, for electrical test procedure.

9. Insulation Maintenance Testing [Richard – Clause head]

9.1 Principles of Maintenance Testing

A list of electrical tests designed to detect particular areas of weakness is included in this maintenance guide. It should be noted that all tests are not applicable to all machines. The tests listed below have been used generally either to establish the long-time trends in parts of the insulation structure, or to detect specific types of flaws that may develop in portions of the insulation. With many maintenance tests, the trends measured over a period of years are normally more important than absolute measured values determined at a specific inspection period. A sudden change in the values for a given machine should be investigated and the cause determined. Some electrical tests may be potentially damaging to the insulation. The risks with such tests should also be recognized. For example, when a winding is tested that has been providing good service, but has a fault near ground or neutral connections, the fault may be worsened to a potentially nonfunctional condition. Where there is uncertainty about insulation condition, it is recommended that the manufacturer be consulted or if the manufacturer is no longer available, an insulation specialist should be consulted.

Insulation maintenance tests in this guide has been grouped as follows:

– Tests conducted on the field – Tests conducted on the armature

Their classification and inclusion in this manner is for convenience and the selection of maintenance tests will depend on the user's own philosophy, performance records, production, and economics. The user is encouraged to discuss findings with the manufacturer or insulataion specialist for interpretation and trends.

The tests are given in synopsis form in this guide; however, further details are provided in Annex A and in the appropriate IEEE standards to which reference is made.

9.2 Tests Conducted On The Field (Rotor)

All field windings have two types of insulation, ground insulation and turn insulation. The insulation resistance of the ground insulation can be measured according to IEEE Std 43-2000. In addition the suitability for service can be determined using an over potential test. The insulation resistance or condition of the turn insulation can be determined by the tests described in the following sections.

There are times when continued operation of a machine in a fault mode may be essential to avoid an outage. In generators, shorted turns of a minor nature, unlike shorted turns in the stator, may



Working Group Copy IEEE P56 Rev. 16 Draft not necessarily require immediate reinsulation. Although not recommended, rotors have been known to operate for years with a few random short circuits between successive turns in the rotor winding. However, should subsequent periodic impedance testing show the shorting to be progressive in nature, reinsulation will likely be necessary to assure reliable operation. Continued operation of large salient pole motors known to have shorted turns is not recommended, since further damage can occur due to high currents in the shorted turns, particularly during start operations.

9.2.1 Insulation Resistance. (numbering?)

9.2.2 Winding Resistance.

A reduction in resistance of a field winding or coil may indicate shorting of conductors possibly caused by a deterioration of the insulation between them. For complete windings or coils with many turns this method may not be accurate enough to detect shorted turns.

The rotor winding should be at room temperature before the cold resistance measurement is made, and the temperature of the winding carefully determined. For synchronous machines, it is necessary that field resistance and the corresponding temperatures be accurately measured since the temperature rise of the field winding during operation is commonly determined from the change in resistance. The relevance of this paragraph to insulation resistance escapes me.

In measuring the rotor resistance by the voltage-drop method, it is essential that voltage contacts for the voltmeter be placed directly on the collector rings or exposed leads of the rotor winding. The relevance of this paragraph to insulation resistance escapes me.

9.2.3 Field Winding Voltage Drop Test.

This test, commonly known as a "voltage drop" test, is sometimes made in the factory and can also be used as a maintenance test. The field winding is energized with low-potential alternating voltage (such as 120 V or at least the equivalent of 10 volts per pole) at conventional power frequency. With the field coils connected in series, similar coils should have a comparable voltage drop. (On salient pole rotors with springs behind the coils, an ac test may overheat the springs if the test continues for more than 10 minutes.)?????????

This test is suitable for cylindrical rotor windings of turbine generators and for windings of multiple fields of other machines. When a coil with short-circuited turns has been discovered, the test may be expanded to measure the voltage drops across individual coil turns.

If the rotor has brushless excitation, the manufacturer's instructions should be reviewed carefully before making impedance tests.

9.2.4 Impedance Test

The presence of short-circuited turns in the windings of cylindrical rotors of turbine-generators or individual field coils of salient-pole generators or motors may be detected by impedance measurements on individual coils These measurements are usually obtained by applying a known current or voltage across the coil. (A power frequency source is acceptable but higher frequency sources are preferred.) Other parameters such as watts, power factor VA are then



Working Group Copy IEEE P56 Rev. 16 Draft measured Similar coils should have a comparable impedance. A coil with a shorted turn will have substantially different values for watts, impedance and power factor

The overall ohmic value of winding impedance obtained from the impedance test is useful if an initial reading, with no short-circuited turns, is available for comparison. When ohmic values are used for comparison purposes, test results should have been obtained at approximately the same frequency and voltage or current (depending which is the dependent and independent quantity) for the two tests being compared.

In operation, the first signs of short-circuited rotor turns may be increased rotor vibration or increases in excitation requirements.

The effects of short-circuited turns on rotor vibration may be due to electromagnetic or thermal influences. Electromagnetic effects would inherently be more pronounced on rotors with four or more poles. Removal of excitation will often indicate whether the effects are electromagnetic, thermal, or both. If short-circuited turns cause thermal unbalance, the vibration will vary with temperature and hence will lag any increase in excitation by the length of time required for heating to occur. If variations from the cold to the hot condition are not too great, weight adjustments can sometimes be made to keep the vibration amplitude entirely within a satisfactory range for all temperatures. Otherwise, either thermal balancing or re-insulation of the short-circuited turns is necessary.

If the primary effect of short-circuited rotor turns is an increase in excitation requirements, re-insulating would be dependent on the ability to supply sufficient field excitation, under normal reactive load conditions, without exceeding exciter or rotor recommended operating temperature limits.

Experience on generators has shown that short-circuited rotor turns are not usually progressive in nature. Due to increased excitation current requirements, the average rotor temperature will likely increase.Changes in excitation requirements may be detected by comparison of a recent no load saturation curve with the original curve. If the rotor has a temperature recorder, the chart should be examined for indications of a sudden drop in rotor resistance at the time vibration appeared.

9.2.5 Flux Distribution Tests

9.2.6 .

In addition to the impedance measurements referred to in Error! Reference source not found.of this guide, several other tests are available by which short circuits between turns of cylindrical rotors can often be detected. Among these are the following:

The flux distribution over the rotor body surface, when a potential of 120 V at 60 Hz is applied to the collector rings, is observed by a test coil connected to a galvanometer, The test coil is arranged to span adjacent rotor teeth and is usually placed on the part of the rotor body close to the coil retaining rings. The magnitude and sign of the voltage induced in the coil for each pair



Working Group Copy IEEE P56 Rev. 16 Draft of teeth is plotted. The flux pattern shows a significant change in magnitude and sign whenever a short circuit exists in the slot being tested.

In another test method, an ac voltage is applied to the collector rings and the induced voltage in the test coil is read on a voltmeter. An iron-cored coil is used on the rotor body or an air-cored coil can be used in the end windings.

In another method a probe is permanently affixed to the stator core and it measures the flux during operation. By measuring the flux at different watt and var points the coil containing the shorted turn can be identified.

9.3 Tests Conducted On The Armature (Stator)

9.3.1 Insulation Resistance Test at Low Voltage.

Insulation resistance test is based on determining the current through the insulation and across the surface when a direct voltage is applied. The current is dependent on the voltage and time of application, the area and thickness of the insulation, and on temperature and humidity conditions during the test.

This test is usually made on all or parts of an armature to ground. The test can be performed with or without a guard electrode. The test primarily indicates the degree of contamination of the insulating surfaces or solid insulation and will not usually reveal complete or uncontaminated ruptures.

IEEE Std 43-2000 [2] outlines a recommended practice for insulation resistance testing and the corrections to be made for temperature and humidity conditions. IEEE Std 43-2000 also provides recommended values for minimum insulation resistance for safe operation.

The insulation resistance test is used to determine the insulation condition prior to application of an over voltage test.

9.3.2 Dielectric Absorption Test.

Dielectric absorption testing is the determination of insulation resistance as a function of time. This test like the insulation resistance test is made on all or parts of an armature circuit to ground.

IEEE Std 43-2000 [2] outlines the test procedures and equipment used for the standard method of performing the test, which is usually made at a test potential of 500 to 5000 V dc. Tests at higher potentials are commonly made using the voltmeter-ammeter method of resistance determination as outlined in IEEE Std 95-2002.

During this test the potential is held until the insulation resistance stabilizes or for a period of 10 min. The slope of the time resistance characteristic gives information on the relative condition of the insulation with respect to moisture and other contaminants. The ratio of the 10 min value to the 1 min value of insulation resistance is termed the "polarization-index," or “PI”. This “PI” value is useful in comparing the results of previous tests on the same machine.

Further details of these tests and suggested acceptable polarization indices for certain insulation systems are contained in Annex A.



Working Group Copy IEEE P56 Rev. 16 Draft 9.3.3 Over Voltage Tests

Over voltage tests, also referred to as high-potential or hi-pot tests, are used to assure minimum dielectric strength of the insulation. Such tests are made on all or parts of the circuit-to-ground insulation of the armature or field winding.

Many users of large rotating machines apply over voltage tests periodically and generally at the beginning of a machine overhaul or the overhaul of related equipment. This allows the detection and possible repair of insulation weaknesses during the scheduled outage.

Over voltage tests should be applied wherever possible to each phase in sequence, the remaining two phases not under test being solidly grounded. In this way, the insulation between phases is also tested. This is only practical, however, where both ends of each phase are brought out to separate terminals, as is usually the case in generators. Except for the larger horsepower ratings, most motors have either three or four leads brought out, precluding a test between phases.

The level of over potential that should be applied will depend to a very large extent on the type of machine involved, the degree of exposure to over voltages, and the level of serviceability required from the machine in question. Such tests should, however, be sufficiently searching to discern any weakness or incipient weakness in the insulation structure that might lead to service failure. The test voltage should not however, be so high as to cause an unnecessary breakdown and the user should be aware that over voltage test can be destructive.

Over voltage tests may be performed either by alternating or direct voltage methods. The values of test voltages usually are selected as follows: – For 60 Hz tests, the over voltage may be related to the rated machine voltage, and tests in

the range of 125 to 150% of the line-to-line voltage are normal. For recommended test levels refer to …….. Over voltage tests are typically conducted for 60 s. For test procedures, refer to IEEE Std 4-1978 [1].

– Equipment for making over voltage tests at very low frequency (0.1 Hz) has become commercially available. Such equipment is typically less in cost and weight and smaller in size than the equivalent 60 Hz equipment. For additional information, see IEEE Std 433-1974 [8].

– For dc tests, the recommended test voltage is a function of the rated machine voltage multiplied by a factor to represent the ratio between direct (test) voltage and alternating (rms) voltage. The recommended value is from 125 to 150% of the rated line-to-line voltage multiplied by 1.7. For test procedures, refer to IEEE Std 95-2002 [6].

It should be recognized that if the windings are clean and dry, over voltage tests may not detect defects that are in the end turns or in leads remote from the stator core.

9.3.4 Controlled Over Voltage Test (DC).

A controlled over voltage test is one in which the increase of applied direct voltage is controlled, and measured currents are continuously observed for abnormalities, with the intention of stopping the test before breakdown occurs. This test is often referred to as a "direct-current leakage test" or a "step voltage test."



Working Group Copy IEEE P56 Rev. 16 Draft Methods of conducting the test and interpretation of the results are detailed in IEEE Std 95-2002 [6]. IEEE Std 95-1977 was developed to provide uniform procedures for the following:

Performing high direct-voltage acceptance tests and routine maintenance tests on the main ground insulation.

Analyzing the variations in measured current so that any possible relationship of the components of these variations to the condition of the insulation can be more effectively studied.

Many machine operators have found this test to be a useful maintenance tool, although there is some controversy on the interpretation of the test results, and breakdown sometimes occurs without prior indication. The operator is urged to study IEEE Std 95-2002 to derive significant benefit from controlled over voltage testing.

9.3.5 Alternative Method of Controlled Over Voltage Test.

An alternative test method that has been adopted by some users is the "graded time test," detailed in the Annex.

In this test, an attempt is made to provide a linear relationship of the time-dependent absorption current with the total leakage current and therefore obtain a more significant impression of the behavior of the insulation when subjected to increased voltage steps. This phenomenon is discussed in [B7].

An alternate analysis of the graphical solution is presented in [B17] by the use of a set of templates for proportioning magnitude of leakage current and determining the time schedule for the test.

Ramp DC Over Voltage Test

The ramped DC voltage test should be made, step and graded time tests are already mentioned

9.3.6 Other Over Voltage Methods.

Other specialized procedures for controlled dc over voltage testing have been developed for certain applications. The requirements of the application and the specific information desired from the test will dictate whether these methods should be considered.

In addition to the tests outlined above, there are a number of other special tests that may be useful. Some of the more frequently used tests and a summary of their performance follow.

9.3.7 Insulation Power-Factor Test or Dissipation Factor Test

These tests are mainly useful on the larger high-voltage machines (6000 V or higher).

The power factor of the insulation from the winding to core may be measured by special bridge circuits or by the volt-ampere-watt method. Equipment that is especially designed for such tests is available.

The power factor of the stator insulation will be affected by the test voltage, the type of insulation, temperature of the insulation and moisture, and voids in the insulation. Results are



Working Group Copy IEEE P56 Rev. 16 Draft also affected by conditions external to the main insulation, such as the condition of the outer wrapper or slot liner, and the type of corona control used.

Increasing power factor on the same machine over a period of time is believed to denote a general deterioration of the insulation. Generally, power-factor increase with age is usually small for machines having corona control treatment on the slot portion, whereas the increase is usually much greater on machines having coils with slot liners constructed of organic materials.

Power-factor values on complete windings are an average of the insulation of all coils. When the coils in a machine can be individually tested, power factor can be used to compare the amount of deterioration among coils that have been operating at different voltages; e.g., between line coils and neutral coils.

9.3.8 Insulation Power-Factor Tip-up Test or Dissipation Factor Tip-up Test.

IEEE Std 286-2000, Recommended Practice for Measurement of Power-Factor Tip-Up of Rotating Machinery Stator Coil Insulation, details the recommended practice for the power factor tip-up test.

The change in power factor of the stator insulation as the test voltage is raised from some low value to a voltage that may be as much as twice the normal operating voltage may be indicative of the amount of ionization loss in or adjacent to the insulation. It is believed that an increase in ionization loss over a period of years indicates an increase in the size and number of voids and, hence, is an indication of deterioration within the insulation.

9.3.9 Slot Discharge and Corona Probe Tests.

The slot discharge test is conducted for the purpose of checking the adequacy of the ground connection between the surfaces of the coil and the core. Should surface discharging exist, it is important that it be detected, since accelerated deterioration of the ground wall insulation may occur. This test is usually applicable to machines with operating voltages in excess of 6000 V.

The corona probe test is intended to be an indicator and locator of unusual ionization within the insulation structure. The ability of this test to discriminate between harmful and acceptable levels of general ionization phenomena, such as occur in high-voltage windings, has not yet been demonstrated but is under study.

Further details of these tests are given in Annex A, and in [B9] and [B12].

9.3.10 Corona-Probe Test.

The corona-probe test is intended to be an indicator and locator of unusual ionization about the insulation structure. This test is sensitive to endwinding surface corona, as well as internal-cavity ionization in the insulation structure. Compared to slot discharge, the discharge energies involved in surface corona or internal-cavity ionization may be of a much lower order of magnitude. The energy in the discharge varies as the square of the voltage across the gap and directly as the effective capacitance at the point of breakdown.

Partial Discharge (Corona) has several undesirable effects, such as chemical action, production of heat, and ionic bombardment. The deteriorating effects of corona are dependent on its intensity and repetition rate as well as the design of the insulation system involved.



Working Group Copy IEEE P56 Rev. 16 Draft Inorganic insulation components such as mica and glass are not affected seriously by corona. Charring or decomposition of organic materials will occur in the vicinity of continued corona activity. However, surface effects may be limited by insulating finish treatments incorporating pigmentation to resist attack from the weak acid deposits formed by surface corona in the presence of oxygen and moisture.

Corona-probe test equipment consists of three basic units:

(1) Equipment capable of energizing the stator winding at its normal operating line-to-neutral voltage at rated frequency.

(2) An antenna or corona probe.

For end-winding corona measurement, the antenna is usually about 1 in long, surrounded by an insulation housing, and mounted on the end of a long insulating handle.

For internal-cavity-discharge (corona) measurements, a coil is used that is wound on a ferrite rod approximately 2 in long by 0.25 in diameter and mounted on the end of an insulating handle. Measurements are made by placing the ferrite rod over the teeth enclosing the coil being tested.

(3) An amplifier and indicator (for connection to the antenna) or a peak-pulse meter (for connection to the ferrite corona probe).

The amplifier is one of the usual type for audio frequencies and must reject 60 Hz and radio frequency signals. The indicator may be earphones, an output meter, or a cathode-ray oscilloscope.

The peak-pulse meter is a broadband instrument calibrated in units of pico-coulombs (pC) of apparent charge. Measurements may be obtained from the meter itself or by connecting the meter output to an oscilloscope or chart recorder.

The ability of the test to distinguish varying intensities of external corona activity and internal cavity corona has been established. However, the evaluation of data, to permit discrimination between harmful and acceptable levels, has not yet reached the stage where industry standards are established.

It should be noted that when performing this test personnel may encroach on recommended limits of approach to energized equipment. For this reason this test should only be carried out by experienced personnel and recommended minimum limits of approach maintained at all times.

9.3.11 Partial-Discharge Tests

The off-line partial discharge test is used to help determine the condition of the ground insulation in the slot sections of a stator winding. Also in windings rated 6kV and above and lower voltage ones motors supplied from voltage source converter drives it can determine the condition of semi-conductive voltage stress control coating in the slot regions. It can also identify degradation of the interfaces between the semi-conductive and grading coatings in high voltage windings.



Working Group Copy IEEE P56 Rev. 16 Draft During an off-line test the machine is stationary, de-energized from the system and energized by an ac test source, and so it will be exposed to different stresses from those present in operational service. This is so because when the winding is energized in service at rated line-to-line voltage (VLL) the phase-to-ground voltage varies from about VLL/√3 at the line end of each phase to virtually zero at the neutral end. On the other hand, for an off-line PD test the voltage throughout the whole phase is at the applied phase-to-ground level.

Typically during an off-line test there are:

(i) There are higher groundwall voltages towards the neutral end of the winding as it is energized to the same voltage potential as the line end and no interphasal voltages are present.

(ii) The winding is at a lower temperature and so voids in the ground insulation are larger.

(iii) There are no mechanical forces, vibration etc….

In all cases, it is not possible, nor practical to directly compare off-line results with on-line results because of the differences in electrical, mechanical, and thermal stresses between the two test conditions.

Some guidelines on how to perform this test are given in IEEE 1434-2000 and IEC 60034-27(2006). For this test, to ensure that it is only the stator winding that is being tested, it should be disconnected from all external buswork, and auxiliary equipment such as transformers, surge arrestors, surge capacitors, etc. Also, if possible the three winding phases should be disconnected from one another.

This test is designed to measure partial discharge activity in a winding as described in section 7.2.8 to allow analysis of this to identify winding insulation system degradation.

The key measurement in a PD test is the peak PD mag f the highest PD pulse. Generally this value is determined for a ure ? gives an example for a pulse repetition rate of 10 pulses alues measured in terms of mV, pC, mA, or µV. The phase angl ation of the source of the measured activity while the magnitu f the insulation degradation. For example for ground insulation ng of the bonding resin, as described in Section 7.2.1, what is k nd is indicated by clumps of activity centred around 45° and e-to-ground voltage (see Figure ?) The magnitudes of the pos e an indication of the severity of such degradation. The rel and negative Qm values give an indication of the type of slotapproximately equal the ground insulation degradation isthickness. If Qm+ is much larger than Qm- then thdegradation as described in Section 7.2.6. On the other hthere is likely separation between the groundwall insuthermal cycling as described in Section 7.2.2

Revis

nitude Qm, i.e. the magnitude o specific pulse repetition rate. Fig per second (PPS) and for Qm ve of the PD activity gives an indicde of Qm indicates the severity o delamination due to thermal aginown as “classic” PD is present a 225° relative to a cycle of phasitive and negative Qm values givative magnitudes of the positive
in e n is g an + la y
io

sulation degradation e.g. if they ardistributed throughout the insulatio indicates semi-conductive coatind if Qm- is much greater than Qmtion and conductor stack from sa

Printed: 10/18/2007 7:07:00 PM -30

-20

-10

0

10

20

30

-30

-20

-10

0

10

20

30

0 45 90 135 180 225 270 315 360

Pul

se M

agni

t

Phase An

Bipolar Slot Total

ude

[mV

]

gle [deg]

0 to 3.16 pps 3.16 to 10 pps 10 to 31.6 pps 31.6 to 100 pps

100 to 316 pps 316 to 1000 pps > 1000 pps Subset 8

n



a) Test Instrumentation

Typically the following test equipment is required;

– High Voltage Capacitive couplers to facilitate PD measurements when connected to an

appropriate test instrument – A variable voltage ac supply capable of energizing at least one phase of the winding and

preferably all three phases to the stator winding rated phase-to-ground voltage. This power supply should preferably be PD free.

– Test instrument to connect to the capacitive couplers together with a laptop computer to acquire PD measurements

– Short connectors that are PD free when energized to connect the test instrument to the HV capacitive couplers

b) Noise Reduction

The two most common methods of ensuring noise does not affect the measured PD values are to use a PD free power source connected to the same point as the PD coupler(s), or if PD testing in the Megahertz range is being done and the winding phases can be separated, energize the winding from one end and take PD measurements from the other end. The latter procedure uses the stator winding to attenuate and disperse noise so that it is not mistaken for PD.

c) Test Procedure for Individual Phases with other Two Grounded

1. An IR test as described in Section 9.3.1 should be conducted prior to the partial discharge tests to determine the suitability of the winding for further testing. The insulation resistances for this new stator winding should indicated the stator winding to be clean and dry and therefore acceptable for off-line PD testing.

2. Connect the power supply and PD couplers to the stator winding phase to be tested

3. Connect the test instrumentation to the PD couplers

4. Ensure that the machine stator frame and the other two phases are solidly connected to a ground termination

5. Raise the test voltage slowly to the maximum test voltage and maintain this voltage until the PD stabilizes (can take 15-20 minutes) and the record PD data. Note the PD tends to decrease over this time period PD normally reduces due to space charge effects and a build up of pressure with voids in the insulation. Also at this time an ultraviolet camera can be used to look for surface PD activity in the areas of the semi-conductive/grading coating interfaces



Working Group Copy IEEE P56 Rev. 16 Draft 6. Slowly lower the applied voltage until no PD is detected and record this voltage which is

known as the PD Extinction Voltage (PDIV)

7. Reduce the applied voltage further and then slowly increase it until PD is detected and record the value at which this occurs. This voltage is known as the PD Inception Voltage (PDIV)

8. Repeat the test for the other two phases

d) Test Procedure for Individual Phases with All Energized

1. Perform IR test

2. Connect all three phases together and the power supply to a common connection point

3. Connect PD coupler to phase to be tested or one to each phase

4. Ensure stator frame is grounded

5. Perform remainder of tests per steps 9.3.11(b) 5 to 8 above

e) Test Interpretation

1. Evaluate PD levels and characteristics at the maximum phase-to-ground test voltage to evaluate the types of insulation degradation present and the severity, e.g., thermal degradation, semi-conductive coating degradation, etc.

2. Evaluate the findings from ultraviolet camera scans

3. Comparison of the results from Tests 9.3.11(b) and (c) may indicate some phase –to-ground PD activity due to contamination, or inadequate spacing between endwindings, or circuit ring bus connections in different phases. This is done by comparing the results for each phase to see if the PD levels for test (c) are higher than for test (d). If this is so interphasal activity is indicated.

4. For a well consolidated winding the ration of the PDEV or PDIV to the PD at rated phase to ground voltage should be 0.5, or higher [B3]. If the PDEV and PDIV values are low this of a further indication of significant insulation system degradation.

9.3.12 Slot-Discharge Test.

The slot discharge test is made for the single purpose of checking the adequacy of the electrical contact between conducting-coil surfaces and the iron of stator slots. Loss of this electrical contact results in a relatively high-energy discharge between the conducting-coil surface and the core. The energy is that resulting from the discharge of a substantial portion of the coil-side capacitance. Since greatly accelerated deterioration of the major ground insulation is produced by slot discharge, early detection and correction of this condition is important.



Working Group Copy IEEE P56 Rev. 16 Draft Slot-discharge analyzers utilize detection circuits resonant in the frequency range where energy from surface discharging is high (approximately 2500 Hz), while blocking 60 Hz voltage by means of a high-pass filter.

Tests are made with the winding energized at approximately the operating stress to ground. Detection is accomplished by connecting the slot-discharge analyzer to the machine terminals, one phase at a time. When a discharge exists, high-frequency reflections are readily observable on a cathode-ray oscilloscope connected to the slot-discharge analyzer output. Location of specific coils suffering slot discharge is accomplished by a probe test. The probe test utilizes the slot-discharge analyzer in conjunction with a probe that successively contacts the conducting surfaces of individual stator coils.

9.3.13 Turn-To-Turn Insulation Test.

In cases where the integrity of the insulation between adjacent turns in a coil is subject of concern, tests should be made to establish that a desired level of insulation strength is present. Test equipment, employed in the application of turn insulation tests, is usually of the type where a capacitor is alternately charged and then discharged into the coil under test (or into an inducing coil which has been placed in the stator bore, over the coil under test).

Since the insulation between turns of stator coils varies greatly in types of insulating materials, types of construction, and spacing, test values are usually determined after consultation with the coil or machine manufacturer. Any test value selected to verify the adequacy of interturn insulation should be based on the design, physical spacing and electrical strength of the insulating system. Refer to IEEE 522 [this has now been released – need to update the reference, fix the reference format, and add it to the correct section], Guide for Testing Turn-to-Turn Insulation on Form Wound Coils for Rotating Machines, for turn-to-turn testing (see Section 9).

9.3.14 Insulation-Resistance Test of Embedded Temperature Detectors.

Stator winding embedded temperature detectors (resistance or thermocouple types) are connected by cable to a terminal board on the frame of the unit. Often one lead of each of the detectors is connected to a common ground strip at the terminal board. If a detector located in the slot portion should become grounded, circulating currents could occur between that ground and the terminal board ground. To guard against this possibility, insulation resistance measurement should be made on the detectors at convenient intervals. Tests are usually made on all detectors simultaneously at 500 V dc after the terminal board common has been isolated from ground. (Recording equipment, connected externally to the terminal board, should be isolated from the test potential.)

9.3.15 Coil-to-Core Contact Resistance.

It is essential that the corona suppression coatings applied to the surface of coils in high-voltage windings be adequately grounded. A low-resistance grounding path is usually provided by direct-contact resistance with the stator core. Measurement of the coil-to-core contact resistance may provide a useful indication of the condition of the corona suppression system.



Working Group Copy IEEE P56 Rev. 16 Draft 9.3.16 Stator Core Interlaminar Resistance Insulation Test

In cases where several core laminations have been short-circuited by the causes referred to in 5.4, repairs will usually be required to restore the interlaminar insulation.

An effective test can be made by inducing in the stator core a flux at rated frequency at approximately the flux density in the core corresponding to 105% of rated voltage. This test is known as the loop test and can be performed by passing a temporary coil through the stator bore and then around one side of the frame. This coil should be insulated from the core and frame and be braced securely in position. A single-turn test coil is similarly wrapped around the core and connected to a voltmeter.

Methods of calculating the test-coil voltage and ampere-turn requirement are given in Annex A.

Another technique developed, an electromagnetic core imperfection detector, offers a simpler procedure with lower kVA requirement for core-fault detection.

The following was from rev 13 and is included because I received editorial changes independent to Richard’s:

9.3.17 Slot Discharge and Corona Probe Tests.

The slot discharge test is conducted for the purpose of checking the adequacy of the ground connection between the surfaces of the coil and the core. Should surface discharging exist, it is important that it be detected and addressed, since accelerated deterioration of the ground wall insulation may occur. This test is usually applicable to machines with operating voltages in excess of 6000 V.

The corona probe test is intended to be an indicator and locator of unusual ionization within the insulation structure. The ability of this test to discriminate between harmful and acceptable levels of general ionization phenomena, such as occur in high-voltage windings, is difficult; however, it is an excellent comparative tool to detect and locate localized areas of high discharges. .

Further details of these tests are given in Annex A, and in [B9] and [B12].

9.3.18 Corona-Probe Test.

The corona-probe test is intended to be an indicator and locator of unusual discharges throughout the insulation structure. This test is sensitive to endwinding surface corona type discharges, as well as internal-cavity discharges in the insulation structure. Compared to slot discharge, the discharge energies involved in surface corona or internal-cavity type discharges may be of a much lower order of magnitude. The energy in the discharge varies as the square of the voltage across the gap and directly as the effective capacitance at the point of breakdown.

Corona type partial discharge has several undesirable effects, such as chemical action, production of heat, and ionic bombardment. The deteriorating effects of these discharges are dependent on its intensity and repetition rate as well as the design of the insulation system involved.



Working Group Copy IEEE P56 Rev. 16 Draft Inorganic insulation components such as mica and glass are not affected seriously by partial discharges. Charring or decomposition of organic materials will occur in the vicinity of continued partial discharge activity. However, surface effects may be limited by insulating finish treatments incorporating pigmentation to resist attack from the weak acid deposits formed by surface discharges in the presence of oxygen and moisture.

Corona-probe test equipment consists of three basic units:

(1) Equipment capable of energizing the stator winding at its normal operating line-to-neutral voltage at rated frequency.

(2) An antenna or corona probe.

For end-winding corona type discharge measurement, the antenna is usually about 1 in long, surrounded by an insulating housing, and mounted on the end of a long insulating handle. This is a dangerous test, and is not recommended.

For internal-cavity-discharge measurements, these utilize a multiturn coil wound on a ferrite rod approximately 2 in long by 0.25 in diameter and mounted on the end of an insulating handle. Measurements are made by placing the ferrite rod over the teeth enclosing the coil being tested. The probe should not be used beyond the stator core boundaries. Insulated gloves and ground instrumentation procedures should be used.

(3) An amplifier and indicator (for connection to the antenna) or a peak-pulse meter (for connection to the ferrite corona probe).

The amplifier is a broad band radio frequency (RF) intensity meter and must reject 60 Hz and high frequency signals. The indicator is usually, an output meter, earphones, or a cathode-ray oscilloscope.

The peak-pulse meter is a broadband instrument usually tuned between 200 kHz and 10 MHz. Measurements may be obtained from the meter itself or by connecting the meter output to an oscilloscope or chart recorder.

The use of the corona-probe test and the evaluation of test data obtained have been around since the 1960s and is sometimes referred to as the TVA corona probe. The ability of the test to distinguish varying intensities of external corona activity and internal cavity corona has been established. However, the evaluation of data, to permit discrimination between harmful and acceptable levels, has not yet reached the stage where industry standards are established.

9.3.19 Partial-Discharge Tests

[Need someone to write this section]



Working Group Copy IEEE P56 Rev. 16 Draft 9.3.20 Slot-Discharge Test.

The slot discharge test is made for the single purpose of checking the adequacy of the electrical contact between conducting-coil surfaces and the iron of stator slots. Loss of this electrical contact results in a relatively high-energy discharge between the floating coil surface and the core. The energy is that resulting from a substantial portion of the coil-side capacitance discharging to the grounded core. Since greatly accelerated deterioration of the major ground insulation is produced by slot discharge, early detection and correction of this condition is important.

Slot-discharge analyzers utilize detection circuits resonant in the frequency range where energy from surface discharging is high (approximately 2500 Hz), while blocking 60 Hz voltage by means of a high-pass filter.

Tests are made with the winding energized at approximately the operating stress to ground. Detection is accomplished by connecting the slot-discharge analyzer to the machine terminals, one phase at a time. When a discharge exists, high-frequency reflections are readily observable on a cathode-ray oscilloscope connected to the slot-discharge analyzer output. Location of specific coils suffering slot discharge is accomplished by a probe test. The probe test utilizes the slot-discharge analyzer in conjunction with a probe that successively contacts the conducting surfaces of individual stator coils.

If the rotor is removed to provide access to the stator-bore surface, an alternate test may be made to provide a partial check of the adequacy of coil surface grounding in the slot portion of the stator. With slot wedges removed, contact-resistance measurements between exposed top-coil surfaces and the iron of the slot may be taken with a low-voltage ohmmeter, used with a suitable probe. If slot wedges are not removed, it is often possible to obtain these resistance measurements, through core vents, from the coil side to the iron of the core. Resistance measurements should be made at several locations along the slot length of every stator coil. Since bottom coil sides are relatively less accessible, evaluations are usually based on values measured on top-coil sides. Values of coil surface contact resistance, when in accordance with manufacturer's recommendations, verify adequate coil surface grounding and the absence of slot discharge.

9.3.21 Resistance Temperature Detectors (RTDs)

Resistance temperature detectors are resistance coils so constructed to allow the temperature to be measured by a change in resistance.

Resistance temperature detectors are made with a resistance element constructed using a material for which the electrical resistivity is a known function of the temperature so as to allow the temperature to be measured by a change in resistance.

Measurements are usually made in order to verify that RTDs are properly connected and that they are free of undesired ground contacts or open circuits. Measurements consist of



Working Group Copy IEEE P56 Rev. 16 Draft comparisons of readings from each RTD with all others, and should be made at room temperature. Refer to IEEE Std 118-1949 and IEEE Std 119-1974 (see Section 9).

For the measurements, a resistance bridge is normally used. A special RTD meter for directly reading temperatures of detectors can also be used. Referring to Fig. 1, all three leads of a given RTD must be of equal length and wire size so that the three lead resistances all have the same value. By subtracting the resistance measured between terminals A and B from the resistance measured between terminals A and C (or B and C), the resistance of the temperature sensing element alone can be accurately determined. The temperature element is usually made of copper alloy wire and is appropriately sized so that its resistance at 25°C is 10Ω. From the measured change in resistance, the temperature of the element may be calculated. After proper meter corrections are applied, temperature readings of each RTD and the thermometer readings should agree to within plus or minus 3°C.

Figure 1. RTD (3 wire) and Wheatstone Bridge Circuit

9.3.22 Insulation-Resistance Test of Embedded Temperature Detectors.

Stator winding embedded temperature detectors (resistance or thermocouple types) are connected by cable to a terminal board on the frame of the unit. Often one lead of each of the detectors is connected to a common ground strip at the terminal board. If a detector located in the slot portion should become grounded, circulating currents could occur between that ground and the terminal board ground. To guard against this possibility, insulation resistance measurement should be made on the detectors at convenient intervals. Tests are usually made on all detectors simultaneously at 500 V dc after the terminal board common has been isolated from ground.



Working Group Copy IEEE P56 Rev. 16 Draft (Recording equipment, connected externally to the terminal board, should be isolated from the test potential.)

9.3.23 Insulation Resistance Test of Insulated Stator-Through-Bolts.

Insulation resistance to ground of stator through-bolts should be measured at a voltage level recommended by the manufacturer.

9.3.24 Coil-to-Core Contact Resistance.

It is essential that the semi-conducting corona suppression coatings applied to the surface of coils in high-voltage windings be adequately grounded. A low-resistance grounding path is usually provided by direct-contact resistance with the stator core. Measurement of the coil-to-core contact resistance may provide a useful indication of the condition of the corona suppression system.

9.3.25 Winding Resistance.

A reduction in winding resistance may indicate shorting of conductors. An increase in winding resistance may indicate poor connection.

Resistance of the stator winding is usually measured with a low-resistance (Kelvin) bridge or by the drop-in-potential method. Refer to IEEE Std 118-1949, Master Test Code for Resistance Measurement (see Section 9). The measurement is normally made for each phase separately. The stator winding should be at room temperature when the cold resistance measurement is made, and the temperature of the winding carefully determined. Refer to IEEE Std 119-1974, Recommended Practice for General Principles of Temperature Measurement as Applied to Electrical Apparatus (see Section 9).

The resistance-temperature characteristic of copper in the range of temperatures usually encountered is a straight line which, if extrapolated, intersects the zero resistance axis at -234.5°C. Based on this characteristic, the temperature corresponding to any resistance of a copper winding may be determined from the formula:

234.5 - ) t (234.5 RR t 1

1

2 2 +=

where R1 and R2 are winding resistance in ohms measured at temperatures of t1 and t2°C, respectively.

9.3.26 Test of Interlaminar Insulation of Stator Core.

A test, often referred to as a loop test or a ring test, has been found to be an effective and relatively easy method of testing interlaminar resistance of stator core. Prior to conducting this test it is recommended that the manufacturer be consulted.



Working Group Copy IEEE P56 Rev. 16 Draft In cases where significant damage is visually evident, the interlaminar insulation must be reestablished before the application of the test. Otherwise, additional overheating and burning damage may occur during the test.

9.3.26.1 Safety Considerations.

Considerable hazard may exist in connection with this test. All test personnel involved should be familiar with the safety precautions that should be observed. The following is a list of some of the major safety items related to this test:

(1) Shielded cable should never be used for the magnetizing coil as applied voltage will also be induced into the cable shield.

(2) Do not go near the magnetizing coil or the stator core when the test setup is energized.

(3) All electrical connections should be checked before a trial application of power is made.

(4) Appropriate fire protection should be made available during the test.

(5) Liquid cooling system, if present, should be drained and remain empty.

(6) Machine terminals should be opened and safely covered and flagged.

(7) Stator RTDs and their recorder should remain in service during tests.

(8) Adequate phone and other communication systems should be established among various points for proper test control.

(9) If thermocouples are used for temperature measurements, a considerable personnel hazard may exist since up to full search coil voltage can be induced in the thermocouple lead. Also, care should be exercised to avoid short circuiting laminations with the thermocouple lead.

(10) On large machines, for example, steam-turbine generators, cables should be secured against motion during energizing.

(11) Care must be taken that no metallic objects are in contact with the air-gap side of the core laminations. Also extraneous metallic structural objects, metal ladders, crane cables, etc, which might form a conducting circuit around the core, should not be left in the machine during test.

9.3.26.2 Test Procedure.

A summary of helpful notes and precautions concerning the performance of this test are given in Annex A.

9.3.27 Stator Core Interlaminar Resistance Insulation Test

In cases where several core laminations have been short-circuited by the causes referred to in 5.4, repairs will usually be required to restore the interlaminar insulation.



Working Group Copy IEEE P56 Rev. 16 Draft An effective test can be made by inducing in the stator core a flux at rated frequency at approximately the flux density in the core corresponding to 105% of rated voltage. This test is known as the loop test and can be performed by passing a temporary coil through the stator bore and then around one side of the frame. This coil should be insulated from the core and frame and be braced securely in position. A single-turn test coil is similarly wrapped around the core and connected to a voltmeter. However, in some cases, such as in a core that has splits, the results of the two tests do not always agree. The loop test is the preferred method, but in some long horizontal steam units, it is not performed due to safety considerations. Three concerns involving the low flux test are that it does not generate operational-type core vibrations, the core plate voltages are only a small fraction of what is seen in normal operation, and there is very little core heating. Thus, conditions to detect various core shorts are less than ideal.

Methods of calculating the test-coil voltage and ampere-turn requirement are given in Annex A.

A recently developed technique, an electromagnetic core imperfection detector, appears to offer a simpler procedure with lower kVA requirement for core-fault detection.ii

10. Cleaning [Lori Rux – Clause head]

[Lori Rux volunteered to revise this section]

10.1 General

Care and good judgment must be used in any electric machinery cleaning program. Excessive or harsh cleaning procedures can damage an otherwise good machine and may result in expensive repairs or replacement. However, cleaning is sometimes necessary, such as when surface contamination degrades electrical insulation performance or reduces the heat transfer capability of the machine. Electric machines, or components thereof, may undergo cleaning on site or at a field service shop.

The need for cleaning may be indicated from: (1) Operation and maintenance history (2) Equipment application (e.g., location in a polluted environment) (3) Visual inspection (4) Low insulation resistance measurements (5) Overheating

Once the need for cleaning has been established, the cleaning method should be tailored to the type of contamination and the extent of the contamination buildup. Prior to any cleaning, it may be prudent to test for lead and asbestos, as the presence of these substances will affect the choice

iiAn electromagnetic core imperfection detector is a device for detection of current flow between core laminations with excitation.



Working Group Copy IEEE P56 Rev. 16 Draft of PPE (such as respirators, goggles, and rubber gloves), cleanup methods, and waste disposal procedures.

After cleaning (and drying, if necessary) the machine’s surface condition should be checked for cracks, porosity, or other damage caused by harsh cleaning methods. The desired surface finish should be reestablished by the application of suitable varnishes, paints, or resins.

10.2 Cleaning Techniques

The preferred method of cleaning depends on the component to be cleaned, as well as on the type and severity of the contamination to be removed. Whenever possible, consult with the equipment manufacturer to select cleaning materials and methods which are safe for workers and not damaging to the equipment. It may be necessary to provide forced ventilation (via fans, ventilation tubes, an air hood, etc.), particularly when work is being performed in the bottom of the stator pit or in other areas where air flow is restricted. The following cleaning methods are listed in increasing order of severity and potential harm to workers and the equipment being cleaned.

10.2.1 Vacuum Cleaning Loose dirt deposits such as carbon dust, coal dust, and fly ash can be removed by vacuum cleaning with an industrial type vacuum cleaner and long hose. Nozzle shapes may be varied to facilitate cleaning hidden or difficult-to-reach areas. Contaminants can be dislodged for vacuum pickup by:

(1) Rubbing with dry cloths (2) Brushing with a plastic or natural bristle brush (with the bristles cut short if a stiff

brush is needed) (3) Scraping with soft wood or fiber scrapers

Note that wire brushes or metal scrapers should not be used to loosen surface dirt because of possible damage to the surface being cleaned and the dangerous possibility of introducing magnetic or other metallic particles into the stator winding or core assembly.

10.2.2 Air-Lance Cleaning

Clean, dry compressed air may be used to blow out air vents or to dislodge trapped contaminants. It is recommended that the air pressure not exceed 100 psi to avoid damaging the insulation or other fragile components. A second round of vacuum cleaning may be necessary to remove materials that were dislodged by the compressed air.

10.2.3 Solvent Cleaning

To prevent harming workers and damaging equipment, extra care must be exercised when using liquid cleaning solvents. Avoid the use of excess solvent, which can wash dirt deposits into cracks, crevices, and other inaccessible areas. Try the cleaning fluid on a small area first and then check for damage and effectiveness. Do not smoke or eat around solvents. If the solvent is



Working Group Copy IEEE P56 Rev. 16 Draft flammable, keep ignition sources out of the cleaning area and have fire extinguishers available for use. It may be advisable to limit worker exposure to cleaning solvents and vapors by having more people working on the cleanup for shorter periods of time.

Mild detergents and diluted alcohol are often effective in cleaning electrical equipment, and their use should be considered before applying harsher chemicals.

Petroleum solvents may be used sparingly for removing oily and greasy contaminants from machine components, including asphaltic or synthetic-resin types of insulation. Quite often a lint-free cloth lightly dampened with solvent is effective for surface cleaning. Avoid saturating asphaltic-type insulations, which could lead to softening of the insulating materials. Gasoline, naphtha, and similar liquids are not to be used for cleaning because of fire and explosion hazards.

If a stronger or faster-drying solvent is required, a chlorinated safety solvent can be used on asphaltic and synthetic-resin types of insulation. Again, solvent-dampened cloths are often sufficient for wiping off contaminants. Chlorinated solvents must not be used on stainless steel components without first consulting with the equipment manufacturer because of the possibility of stress corrosion caused by the chlorides. Chlorinated solvents must not be used on aluminum or copper components because of chloride attack.

Mixtures of petroleum solvents and chlorinated solvents can provide better cleaning capability than the petroleum solvents alone. Such mixtures must be considered flammable even though in some proportions they might not be. However, differences in evaporation rates can change the flammability characteristics of the blend over time.

Neither petroleum solvents nor chlorinated solvents should be used on silicone insulated windings because of the degrading effect on this type of insulation.

Carbon tetrachloride and benzene are highly toxic solvents and are not to be used for cleaning.

Solvent cleaning of cylindrical rotors should be avoided. Cleaning of cylindrical rotors should be limited to vacuuming, blowing with dry compressed air, wiping with dry or solvent dampened cloth, or combinations of these three methods. The need for more extensive cleaning may involve retaining ring removal to provide access to areas where contaminants are trapped. Carbon brushes should not be allowed to absorb solvents, particularly the chlorinated types.

10.2.4 Abrasive Blasting

Abrasive blasting is used to remove paint, oil, dirt, grime, and other contaminants from hard surfaces such as stator cores. Ground corncobs, pulverized walnut shells, or other abrasive materials are discharged from a pressurized blasting machine through a nozzle attached to a flexible hose. The abrasive particulates impact the surface and knock off the unwanted coating or contaminant. Crushed corncobs are especially effective in removing oily contaminants, whereas highly abrasive walnut shells rapidly remove unwanted surface paint. Regardless of the abrasive media being used, the air-abrasive blast must not be held too long on any one area or the



Working Group Copy IEEE P56 Rev. 16 Draft component being cleaned could be damaged by abrasion. In addition, care must be taken to avoid blowing the abrasive material into inaccessible areas where it cannot be completely removed and may block ventilating passages or cause mechanical imbalance during operation.

10.2.5 CO2 Blasting (Cryogenesis)

CO2 cleaning uses conventional blasting technology in combination with dry ice pellets. Vaporized liquid nitrogen is used to propel the dry ice particles toward the surface to be cleaned. Unlike other blast-cleaning technologies, which strictly rely on abrasive media impacting the surface to be cleaned, dry ice blasting also creates thermal differentials between the contaminant and the surface (due to different rates of shrinkage caused by differing thermal coefficients of expansion). These thermal differentials loosen the bonds between the contaminant and the surface, and improve the effectiveness of the blasting process. Reverse fracturing further aids in the cleaning action when molecules of the vaporized nitrogen and vaporized CO2 enter the pores of the contaminants. As the gas molecules warm and expand, they help break the bonds between the surface and the contaminants. When used properly, CO2 cleaning is a totally dry process and produces no secondary waste. It works best removing loose, non-oily surface contamination on hard, non-porous surfaces. As with other blasting techniques, some areas may be difficult to reach, depending on the geometry of the component being cleaned. Furthermore, poorly adhering surface paint may be knocked off and more serious damage may occur if the procedure is not performed carefully.

10.2.6 Steam Cleaning

Steam cleaning utilizes a high velocity jet of steam and water containing a mild nonconductive detergent. The detergent spray is followed by multiple clean water rinses. The steam cleaning method is effective on heavily contaminated windings and windings subjected to flooding or salt contamination. The steam cleaning method usually can be used on silicone-insulated windings.

Prior to returning a steam-cleaned machine to service, it must be dried or baked to remove all moisture from the windings and to obtain an acceptable insulation resistance value. If voltage is applied before all moisture has been removed, there is a risk of insulation failure. Regardless of the method used for drying the insulation system, dryout temperatures should not exceed 75°C to 85°C, and the rate of temperature rise should be limited to 5°C per hour. In exceptional cases, where insulation resistance has not reached acceptable levels after 24 hours (or more) of drying, consideration may be given to increasing the maximum temperature to 100°C to 105°C. However, at temperatures of 100°C and above, the possibility of insulation damage increases as gases and vapors generated within the insulation by high temperature develop pressure and are forced through the insulation. This can break the continuity of the layers and cause delamination, or actually rupture the material. Ventilation is required to remove the water vapor during the heating cycle.



Working Group Copy IEEE P56 Rev. 16 Draft 10.2.7 Cleaning By Water Immersion or Water Hose

Many of the machines covered by this guide are too large for immersion, although heavily contaminated or flooded machines can be washed with a hose. Baking and drying precautions noted under steam cleaning would also apply for water immersion or water hose cleaning.

Silicone-insulated windings can be generally cleaned using the water hose method with a non-ionic, non-sudsing detergent.

Add to Bibliography:

Dionne, Don and Stefano Bomben, “Blasting Away the Dirt,” Hydro Review, June 2004, pp. 57-58.

Rux, Lori., “Rehabilitation of Flood-Damaged Hydroelectric Generators,” Proceedings of the 1999 International Electric Machines and Drives Conference, February 1999.

10.2.4 Cleaning Instructions

Proper maintenance of electrical equipment requires periodic visual examination of the machine and windings and appropriate electrical and thermal checks. Insulation surfaces should be examined for cracks and accumulations of dirt and dust to determine required action. Lower than normal insulation resistance can be an indication that conductive contaminant is present. The contaminant may be carbon, salts, metal dusts, or virtually any dirt saturated with moisture. These contaminants develop a conductive path to produce shorts or end turn tracking that can eventually lead to failure. Cleaning is also advisable if heavy accumulations of dirt and dust can be seen, or are suspected to be restricting ventilation as manifested by excessive heating.

With no visual, electrical, or thermal evidence that dirt is present, cleaning should not be initiated since more harm than good may result.

If harmful dirt accumulations are present, a variety of cleaning techniques are available. The one selected will depend on – The extent of the cleaning operation to be undertaken – The type, rating, and insulation structure of the machine involved – The type of dirt to be removed

10.2.5 Field Service Cleaning of Assembled Machines.

Prior to any cleaning, samples should be taken and tested for lead and asbestos, as this will determine cleanup methods and procedures.

Where cleaning is required at the installation, and complete disassembly of the machine is unnecessary or not feasible, dry dirt, dust, or carbon should first be picked up by a vacuum cleaner to prevent the redistribution of the contaminant. A small nonconductive nozzle or tube



Working Group Copy IEEE P56 Rev. 16 Draft connected to the vacuum cleaner may be required to reach dusty surfaces or to enter into narrow openings such as between commutator risers. After most of the dust has been removed, a small brush can be affixed to the vacuum nozzle to loosen and allow removal of dirt more firmly attached.

Suction should be used to remove dust produced within a machine, such as in the stoning of commutators, collector rings, or the seating of brushes.

After the initial cleaning with vacuum, compressed air (not to exceed 30 lb/in2 [207 kPA]) may be used to remove the remaining dust and dirt. An exhaust must be provided so that dirt will be removed from the machine. Indiscriminate blowing may produce mechanical unbalance of an armature or rotating field by redistribution of dirt.

Compressed air used for cleaning should be clean and free of moisture or oil. Air pressure or velocity should be adequately controlled to prevent mechanical damage to the insulation.

Disassembly of the machine and more effective cleaning by a qualified service shop may be required if the previously described field service cleaning procedures do not yield effective results.

10.2.6 Service Shop Cleaning of Disassembled Machines.

An initial insulation-resistance reading should be taken on the machine to check electrical integrity. A reading of not less than 1 Mõ/kV of machine rated voltage but not less than 1 Mõ would be expected with severely contaminated machines. A zero reading may indicate an insulation breakdown requiring repair, not just cleaning.

The "steam-jenny" method of cleaning, which sprays a high-velocity jet of hot water and water containing a mild detergent, is normally effective in cleaning windings including those subjected to flooding or salt contamination. The detergent spray is followed by multiple sprays with clean water to remove or dilute the detergent. Water immersion with multiple rinses and changes of water may also be used to remove contamination. The machine should then be dried (low-temperature oven may be used) until normal insulation resistance values are obtained at room temperature. Other cleaning methods may be used, such as abrasive blasting at light pressures using ground corncobs or nutshells. Large high-voltage windings should never be soaked or wet down.

Solvents are effective for removing oil or grease and may be required if water or detergent is not adequate. However, solvents may carry contamination, such as conductive dusts, metals, salts and carbon, into cracks, crevices, and interstices by direct flow or by capillary action. Removal of contamination from such inaccessible areas is virtually impossible. A solvent dampened cloth is the preferred cleaning method rather than direct application of liquids. If the cloth is dripping, then too much liquid is being applied to the insulation. Extreme care should be taken when using solvents both with respect to the equipment and the personnel. Environmental concerns, the disposal of waste products, the possibility of damage to the insulation, and health and safety hazards for the worker should be considered when selecting a solvent. Adequate ventilation and



Working Group Copy IEEE P56 Rev. 16 Draft protective gear, such as a dust mask, glasses and gloves, should be used during equipment cleaning. Of course, equipment should not be cleaned when energized.

10.3 General

Care and good judgment must be used in any program for cleaning of electrical machinery and windings. Excessive cleaning and unwise use of solvents can do more damage than good, and result in expensive rewinding or repairs.

The need for cleaning may be indicated from: (1) Previous history of machine (2) Equipment application (3) Visual inspection (4) Low insulation resistance (5) Overheating

Once the need for cleaning is established, the cleaning method can be tailored to the type of contamination and the severity of the contamination buildup. It is most important to keep the core air vents open and the voltage grading paint clean.

After cleaning (and drying if necessary) the insulation surface condition should be checked for surface cracks, porosity, or the effects of harsh cleaning methods. The desired surface insulating finish should be reestablished by the application of suitable varnishes, paints, or resins. Depending on accessibility or size, surface treatments may be applied by dipping, spraying, flooding, or brushing.

10.4 Cleaning Techniques

The method of cleaning can be adapted to the type of contamination and the buildup of contaminants. The methods listed in the following section are in increasing order of severity and possible damage to windings.

10.5 Vacuum Cleaning.

Dry contaminants such as carbon dust, coal dust, and fly ash can be removed with a vacuum cleaner. Contaminants can be dislodged for vacuum pickup by:

(4) Rubbing with dry cloths

(5) Brushing with a bristle brush

(6) Scraping with soft wood or fiber scrapers (Wire brushes or metal scrapers should not be used because of possible damage to the insulation and the dangerous possibility of introducing magnetic or other metallic particles into the winding or core assembly.)



Working Group Copy IEEE P56 Rev. 16 Draft (7) Nozzle shapes should be varied as required to enable directing the vacuum cleaning

into hidden, difficult to clean areas.

10.6 Air-Lance Cleaning.

After vacuum cleaning, additional cleaning can be done employing shaped nozzles to direct high-velocity clean dry air to dislodge trapped contaminants. It is recommended that air pressure be limited to avoid damaging the insulation.

10.7 Solvent Cleaning.

Care must be exercised in the choice and application of cleaning solvents from the standpoint of worker safety, and risk of damage to the insulation. The manufacturer should be consulted to select a solvent and method of application that is non-injurious to the winding.

10.8 Types of Solvents

Petroleum solvents of the safety type can be used for removing oily and greasy contaminants from asphaltic or synthetic-resin types of insulation. These solvents should be used sparingly. Quite often a lint-free cloth, dampened in solvent, is adequate for rubbing off the contamination. Saturation of asphaltic-type insulations should be avoided to prevent softening of the insulating materials.

Where a stronger or faster-drying solvent is required, a chlorinated safety solvent can be used on asphaltic and synthetic-resin types of insulation. For recommendations on specific solvents to be used with each given insulation the manufacturer of the machine should be consulted. Here, again, solvent-dampened cloths are often sufficient for wiping off contaminants. Refer to Section 7.2.3.3 for risk of damage.

Mixtures of petroleum solvents and chlorinated solvents can be used with better cleaning capability than the petroleum solvents alone. Such mixtures must not be considered nonflammable, even though in some proportions they might be. Differences in evaporation rates can change characteristics of the blend.

Carbon tetrachloride and benzene are highly toxic solvents and are not to be used for cleaning. Gasoline, naphtha, and similar liquids are not to be used for cleaning because of fire and explosion hazards.

10.9 Risk of Damage

Liquid solvents are effective in removing Oily contaminants, but there are risks involved, particularly from spray applications of solvents. The solvents may carry contaminants into cracks, crevices, or inaccessible areas and cause the insulation resistance to decrease to unsafe levels.



Working Group Copy IEEE P56 Rev. 16 Draft Chlorinated solvents must not be used on stainless steel components unless agreed upon with the manufacturer because of the possibility of stress corrosion caused by the chlorides. Examples of stainless steel components are: (1) nonmagnetic retaining rings and wedges on turbine-generator rotors, and (2) stator cooling oil or water systems on turbine-generators. Chlorinated solvents must not be used on aluminum or copper components because of chloride attack.

Solvent cleaning of cylindrical rotors should be avoided. Cleaning of cylindrical rotors should be limited to vacuuming, blowing with dry compressed air, wiping with dry or solvent dampened cloth, or combinations of these three methods. The need for more extensive cleaning may involve retaining ring removal to provide access to areas where contaminants are trapped. Carbon brushes should not be allowed to absorb solvents, particularly the chlorinated types.

Neither petroleum solvents nor chlorinated solvents should be used on silicone insulated windings because of the degrading effect on this type of insulation.

10.10 Abrasive Blasting

Another method for removing contaminants utilizes an air blast of ground corncobs or pulverized nutshells. This method is often successful for removing oily contaminants. The air-abrasive blast must not be held too long on any one area or the insulation will be damaged by abrasion. Care must be exercised to avoid blowing the abrasive material into inaccessible areas where it cannot be completely removed and may block ventilating passages or cause mechanical imbalance during operation.

10.11 Cleaning with Solid CO2

[Need a volunteer to write this section] Eric Eastment, Bureau of Reclamation, has managed the CO2 cleaning of at least five hydrogenerators. He is available to assist and may be reached at 303-445-2324.

10.12 Steam Cleaning

The steam-jenny method of cleaning utilizes a high-velocity jet of steam and water containing a mild nonconductive detergent. The detergent spray is followed by multiple steam and water sprays without detergent to provide adequate rinsing. The machine must then be dried or baked to remove all moisture from the windings and to obtain an acceptable insulation resistance value. If an overvoltage test is applied after steam cleaning, there is a risk of insulation failure if all moisture has not been removed or the insulation is defective.

Regardless of the procedure used for drying insulation systems, initial dryout temperatures should not exceed 75 to 85°C (reached at a maximum rate of 5°C/h). In exceptional cases, where insulation resistance does not respond to this limit after 24 h, the maximum temperature may be carefully increased to 100-105°C. At temperatures of 100°C or higher, the possibility of insulation rupture (as water changes to steam) should be a prime consideration in the time-



Working Group Copy IEEE P56 Rev. 16 Draft temperature schedule selected. Ventilation is required to remove the water vapor during the heating cycle.

The steam cleaning method is effective on heavily contaminated windings and windings subjected to flooding or salt contamination.

The steam cleaning method usually can be used on silicone-insulated windings.

10.13 Cleaning By Water Immersion or Water Hose

Many of the machines involved in this guide are too large for immersion, but heavily contaminated or flooded machines can be washed with a hose. Baking and drying precautions noted under steam cleaning would also apply for water immersion or water hose cleaning.

Silicone-insulated windings can be generally cleaned using the water hose method with a non-ionic, non-sudsing detergent.

11. Bibliography [Doug Conley– Clause head]

[B1] Alke, R. J. and Sexton, R. M., "Detection of turn-to-turn faults in large high voltage turbine generators," AIEE Transactions, pt. I, vol. 70, 1951.

[B2] Bhimani, B. W., "Very low-frequency high-potential testing," AIEE Transactions Paper 61-138.

[B3] Culbert, I. M., Dhirani, H., and Stone, G. C., Handbook to Assess the Insulation Condition of Large Rotating Machines, EPRI EL5036, vol. 16, June 1989.

[B4] Duke, C. A., Smith, L. E., Roberts, C. A., and Cameron, A. W. W., "Investigation of maintenance tests for generator insulation," AIEE Transactions on Power Apparatus and Systems, pt. III, vol. 80, pp. 471_480, Aug. 1961.

[B5] Hermann, P. K., Mahrt, R., and Doon, H. H., "Detecting and locating interturn short circuits on turbine generator rotors," AIEE Transactions, vol. 82, pp. 686_698, 1963.

[B6] Hill, G. Leslie, "Testing electrical insulation on rotating machinery with high voltage dc," AIEE Transactions, 1953, AIEE Technical Paper 53-3.

[B7] "Insulation Testing by DC Methods," James G. Biddle Company Technical Publication 22T1, 1964.

[B8] Johnson, J. S., "Inspection and maintenance testing of high-voltage generator winding," AIEE Transactions, pt. I, vol. 70, 1951.



Working Group Copy IEEE P56 Rev. 16 Draft [B9] Johnson, J. S., "Preventive maintenance inspection and testing of motors and generators," Westinghouse Maintenance News, vol. XXIII, no. 4; vol. XXIV, nos. 1 and 2; vol. XXV, no. 1; 1960.

[B10] Kelley, W. E., "Maintenance testing of insulation resistance on diesel-electric locomotives," AIEE Transactions, 1954, AIEE Transactions Paper 54-339.

[B11] Kilman, L. B. and Dallas, J. P. A., "Discussion of dc high potential test voltage for aircraft electrical insulation," AIEE Transactions Paper 58-845, presented at the 1958 AIEE Summer General Meeting.

[B12] Kurt, M., Lyles, J. F., and Stone, G. C., "Application of partial discharge testing to hydro generator maintenance," IEEE Transactions on Power Apparatus and Systems, p. 2148, Aug. 1984.

[B13] Moses, Graham Lee, "AC and dc voltage endurance studies on mica insulation for electrical machinery," AIEE Transactions, 1951, AIEE Technical Paper 51-127.

[B14] Moses, Graham Lee and Harter, E. F., "Winding-fault detection and location by surge-comparison testing," AIEE Transactions, vol. 64, pp. 499_503, July 1945.

[B15] Moses, Graham Lee, "Review of some problems in dc testing low voltage electric machine insulation," AIEE Conference Paper, presented at the 1953 AIEE Winter General Meeting.

[B16] Oliver, F. S., "Medium voltage ac testing of rotating machinery insulation," AIEE Conference Paper 58-203, presented at the 1953 AIEE Winter General Meeting.

[B17] Schleif, F. R. and Engvall, L. R., "Experience in analysis of dc insulation tests for maintenance programming," AIEE Transactions on Power Apparatus and Systems, pt. III, vol. 78, pp 156_162, 1959.

[B18] Sidway, C. L. and Loxley, B. R., "Techniques and examples of high voltage dc testing of rotating machine windings."

[B19] Tomlinson, H. R. "Interlaminar insulation test for synchronous machine stators," IEEE Transactions Paper 52-174, pp. 676_677, Aug. 1952.

[B20] Walker and Flaherty, "Severe moisture conditioning uncovers weaknesses in conventional motor insulation systems for naval shipboard use," AIEE Transactions Paper 61-219.

[B21] Weddendorf, W. A., "The use of dc overpotential testing as a maintenance tool in the industrial plant," AIEE Transactions Paper 59-511.

[B22] Wieseman, R. W., "Maintenance overpotential tests for armature windings in service," General Electric Review (Schenectady, NY), pp. 24_28, Aug. 1950.



Working Group Copy IEEE P56 Rev. 16 Draft [Old Bibliography from ?? ] – [Resolve this question, or remove the section]

[B23] VAN HESS, B. Generator-Insulation Tests. Bulletin, Edition Electric Institute, vol 6, May 1938, pp 204-206.

[B24] SUBJECT COMMITTEE ON GENERATOR INSULATION AND TESTING. Field Testing of Generator Insulation. AIEE Transactions, vol 60, Dec 1941, pp 1003-1011.

[B25] DAVIES, E. R., and LEFTWICH, M. F. Progress Report of D-C Testing of Generators in the Field. AIEE Transactions, vol 61, Jan 1942, pp 14-18.

[B26] FIELD, R. F. The Basis for the Nondestructive Testing of Insulation. AIEE Transactions, vol 60, Sep 1941, pp 890-895.

[B27] MONTSINGER, V. M. Breakdown Curve for Solid Insulation. AIEE Transactions, vol 54, Dec 1935, pp 1300-1301.

[B28] PEEK, F. W., Jr. Dielectric Phenomena in High-Voltage Engineering. New York: McGraw-Hill Book Company, 1929.

[B29] HAYDEN, J. L. R., and EDDY, W. N. Dielectric Strength Ratio Between Alternating and Direct Voltages. General Electric Review, vol 26, 1923, pp 645-652.

[B30] GILT, C. M., and BARNS, B. L. Insulation Tests of Electrical Machinery Before and After Being Placed in Service. AIEE Transactions, vol 48, Apr 1929, pp 656-665.

[B31] 1 American Society for Testing and Materials, 1916

[B32] Race Street, Philadelphia, Pa., 19103.

[B33] HURD, D. T. Mechanism of Dielectric Breakdown. General Electric Review, vol 51, Dec 1948, pp 26-33.

[B34] ATKINSON, F. W., and TAYLOR, R. B. A Portable Instrument for Measuring Insulation Resistance at High Voltage. AIEE Transactions, vol 64, Apr 1945, pp 164-166.

[B35] ATKINSON, F. W., and HEWSON, J. Direct-Current Over-Potential Testers for High-Voltage Insulation Fault Detection, AIEE Technical Paper No 51-128, presented at the AIEE Winter General Meeting, New York, NY, Jan 22-26, 1951.

[B36] WISEMAN, R. W. Maintenance Over-Potential Tests for Armature Windings in Service. General Electric Review, vol 53, Aug 1950, pp 24-28.

[B37] HUNT, J. F., and VIVIAN, J. H. Operation of Synchronous Machines for Continuity of Service on Southern California Edison Company System. Minutes, Electrical Equipment Committee, Edison Electric Institute (New York), May 3, 1949.



Working Group Copy IEEE P56 Rev. 16 Draft [B38] HUNT, J. F., and VIVIAN, J. H. Routine Insulation Testing of Synchronous Machines. AIEE Transactions, vol 70, part 1, 1951, pp 756-762.

[B39] JOHNSON, J. S. A Maintenance Inspection Program for Large Rotating Machines. AIEE Transactions, vol 70, part 1, 1951, pp 749-755.

[B40] MOSES, G. L. Alternating and Direct Voltage Endurance Studies on Mica Insulation for Electric Machinery. AIEE Transactions, vol 70, part 1, 1951, pp 763-769.

[B41] LAFFOON, C. M., HILL, C. F., MOSES, G. L., and BERBERICH, L. J. A New High-Voltage Insulation for Turbine-Generator Stator Windings. AIEE Transactions, vol 70, part 1, 1951, pp 721-730.

[B42] ASKEY, J. S., and JOHNSON, J. S. Insulation Resistance and Dielectric-Absorption Characteristics of Large A-C Stator Windings. AIEE Transactions, vol 64, Jun 1945, pp 347-351.

[B43] JOHNSON, J. S., and WEIL, C. Factors Affecting Insulation Resistance of Large D-C Machines. AIEE Transactions, vol 65, Nov 1946, pp 705-710.

[B44] WIESEMAN, R. W. Insulation Resistance of Armature Windings. AIEE Transactions, vol 53, Jun 1934, pp 1010-1021.

[B45] RYLANDER, J. L. Effect of Temperature on Insulation Resistance. Electric Journal, Aug 1935, Sep 1937.

[B46] MARCROFT, H. C. Use of Dielectric-Absorption Tests in Drying Out Large Generators. AIEE Transactions, vol 64, Feb 1945, pp 56-60.

[B47] RYLANDER, J. L. A High-Frequency Voltage Test for Insulation of Rotating Electrical Apparatus. AIEE Transactions, vol 45, Feb 1926, pp 459-465.

[B48] FOUST, C. M., and ROHATS, N. Insulation Testing of Electric Windings. AIEE Transactions, vol 62, Apr 1943, pp 203-206.

[B49] SEXTON, R. M., and ALKE, R. J. Detection of Turn-to-Turn Faults in Large High-Voltage Turbine Generators. AIEE Transactions, vol 70, part 1, 1951, pp 270-274.

[B50] RAWLS, J. A. Power Factor Testing of Electric Equipment to Determine Insulation Values. AIEE Technical Paper No 51-37, presented at the AIEE Winter General Meeting, New York, NY, Jan 22-26, 1951.

[B51] DOBLE, F. C. The A-C Dielectric Loss and Power-Factor Method for Field Investigation of Electrical Insulation. AIEE Transactions, vol 66, 1941, pp 934, 939.

[B52] POVEY, E. H., and OLIVER, E. F. Nondestructive Testing of Generator Insulation. AIEE Technical Paper No 51-130, presented at the AIEE Winter General Meeting, New York, NY, Jan 22-26, 1951.



Working Group Copy IEEE P56 Rev. 16 Draft [B53] JOHNSON, J. S. Slot Discharge Detection Between Coil Surfaces and Core of High-Voltage Stator Windings. AIEE Transactions, vol 70, part II, 1951, pp 1993-1997.

[B54] JOHNSON, J. S., and WARREN, M. Detection of Slot Discharges in High-Voltage Stator Windings During Operation. AIEE Transactions, vol 70, part 11, 1951, pp 1998-2000.

[B55] TOMLINSON, H. R. Interlaminar Insulation Test for Synchronous Machine Stators. AIEE Transactions, vol 71, part 111, Aug 1952, pp 676-677.

[B56] Capacitance of Synchronous-Machine Armature Windings Determined for High-Potential Test. General Electric Review, Jul 1947, pp 26-30.

[B57] CAMERON, A. W. W. Diagnoses of A-C Generator Insulation Condition by Nondestructive Tests. AIEE Transactions, vol 71, part 111, Jan 1952, pp 263-274.

[B58] MARCROFT, H. C. Field Studies of Generator Windings. AIEE Transactions, vol 71, part III, Oct 1952, pp 822-829.

[B59] ALKE, R. J. D-C Overpotential Testing Experience on High-Voltage Generators. AIEE Transactions, vol 71, part 111, Aug 1952, pp 567-570.

[B60] HILL, G. L. Testing Electrical Insulation of Rotating Machinery with High-Voltage Direct Current. AIEE Transactions, vol 72, part III, Apr 1953, pp 159-174.

[B61] DUKE, C. A., SMITH, L. E., ROBERTS, C. A., and CAMERON, A. W. W. Investigation of Maintenance Tests for Generator Insulation. AIEE Transactions, vol 80, part 111, Aug 1961, pp 471-480.

[B62] SIDWAY, C. L., and LOXLEY, B. R. Techniques and Examples of High-Voltage D-C Testing of Rotating Machine Windings. AIEE Transactions, vol 72, part III, Dec 1953, pp 1121-1129.

[B63] MOSES, G. L., and HARTER, E. F. Winding-Fault Detection and Location by Surge-Comparison Testing, AIEE Transactions, vol 64, Jul 1945, pp 499-503.

[B64] JOHNSON, J. S. Preventive Maintenance Inspection and Testing of Motors and Generators. Westinghouse Maintenance News (Westinghouse Electric Corporation, East Pittsburgh, PA), vol 23, no 4, 1959, vol 24, nos I and 2, vol 25, no 1, 1960.

[B65] HERMANN, P. K., MARHT, R., and DOON, H. H. Detecting and Locating Interturn Short Circuits on Turbine Generator Rotors. IEEE Transactions on Power Apparatus and Systems, vol 82, Oct 1963, pp 686-698.

[B66] CURDTS, E. B. Insulation Testing by D-C Methods. Technical Publication 22T1, James G. Biddle Company.

[B67] HARROLD, R. T., FORT, E. M., and GOODWIN, T. A. The Interpretation of Corona and Dielectric Measurements on the Mica-Asphalt Insulation of a 30-Year-Old Waterwheel



Working Group Copy IEEE P56 Rev. 16 Draft Generator at Grand Coulee Dam. IEEE Transactions on Power Apparatus and Systems, vol 92, Nov/Dec 1973, pp 1935-1944.

[B68] FINDLAY, D. A., BREARLEY, R. G., and LOUTTIT, C. C. Evaluation of the Internal Insulation of Generator Coils Based on Power-Factor Measurements. AIEE Transactions, vol 78, part IIIA, Jun 1959, pp 268-279.

[B69] DAKIN, T. W. Corona Pulse Detection Circuits and Their Calibration, AIEE Technical Paper No 62-260, presented at the AIEE Winter General Meeting, New York, NY, Jan 28-Feb 2, 1962.

[B70] DAKIN, T. W., WORKS, C. N., and JOHN-SON, J. S. An. Electromagnetic Probe for Detecting and Locating Discharges in Large Rotating-Machine Stators. IEEE Transactions on Power Apparatus and Systems, vol 88, Mar 1969, pp 251-257.

[B71] OLIVER, J. A., WOODSON, H. H., and JOHNSON, J. S. A Turn Insulation Test for Stator Coils. IEEE Transactions on Power Apparatus and Systems, vol 87, Mar 1968, pp 669-678.




Annex A (informative) (formally from 56) Test of Laminar Insulation in Stator Core

[This Annex is not a part of IEEE Std 56-1977, Guide for Insulation Maintenance of Large Alternating-Current Rotating Machinery (10 000 kVA and Larger).]

Included in this Annex are some helpful considerations to aide in the testing of laminar insulation in stator cores.

A.1. Design of Magnetizing Coil

In order to test the stator core adequately, it is necessary to magnetize the core at approximately its normal operating peak.

The turns of the magnetizing coil should encircle the stator through the main bore (after rotor is removed) and around the outer frame. A preferable return route, if available, is near the outside diameter of the core, within the frame. On large-diameter machines (such as waterwheel generators), the magnetizing coil should be distributed around the periphery of the stator to ensure uniform flux distribution around the entire core. A clearance of 3 in to 12 in should be maintained between the magnetizing-coil conductor and solid metal (that is, metal floor, frame, etc).

A.2. Search Coil

A single turn of AWG 12 to 18 wire, insulated adequately for the volts per turn applied, should be placed around the core, preferably diametrically opposite from the magnetizing coil. The actual core flux density can be measured by placing the search coil so that it encircles only the core and does not include the frame members. On some machines this is not possible and the error in measured flux density may or may not be acceptable. An alternative is to route the search coil leads through air vents and adjust the voltage reading for the percent of laminations not included in the search coil loop.

A voltmeter connected to the search coil will read approximately the volts per-turn value calculated in Section A3.

A.3. Calculations

The following calculations are performed in designing the test. Volts-per-turn value for the magnetizing coil as well as the search coil is given by:

4.44 VPT φf= (Eq 1)



Working Group Copy IEEE P56 Rev. 16 Draft Max ( Webers/max)

( )eff

21 2

DD LB −=φ (Eq 2)

[Conversion factor = 64516.0000003 line/square inch / Tesla * 1 in / .0254 m * 1 in / 0.0254 m - DJC Note ]

where

VPT = volts (rms) per turn

f = frequency in hertz

φ = peak core flux in Webers

B = Peak flux density in Tesla (from manufacturer)

D1 = outside diameter of core in meters

D2 = diameter to bottom of stator slots in meters

Leff = effective length of core in meters

In CGS units [are these still needed?] the equation is given by:

(Eq 1) 4.44 VPT φf=

( )eff

21 2

DD LB

−=φ (Eq 2)

where

VPT = volts (rms) per turn

f = frequency in Hertz

φ = peak core flux in lines

B = peak core-flux density in lines per square inch (from manufacturer)

D1 = outside diameter of core in inches



Working Group Copy IEEE P56 Rev. 16 Draft D2 = diameter to bottom of stator slots in inches

Leff = effective length of core in inches

The effective length of core should be obtained from the manufacturer. If that is not possible, the value can be calculated as follows:

Leff = (L - Nv bv) Fs (Eq 3)

where

L = gross length in meters

Nv = number of ventilation ducts

bv = width of ventilation duct in meters

Fs = core stacking factor (from manufacturer typically 0.93 – 0.95)

From the known supply of voltage and the volts-per-turn value from Eq (1), the number of t urns for the magnetizing coil can be determined by direct division. The result should be rounded to the next higher integer. This number of turns should be used in the first trial test.

In order to determine the size of the cable necessary for the magnetizing coil, data on ampere-turns per inch of mean back iron periphery corresponding to the core-flux densities will be required. These data should be obtained from the manufacturer.

The magnetizing-coil current requirement is given by:

π

2D D I 21

tNATI

t ⎟⎠⎞

⎜⎝⎛ +

=

(Eq 4)

Comment from JRS: This is the magnetizing current. For a more accurate estimation of current requirements, the watts loss current, Iw, should be calculated as well. These two currents can then be added as vectors (90 degrees apart) (Iexc = sqrt (It2+Iw2) ).

Iw = NtVPT

CLDWKGLeffDDDD

*

***2

21**2

21⎟⎠⎞

⎜⎝⎛ −

⎟⎠⎞

⎜⎝⎛ + π

(Eq 5)




Where

Iw = the watts loss current

WKG = stator core lamination loss in Watt per kilogram (from manufacturer) at the core test peak flux density (Tesla) and frequency (Hertz)

CLD = core lamination density in kilogram per meter (typically 7600 kg/m)

D1, D2, VPT see Eq 1

Leff see Eq 3

Nt see Eq 4

where

It = magnetizing coil current in amperes

ATI = ampere-turns per inch (from manufacturer)

Nt = number of turns

π = 3.14

D1 + D2 = (see Eq 2)

Using the results from Eq (4), the approximate minimum conductor area can be calculated.

A.4. Temperature Measurements

The magnetizing coil should be located remote from the areas suspected as damaged in order to facilitate temperature measurement. Thin shavings of paraffin, thermometers affixed with a suitable putty, thermocouples, portable pyrometers, or infrared cameras can be used to detect hot spots. These should be detectable in 15 to 30 s if the low interlaminar resistance is located at the bore surface. If the low interlaminar resistance is radially outward from the tooth surface or in core areas below the bottom of stator slots, 10 to 20 min of excitation may elapse before the heat



Working Group Copy IEEE P56 Rev. 16 Draft becomes evident at the tooth bore surfaces. A final heat run of 1 to 3 h should be made after all repairs are completed.

Comment from DP16: Need a section on how temperature is measured. Send them the core test paper.



Working Group Copy IEEE P56 Rev. 16 Draft Annex B (normative)

[formally from IEEE 432…??]

[These Annexes are not a part of IEEE Std 56-200?, IEEE Guide for Insulation Maintenance of Electric Machines (5 hp to less than 10 000 hp), but are included for information only.]

B.1. Discussion of Tests Described in the Guide

B.1.2 Dielectric Absorption Test

Dielectric absorption testing involves a determination of insulation resistance as a function of time (usually up to 10 min or until insulation resistance stabilizes). This test is also used on all, or parts of, circuit-to-ground insulation of either armature or field. IEEE Std 43-2000 [2] outlines test procedures and equipment for the standard method, which is usually made at a test potential of 500 to 5000 V dc or at a value appropriate to the voltage rating of the winding or the basic insulation condition.

Insulation resistance versus time curves, particularly on higher-voltage stator winding ground insulation, are sometimes made at test potentials higher than the value specified in IEEE Std 43-2000. Test voltages used vary from 500 V dc up to potentials above normal operating levels. High-voltage direct-current (HVDC) test sets utilizing the voltmeter-ammeter method of resistance determination are usually employed.

Dielectric absorption tests are considered to be more significant than the 1 min insulation resistance tests, particularly on the higher voltage windings, because the slope of the time-resistance characteristic gives further information concerning the relative condition of the insulation with respect to moisture and other contaminants. The ratio of the ten-to-one minute insulation resistance is called the "polarization index," or PI. In terms of measured current at some fixed voltage, PI is the ratio of the one-to-ten minute values. These ratios are nondimensional in character and aid in making comparisons between insulation on machines of different physical size and between tests performed at different times on the same machine.

The voltage for making polarization index checks is usually 500 V dc. The suggested polarization index values for clean, dry windings (according to IEEE Std 43-2000) are as follows:

Class 105 Insulation: 1.5 or more



These values are considered satisfactory for varnish- or asphalt-impregnated windings. Polyester- or epoxy-impregnated windings may have different PI values, but such values have not been established. The values of PI given above are for machine windings only. Windings



Working Group Copy IEEE P56 Rev. 16 Draft must be isolated for meaningful results. If the insulation resistance measurements are made with leads, cables, or surge-protective capacitors connected to the machine, the values obtained do not necessarily apply to the winding alone.

A polarization index of less than 1.0 indicates a conduction current increasing with time. This normally would be an unsatisfactory condition and might be due to a leakage path that has not been dried out. If the index is still less than 1.0 after cleaning and drying the insulation, the apparatus manufacturer should be consulted.

An abnormally high polarization index from an older winding; i.e., a PI in the order of 5 or more for a winding in the 20-year age bracket, could indicate intact, but lifeless, dried-out insulation. An "in-service" failure could result from a sudden fracture of the brittle insulation caused by mechanical shock, such as introduced from a short circuit on the system.

Frequently, the insulation resistance is found to be higher than the basic minimum requirements, but the polarization index is below the recommended values. This condition occasionally occurs when testing field windings or dc armatures. Generally, a machine that has been idle for some time could be returned to service if the PI value is 1.0 or above and the insulation resistance is above the minimum requirement. The insulation resistance of a reliable winding structure should be well above the basic requirements and the PI value at least equal to or above the suggested minimum.

B.1.1.3 Slot Discharge and Corona Probe Tests

Two common ways of determining the presence of ionization or corona type discharges are (1) by measuring the power factor of the insulation at two voltages and noting the increase, and (2) by the use of a cathode-ray oscilloscope. The first method gives a measure of the total ionization if readings are made at two voltages, one at or slightly above the operating level and one below any practical corona level (2 kV is commonly used). Some specifications set a limit on the power-factor difference (called tip-up) obtained in this way. If the oscilloscope is connected to an exploring coil, the signal can detect the location of the corona, but will not relate whether the corona is internal or external and gives only a very rough idea of the intensity.

The slot discharge test is made for the purpose of checking the adequacy of the ground connection between the conducting coil surfaces and the core. If such surface discharging exists in a winding, it is important that it be detected. Greatly accelerated deterioration of the major ground insulation is produced by slot discharges. These tests are usually applicable to machines with operating voltages in excess of 6000 V.

Tests with a slot discharge analyzer check the adequacy of the electric contact between the core and the stator coils for those machines having low-resistance conducting surfaces in the slot portions. This test is effective in detecting and locating the presence of surface discharge, which is injurious to the coil insulation. Detection is accomplished by a simple test from the machine terminals. Discharge is located by the probe test described below.



Working Group Copy IEEE P56 Rev. 16 Draft Tests are made with the winding energized at the maximum normal operating stress to ground. When a discharge exists, high-frequency reflections are readily observable at the machine terminals on a cathode-ray oscilloscope connected to the slot discharge analyzer.

The corona probe test is intended to be an indicator and locator of unusual ionization about the insulation structure. Such a test is sensitive in varying degrees to surface corona and unusual internal ionization. The ability of this test to discriminate between harmful and acceptable levels of general ionization phenomena, such as occur in high-voltage windings, has not been demonstrated but is under study.

The corona probe test equipment consists of three basic units: an antenna, a radio frequency amplifier, and an indicator. A typical antenna is about 1 in (2.5 cm) long, surrounded by an insulating housing, and mounted on the end of a long insulating handle. The antenna is connected to the amplifier by a length of shielded lead. The amplifier is one of the usual type for radio frequencies, and must reject 60 Hz. The indicator may be earphones, an output meter, or preferably a cathode-ray oscilloscope. The corona probe is used to explore the insulation for areas of ionization while the winding is energized from a source of test voltage. The probe is potentially useful for determining the effect of the dielectric-stress gradients in the end-turn insulation; however, this is not a recommended test as it is very dangerous; a lights-out test is preferred and safer.

Partial discharge probe techniques are now well established for more precisely determining the location and intensity of slot discharges.

B.1.4 Stator Core Interlaminar Insulation Test (Loop Test)

A test coil may be wound around the stator core and excited to evaluate the quality of the interlaminar insulation. To induce flux in the stator core, equivalent to the operating flux at 105% of rated voltage and rated hertz. (For some equipment, this test should be performed at a flux level corresponding to 100% rated voltage to avoid damage to the core. The following data may be used:

The ampere-turns required may be approximated from Table A1.

The test-coil voltage, however, is a more significant criterion, as it will enable the actual volts per turn in the stator winding to be simulated. The test-coil voltage per turn is calculated as follows:

Voltage per turn of test coil =

where

Phase voltage = Line-to-line voltage for delta connection and line-to-line voltage for wye connection

Kd = Distribution factor = 0.955 for a three-phase wye or delta-connected stator winding (0.955 is a typical value used for this calculation)



Working Group Copy IEEE P56 Rev. 16 Draft Kp = Chord factor of the stator winding

N = Number of turns/phase in series of the stator winding

The number of turns for the test coil is determined by the above equation and the available voltage. The current capability of the cable to be used for the test is determined by dividing the required ampere-turns by the number of turns to be used. However, it should be noted that the calculations are estimations of the current that will be required to achieve rated flux in the stator core. The actual current may be greater or less than the calculated value. The test coil cable should be sized with this in mind.

When the coil is energized, hot spots due to short-circuited laminations will be apparent within a few minutes. Hot spots should not be detected by feel with the hand as this presents a serious shock hazard. Care should be exercised in entering the core of a large machine with the coil energized because of the hazard from voltage induced across large spans of laminations. It is advised that entering or exiting the stator bore area only occur when the coil is de-energized.

Induction Requirements

[Should insert a table here?]

Induction

Magnetization Force

Kilolines per Square Inch

Tesla

Ampere Turns, per Inch of Core Mean Periphery

Oersteds, per Centimeter of Core Mean Periphery 85 13.2 9 1.7 90 13.9 18 3.5 95 14.7 37 7.2 106 16.4 145



Working Group Copy IEEE P56 Rev. 16 Draft 28.2



Working Group Copy IEEE P56 Rev. 16 Draft Annex C (Informative)

A1. Prevention of Moisture Absorption in Winding Insulation of Rotating Machines

The following information describes suggested procedures to prevent moisture absorption in the windings of machines which are out of service. More detailed information may be found in IEEE Std 56-1958, (Reaff 1971)(ANSI C50.25-1972), Guide for Insulation Maintenance for Large AC Rotating Machinery.

A1.1 Machines which are out of service for prolonged periods may absorb sufficient moisture to reduce insulation resistance to a value below the recommended limits suggested in Section 9.

A1.1.1 This can be prevented if the winding temperature is always maintained slightly above the surrounding ambient air.

A1.1.2 The application of heat to the machine windings to keep the winding temperature about 5°C above the ambient is usually sufficient.

A1.1.3 In rooms where wide and rapid temperature changes are experienced, some higher winding temperature rise may be necessary.

A1.1.4 In rooms with limited air exchange, dehumidifiers will reduce dampness in storage.

A1.2 It may not be necessary to maintain heat on the winding continuously. Generally, moisture is more likely to be absorbed in the Spring and Summer months in most parts of the United States, while the low atmospheric humidity which prevails during the winter may itself remove moisture from the windings. Heat may not be required when the polarization index or the insulation resistance is considerably above the recommended minimum values given in Section 9.

A1.3 An approximation of the heat required to raise the winding temperature of an enclosed horizontal generator or motor 5°C above ambient temperature, where the machine is closed, is given by

H = DL/35 (H = DL/376) (Eq A1)

where

H = heat, in kilowatts

D = machine end-bell diameter, in feet (in meters)

L = machine stator length between end-bell centers, in feet (in meters).

A2. Removal of Moisture from Winding Insulation of Rotating Machines

A2.1 Electric machinery should, when necessary, be dried by circulating current in the windings or by heaters maintaining a reasonably constant temperature in the windings.



Working Group Copy IEEE P56 Rev. 16 Draft A2.1.1 Sufficient heat should be provided to produce a temperature on the end windings of not higher than 80°C by thermometer or 90°C by resistance temperature detector.

A2.1.2 For machines which have been flooded, a prolonged drying time is expected. Cycling of temperatures often accelerates dry-out in severe cases of flooding. A temperature in excess of 80°C by thermometer or 90°C by resistance temperature detector may be required for satisfactory drying at atmospheric pressure in a reasonable time. The use of higher temperatures should be made with caution and after consultation with the manufacturer of the equipment. It is not always possible to reclaim a flooded insulation system, especially on aged/delaminated windings

A2.1.3 Drying under a moderate vacuum is recommended, when possible, since this method reduces the maximum temperature required, and drying can usually be accomplished without exceeding the design-temperature values.

A2.1.4 The rate of temperature rise should be limited to 5°C per hour to avoid damage to the insulation by excessive gas or vapor formulation.

A2.2 The process of dry-out may also be observed by means of 1 minute insulation resistance readings, but the effect of temperature variations on insulation resistance readings is sufficient during the drying period. Observed readings should be corrected to 40°C (see Section 4.3, Figure 1) or readings should be taken at a constant winding temperature after the machine is heated.

A2.3.1 With the application of heat, the insulation resistance will initially drop and then will rise again over a period of time, finally approaching a constant value.

A2.3.2 Drying should be continued well beyond the time at which the insulation resistance has started to increase after reaching its minimum value and preferably until it approaches constancy (see Fig 3).

A3. Methods of Heating Machine Windings

Heat may be applied by any convenient method as long as it is safe and proper precautions are taken to prevent fire. The following methods may be considered, the choice between them being principally a matter of convenience, flexibility, cost, and availability. If the machine is enclosed to conserve heat, provision should be made to vent the moisture being removed.

A3.1 Electric Space Heaters

Electric space heaters will frequently be found to be the most suitable for heating since they are usually available in various sizes and can be placed in service quickly at low cost.

A3.1.1 The heaters should preferably be located in the air chambers under the machine and distributed so as to allow for even distribution so as to allow for even distribution of heat along the length of the machine.



Working Group Copy IEEE P56 Rev. 16 Draft A3.1.2 If the rotor is not in place, the ends of the machine may be closed with the end-bells or with large tarpaulins to reduce the heat loss.

A3.2 Field Winding Heating

The main field winding of the generator may be used to introduce heat into the machine when some source of direct current is available which can be separately controlled and allocated to the machine that requires heat. If the rotor of an alternating-current machine remains at stand-still, the current should be conducted to the slip rings through copper straps to prevent damage to the rings caused by pitting at the brush contacts.

A3.2.1 The field current required is

I = √ (1000H/R) (Eq A2)

where

I = field direct-current, in amperes

H = heat required, in kilowatts (see Eq A1)

R = resistance of field winding, in ohms at 25°C, measured at the slip rings

A3.2.2 Normally, about 15 percent of rated full-load field current will be required and the field-winding temperature rise should be 20°C or less above ambient temperature.

A3.3 Armature Winding Heating. The armature of the machine may be used to supply heat by passing current through the armature conductors.

A3.3.1 Alternating current at low potential may be induced in the armature, if the machine can be rotated at reduced speed (as on a hydro-generator), with the generator terminals short circuited or connected to a load. Careful control of speed and field current is required.

A3.3.2 Direct current may be passed through the armature conductors.

A3.4 Steam Heating

The use of steam heaters is not recommended because of the possibility of leakage condensing, damaging the insulation system.



Working Group Copy IEEE P56 Rev. 16 Draft Rotating Electric Machinery Insulation Condition Visual Inspection Appraisal – Visual Inspection Checklist

Component:

Deterioration or Damage:

Condition: NONE/ SLIGHT/ MAJOR

Armature windings (stationary ac, rotating dc)

(a) Puffy coils

(b) Soft insulations

(c) Girth cracking

(d) Separation in ground wall

(e) Bond cracks at slot ends

(f) Bond cracks into slots (wedge removed for inspection)

(g) Contamination of coil or connection surfaces (carbon dust, dirt, oil)

(h) Abrasion damage from chemicals, abrasives, or foreign materials

(i) Cracks/abrasion from mechanical forces, coil movement

(j) Loose bracing structure

(k) Corona damage (white, gray, or red deposits)

(l) Loose wedges or slot fillers

(m) Distorted windings, coils, or commutator risers

(n) Loose restraint bands

(o) Cracked restraint bands

(p) Uneven color commutator bars

(q) Evidence or occurrence of flashover

(r) Evidence of bar faults or band burning at walls of glass-band grooves




Field windings (rotating ac, stationary dc)

(a) Puffy coils

(b) Soft insulations

(c) Contamination of coil, collector, banding, or connection surfaces (carbon dust, dirt, oil)

(d) Abrasion damage from chemicals, abrasives, or foreign materials

(e) Cracks/abrasion from mechanical forces, coil movement

(f) Loose wedges

(g) Distortion of coils

(h) Shrinkage or looseness of coils, washers, or pads from poles

(i) Loose connections

(j) Heating of wedges

(k) Cracks in retaining rings

(l) Loose end-winding blocking

(m) Powdered insulations in air ducts

(n) Red oxide at metallic joints

(o) Loose collector or collector leads

Brush rigging

(a) Evidence or occurrence of flashover

(b) Carbonized leakage paths

(c) Loose parts

(d) Carbon dust accumulation




Cores

(a) Evidence or occurrence of rub or impact damage (rotor rub or objects in air gap)

(b) Burned punchings at bore surface

(c) Heating of adjacent punchings

(d) Loose or broken vent duct separators

(e) Core looseness

(f) Heating of end-finger plates

(g) Loose or broken laminations at clamping flanges

Insulated through bolts

(a) Contamination

(b) Looseness

Bearing insulation

(a) Cracks

(b) Distortion, evidence of excessive heating

(c) Oxidized or corroded conductors/strands

(d) Loose connections



Working Group Copy IEEE P56 Rev. 16 Draft New Annex - Thermosetting Resins Used In Insulation Systems

Polyesters are the polycondensation products of dicarboxylic acids with dihydroxy alcohols. Unsaturated polyester resins usually contain three essential components: the polyester (usually a very viscous liquid), the monomer and the inhibitor. A common monomer is styrene (a thin liquid). During the cure the monomer reacts with the unsaturated acid in the polyester chains to produce a cross-linked structure. As there is no volatile product void free structures can be produced. The inhibitor increases the shelf life but does not interfere with the subsequent polymerization when the mixture is heated.

The term epoxy implies a ring containing one oxygen and two carbon atoms. Four common types of epoxy resins are:

Resin Description

bisphenol-A Most widely used

epoxy-novolacs

High temperature applications

cycloaliphatics Good mechanical and dielectric properties at elevated temperature, good weatherability

aliphatic Flexibilizing resin

Epoxies are cured in one of two ways: • In catalytic curing the epoxy molecules react with each other, initiated by a catalyst. • In hardener initiated curing the hardener reacts with the epoxy and becomes part of the cured

material.



Working Group Copy IEEE P56 Rev. 16 Draft New Annex - STATOR CORE LOW ENERGY TEST

The condition of the interlaminar resistance between stator laminations (punchings) of a machine core is often best evaluated by means of magnetic excitation of the core. This procedure describes a method of accomplishing this using low flux densities, low power requirements, and short set-up time. The procedure has the further benefit of producing a permanent record of the condition of the interlaminar core insulation.

The principle underlying this method is that measurable currents will flow through failed or severely deteriorated interlaminar insulation when a flux of only a few percent of the rated value is induced in the core.

Discussion

A weak magnetic field (2-10% of nominal flux) is induced in the core using an excitation loop consisting of a few turns of small, low-voltage cable. The magnetic excitation field is in a circumferential pattern around the stator bore, and is to be the datum phase to which all other quantities are referenced. This excitation field induces currents to flow between laminations with weakened insulation. These resultant eddy currents due to the interlaminar insulation defects are detected using a Chattock or Rogowski-type pick-up coil, which is also known as Maxwell's worm. The Rogowski-type coil is constructed from many turns of fine wire wound on a flexible, U-shaped magnetic core. The number of turns per unit length and cross-sectional area of the core are kept constant so that a calibrated output from the coil can be obtained. For this reason, during operation, the tips of the coil should be maintained uniformly close to the core iron as when calibrated.

When such a coil is placed across two core teeth the voltage induced by the fault current is approximately proportional to the line integral along its length. If the field in the core is ignored, the voltage output of the coil is proportional to the eddy current flowing in the area encompassed by the pick-up coil, the two teeth it spans, and the core behind these. Unfortunately, due to the circumferential magnetic field component resulting from the excitation coil, the output of the Rogowski-type coil cannot be used directly to indicate the condition of the core insulation. However, the eddy currents due to the faults result in fluxes which are phase shifted with respect to the reference flux. Consequently, the component of the excitation flux measured by the Rogowski-type coil can be eliminated to produce a voltage that is proportional to the axial component of the eddy current. Elimination of the excitation flux component is enabled by placing a second reference coil to provide excitation voltage reference information. From this second coil, one can derive the zero crossing point of the excitation voltage, at which time the output from the Rogowski-type coil is recorded.

The outputs from the Rogowski-type and reference coils are fed to a signal processing unit which performs the excitation voltage component elimination and provides an output of the axial eddy currents detected by the Chattock-Rogowski-type coil in milliamperes. If the stator core insulation has been damaged, relatively high current readings( >100mA) will result.

Test Set-Up and Procedure



Working Group Copy IEEE P56 Rev. 16 Draft An excitation loop should be pulled through the bore and around the outside of the stator frame. One advantage of this test over traditional high power test is that the cable used for the excitation loop is low voltage and typically about 2-4 mm in diameter. The wires constituting the loop should be installed along the central axis of the bore, rather than letting the wires be in contact with the stator core. The core of the machine under test is excited with a weak magnetic flux, in range of 2-10% of nominal flux. Consequently, the exciting coil parameters (number of turns and cross-section) have to be calculated based upon the size of the core. Further, it is advisable to install a separate single turn coil to measure the actual induced flux.

Prior to commencement of the test the stator should be inspected for any conductive material which would short the laminations together and the Chattock-Rogowski-type coil should be adjusted to ride smoothly and freely on two teeth but should be prevented from wobbling or binding. Good practice should also include numbering the core teeth to provide an easy means of referencing any faults located. The reference coil is located in the bore and should be positioned so that its axis is perpendicular to the excitation field and so that it will not impede the progress of Rogowski-type coil as it is driven along the length of the core.

Once all of these requirements have been met the Rogowski-type coil can be set over the slot and the complete slot is scanned with the current readings being observed or recorded. This procedure is repeated with each slot in turn until the entire core, or a selected portion of it has been tested. The first slot to have been scanned should be retested for verification purposes.

Interpretation

This test has high sensitivity, hence it can detect magnetic disturbances which may not prejudice the reliability of the stator. Consequently, interpretation of the results is not simple and there may be some difficulty in determining an appropriate level of response which warrants further investigation and/or repair. In general, responses of greater than 100 mA (expected temperature rise for each 100 mA of fault current measured is 5-10 ˚C) should be regarded as significant faults and should be further investigated. Apart from absolute magnitude, some indication of the location of the fault may be obtained by examination of the polarity of the Rogowski-type coil response. This phenomenon results because faults lying within the span of the Rogowski-type coil give positive responses whereas those not encompassed by the span produce negative responses.

It should be recognized that no reading will be obtained at a fault location if the electrical circuit is not completed elsewhere, i.e. no electrical contact between laminations and building bars. However, it also should be noted that such a fault will not create a hot spot in normal operation due to the same reason.


merger of ieee 56 & 432 draft - iranelectrical...

Documents