IEEE STD 4000 (DP) PDF

IEEE Guide for Partial Discharge
Testing of Shielded Power Cable

Systems in a Field Environment
IEEE Power Engineering Society

Sponsored by the
Insulated Conductors Committee
IEEE
3 Park Avenue IEEE Std 400.3™-2006
New York, NY 10016-5997, USA
5 February 2007
Authorized licensed use limited to: University of Wisconsin. Downloaded on July 23,2012 at 05:47:11 UTC from IEEE Xplore. Restrictions apply.
IEEE Std 400.3TM-2006

Sponsor
Insulated Conductors Committee
of the
IEEE Power Engineering Society
Approved 15 September 2006

IEEE-SA Standards Board
Abstract: This guide covers the diagnostic testing of new or service-aged installed shielded
power cable systems, which include cable, joints, and terminations, using partial discharge (PD)
detection, measurement, and location. Partial discharge testing, which is a useful indicator of
insulation degradation, may be carried out on-line or off-line by means of an external voltage
source. This guide does not include the testing of compressed gas insulated systems or
continuous on-line monitoring at normal service voltage.
Keywords: cable system testing, cable testing, diagnostic testing, off-line partial discharge
testing, on-line partial discharge testing, partial discharge testing
_________________________
The Institute of Electrical and Electronics Engineers, Inc.

3 Park Avenue, New York, NY 10016-5997, USA
Copyright © 2007 by the Institute of Electrical and Electronics Engineers, Inc.

All rights reserved. Published 5 February 2007. Printed in the United States of America.
National Electrical Safety Code and NESC are registered trademarks in the U.S. Patent & Trademark Office, owned by
The Institute of Electrical and Electronics Engineers, Incorporated.
IEEE is a registered trademark in the U.S. Patent & Trademark Office, owned by the Institute of Electrical and Electronics
Engineers, Incorporated.
Print: ISBN 0-7381-5221-8 SH95571

PDF: ISBN 0-7381-5222-6 SS95571
No part of this publication may be reproduced in any form, in an electronic retrieval system or otherwise, without the prior
written permission of the publisher.
IEEE Standards documents are developed within the IEEE Societies and the Standards Coordinating Committees of
the IEEE Standards Association (IEEE-SA) Standards Board. The IEEE develops its standards through a consensus
development process, approved by the American National Standards Institute, which brings together volunteers
representing varied viewpoints and interests to achieve the final product. Volunteers are not necessarily members of the
Institute and serve without compensation. While the IEEE administers the process and establishes rules to promote
fairness in the consensus development process, the IEEE does not independently evaluate, test, or verify the accuracy
of any of the information contained in its standards.
Use of an IEEE Standard is wholly voluntary. The IEEE disclaims liability for any personal injury, property or other
damage, of any nature whatsoever, whether special, indirect, consequential, or compensatory, directly or indirectly
resulting from the publication, use of, or reliance upon this, or any other IEEE Standard document.
The IEEE does not warrant or represent the accuracy or content of the material contained herein, and expressly
disclaims any express or implied warranty, including any implied warranty of merchantability or fitness for a specific
purpose, or that the use of the material contained herein is free from patent infringement. IEEE Standards documents
are supplied “AS IS.”
The existence of an IEEE Standard does not imply that there are no other ways to produce, test, measure, purchase,
market, or provide other goods and services related to the scope of the IEEE Standard. Furthermore, the viewpoint
expressed at the time a standard is approved and issued is subject to change brought about through developments in the
state of the art and comments received from users of the standard. Every IEEE Standard is subjected to review at least
every five years for revision or reaffirmation. When a document is more than five years old and has not been
reaffirmed, it is reasonable to conclude that its contents, although still of some value, do not wholly reflect the present
state of the art. Users are cautioned to check to determine that they have the latest edition of any IEEE Standard.
In publishing and making this document available, the IEEE is not suggesting or rendering professional or other
services for, or on behalf of, any person or entity. Nor is the IEEE undertaking to perform any duty owed by any other
person or entity to another. Any person utilizing this, and any other IEEE Standards document, should rely upon the
advice of a competent professional in determining the exercise of reasonable care in any given circumstances.
Interpretations: Occasionally questions may arise regarding the meaning of portions of standards as they relate to
specific applications. When the need for interpretations is brought to the attention of IEEE, the Institute will initiate
action to prepare appropriate responses. Since IEEE Standards represent a consensus of concerned interests, it is
important to ensure that any interpretation has also received the concurrence of a balance of interests. For this reason,
IEEE and the members of its societies and Standards Coordinating Committees are not able to provide an instant
response to interpretation requests except in those cases where the matter has previously received formal consideration.
At lectures, symposia, seminars, or educational courses, an individual presenting information on IEEE standards shall
make it clear that his or her views should be considered the personal views of that individual rather than the formal
position, explanation, or interpretation of the IEEE.
Comments for revision of IEEE Standards are welcome from any interested party, regardless of membership affiliation
with IEEE. Suggestions for changes in documents should be in the form of a proposed change of text, together with
appropriate supporting comments. Comments on standards and requests for interpretations should be addressed to:
Secretary, IEEE-SA Standards Board

445 Hoes Lane
Piscataway, NJ 08854
USA
NOTE—Attention is called to the possibility that implementation of this standard may require use of subject matter
covered by patent rights. By publication of this standard, no position is taken with respect to the existence or validity of
any patent rights in connection therewith. The IEEE shall not be responsible for identifying patents for which a license
may be required by an IEEE standard or for conducting inquiries into the legal validity or scope of those patents that
are brought to its attention.
Authorization to photocopy portions of any individual standard for internal or personal use is granted by the Institute of
Electrical and Electronics Engineers, Inc., provided that the appropriate fee is paid to Copyright Clearance Center. To
arrange for payment of licensing fee, please contact Copyright Clearance Center, Customer Service, 222 Rosewood
Drive, Danvers, MA 01923 USA; +1 978 750 8400. Permission to photocopy portions of any individual standard for
educational classroom use can also be obtained through the Copyright Clearance Center.
Introduction
This introduction is not part of IEEE Std 400.3-2006, IEEE Guide for Partial Discharge Testing of Shielded
Power Cable Systems in a Field Environment.
Cable systems need to be tested after installation, periodically during their service life, occasionally after
frequent failures in specific types of cables or accessories, and whenever a decision needs to be made about
cable repair or replacement. The main purpose of testing is to provide a high degree of service reliability in
the most economic fashion. To guarantee optimum performance of the power cable system, standards and
guidelines have been developed that address the specific testing requirements for newly installed and
service-aged extruded and laminated dielectic insulation.
This guide was prepared by working group C-19W of the IEEE Insulated Conductors Committee.
Notice to users
Errata
Errata, if any, for this and all other standards can be accessed at the following URL: http://
standards.ieee.org/reading/ieee/updates/errata/index.html. Users are encouraged to check this URL for
errata periodically.
Interpretations
Current interpretations can be accessed at the following URL: http://standards.ieee.org/reading/ieee/interp/
index.html.
Patents
Attention is called to the possibility that implementation of this standard may require use of subject matter
covered by patent rights. By publication of this standard, no position is taken with respect to the existence
or validity of any patent rights in connection therewith. The IEEE shall not be responsible for identifying
patents or patent applications for which a license may be required to implement an IEEE standard or for
conducting inquiries into the legal validity or scope of those patents that are brought to its attention.
iv
Copyright © 2007 IEEE. All rights reserved.
Participants
At the time this guide was completed, the C-19W Working Group had the following membership:
Matthew S. Mashikian, Chair

Willem Boone, Vice Chair
Nezar Ahmed Craig Goodwin Ralph E. Patterson

Lawrence W. Bobb Edward Gulski Johannes Rickmann
Steven A. Boggs Richie Harp Dirk Russworm
Arvid J. Braun Stanley V. Heyer Lawrence Salberg
John R. Densley John Hinkle Albert Spear
Swapan K. Dey William Larzelere William A. Thue
Rachel I. Mosier
The following members of the individual balloting committee voted on this guide. Balloters may have
voted for approval, disapproval, or abstention.
Roy W. Alexander Edgar O. Galyon William M. McDermid

Butch Anton R. B. Gear Jr. Nigel P. McQuin
Ali Al Awazi David L. Gilmer John E. Merando Jr.
Saber Azizi-Ghannad Steven N. Graham Gary L. Michel
Michael P. Baldwin Randall C. Groves Rachel I. Mosier
Earle C. Bascom III Ajit K. Gwal Jerry R. Murphy
Martin Baur Adrienne M. Hendrickson Shantanu Nandi
Michael G. Bayer Lauri J. Hiivala Michael S. Newman
Steven A. Boggs Ajit K. Hiranandani Joe W. Nims
Kenneth E. Bow David A. Horvath Ralph E. Patterson
Chris Brooks Dennis Horwitz Allan D. St. Peter
Kent W. Brown David W. Jackson Johannes Rickmann
William A. Byrd A. S. Jones Michael A. Roberts
Jim Y. Cai James H. Jones Robert L. Seitz
Mark S. Clark Gael Kennedy David Singleton
John H. Cooper Joseph L. Koepfinger Michael J. Smalley
Tommy P. Cooper Jim Kulchisky James E. Smith
Jorge E. Fernandez Daher Saumen K. Kundu S. Thamilarasan
J. P. Disciullo Thomas W. La Rose James E. Timperley
Gary L. Donner Solomon Lee Joseph J. Vaschak
Randall L. Dotson Maurice Linker Waldemar G. Von Miller
Dana S. Dufield William E. Lockley Edward E. Walcott
Donald G. Dunn Lisardo Lourido Carl T. Wall
Gary R. Engmann William Lumpkins Mark D. Walton
Robert B. Fisher G. L. Luri William D. Wilkens
Rabiz N. Foda Glenn J. Luzzi James W. Wilson Jr.
Marcel Fortin Eric P. Marsden Luis E. Zambrano
Carl J. Fredericks Matthew S. Mashikan Donald W. Zipse
v
When the IEEE-SA Standards Board approved this application guide on 15 September 2006, it had the
following membership:
Steve M. Mills, Chair
Richard H. Hulett, Vice Chair
Don Wright, Past Chair
Judith Gorman, Secretary
Mark D. Bowman William B. Hopf T. W. Olsen

Dennis B. Brophy Lowell G. Johnson Glenn Parsons
Joseph Bruder Herman Koch Ronald C. Petersen
Richard Cox Joseph L. Koepfinger* Gary S. Robinson
Bob Davis David J. Law Frank Stone
Julian Forster* Daleep C. Mohla Malcolm V. Thaden
Joanna N. Guenin Paul Nikolich Richard L. Townsend
Mark S. Halpin Joe D. Watson
Raymond Hapeman Howard L. Wolfman
*Member Emeritus
Also included are the following nonvoting IEEE-SA Standards Board liaisons:
Satish K. Aggarwal, NRC Representative

Richard DeBlasio, DOE Representative
Alan H. Cookson, NIST Representative
Catherine Berger
IEEE Standards Project Editor
Angela Ortiz
IEEE Standards Program Manager, Technical Program Development
vi
Contents
1. Overview .................................................................................................................................................... 1
1.1 Scope ................................................................................................................................................... 2

1.2 Purpose ................................................................................................................................................ 2
1.3 Background.......................................................................................................................................... 2
2. Normative references.................................................................................................................................. 3
3. Definitions, acronyms, and abbreviations .................................................................................................. 3
3.1 Definitions ........................................................................................................................................... 3

3.2 Acronyms and abbreviations ............................................................................................................... 4
4. Partial discharge data interpretation ........................................................................................................... 5
4.1 Type and location of defects................................................................................................................ 8

4.2 Insulating materials............................................................................................................................ 12
4.3 Operating conditions.......................................................................................................................... 12
5. Partial discharge detection........................................................................................................................ 13
5.1 General test setup............................................................................................................................... 14

5.2 Partial discharge detection methods .................................................................................................. 15
5.3 Partial discharge detection sensitivity................................................................................................ 16
5.4 Partial discharge location................................................................................................................... 16
5.5 Test limitations .................................................................................................................................. 19
6. Voltage sources ........................................................................................................................................ 20
6.1 Sinusoidal power-frequency voltage sources..................................................................................... 20

6.2 Alternative voltage sources................................................................................................................ 21
7. Practical testing guidelines ....................................................................................................................... 22
7.1 Pre-test owner information ................................................................................................................ 22

7.2 Preliminary inspection and/or testing ................................................................................................ 23
7.3 Pre-test conference ............................................................................................................................ 23
7.4 Test conditions (voltage levels and duration) .................................................................................... 23
8. Test results and recommendations............................................................................................................ 24
8.1 Interpretation of test results—recommendations and implications.................................................... 24

8.2 Partial discharge test documentation ................................................................................................. 26
vii
9. Safety........................................................................................................................................................ 28
9.1 Personal protective equipment........................................................................................................... 28

9.2 Switching and blocking practices ...................................................................................................... 28
9.3 Grounding and responsibility for control of cable ............................................................................. 28
9.4 Safety of the public............................................................................................................................ 29
9.5 Daily safety meeting .......................................................................................................................... 29
10. Conclusions ............................................................................................................................................ 29
Annex A (informative) Partial discharge and water trees............................................................................. 30
Annex B (informative) Effect of cavity shape on partial discharge inception voltage ................................. 31
Annex C (normative) Partial discharge calibration and location accuracy................................................... 32
Annex D (informative) Bibliography ........................................................................................................... 34
viii
1. Overview
This guide is one of a series of guides introduced in IEEE Std 400TM-2001.1 It is divided into 10 clauses, as
follows:
⎯ Clause 1 provides the scope and the purpose of the guide.
⎯ Clause 2 lists the references to other standards or guides that are useful in applying this guide.
⎯ Clause 3 provides definitions that are of particular importance for the understanding of this
guide and lists a glossary of abbreviations and acronyms.
⎯ Clause 4 provides basic interpretation of partial discharge (PD) data.
⎯ Clause 5 describes the different types of PD detection methods.
⎯ Clause 6 briefly describes the voltage sources available commercially.
⎯ Clause 7 offers useful practical testing guidelines.
⎯ Clause 8 discusses the expected test results and recommendations.
⎯ Clause 9 is devoted to the safety precautions to be observed during testing.
⎯ Clause 10 provides conclusions.
This guide also contains four annexes. Annex A expounds, in some detail, on the question of water treeing
and partial discharge in cables with extruded dielectrics. Annex B shows how the partial discharge
inception voltage changes with the cavity shape, location, and material. Annex C discusses in greater detail
the subjects of PD calibration and location accuracy. Annex D is the bibliography.
1
Information on references can be found in Clause 2.
1
IEEE Std 400.3-2006
IEEE Guide for Partial Discharge Testing of Shielded Power Cable Systems in a Field Environment
1.1 Scope
This guide covers the diagnostic testing of new or service-aged installed shielded power cable systems,
which include cable, joints, and terminations, using PD detection, measurement, and location. Partial
discharge testing, which is a useful indicator of insulation degradation, may be carried out on-line or by
means of an external voltage source. This guide does not include the testing of compressed gas insulated
systems or continuous on-line monitoring at normal service voltage.
1.2 Purpose
This guide describes diagnostic methods capable of detecting and locating partial discharges from defects
and damage in installed shielded power cable systems. The results of PD tests are used to assess the
condition of cables and accessories.
1.3 Background
PDs are small electric sparks or discharges that occur in defects in the insulation, or at interfaces or
surfaces, or between a conductor and a floating metal component (not connected electrically to the high-
voltage conductor nor to the ground conductor), or between floating metal components if the electric field
is high enough to cause ionization of the gaseous medium in which the components are located. The
discharges do not bridge the insulation between conductors, and the defects may be entirely within the
insulation, along interfaces between insulating materials (e.g., at accessories) or along surfaces
(terminations or potheads).
Partial discharge characteristics depend on the type, size, and location of the defects, insulating material,
applied voltage, and cable temperature, and they vary with time. The damage caused by PD depends on
several factors and can range from negligible to causing failure within days to years.
Advances in digital (electrical) measurement technology, both in the time and frequency domains, have
improved the sensitivity of PD measurements. This has led to an increasing number of PD measurements
on cable systems, particularly on medium-voltage systems. The purpose of such measurements is to assess
the current condition of a cable circuit. At the current state-of-the-art, very good cables and very bad cables
can generally be identified. It is the remaining life of the cables between these two extremes that cannot be
predicted with great accuracy. As well, this technology cannot determine with complete confidence that a
specific cable is in very good condition with essentially no probability of failure in the near future, as
failure can be caused by phenomena that do not generate PD. However, the PD measurement can, at times,
predict with a high level of confidence that a given cable is in very poor condition and is likely to fail in the
near future.
This guide provides background information on PD detection and location techniques for users of PD
testing services of cables with laminated or extruded insulations, and it provides background information
on the interpretation of PD data. In this guide, cable may also refer to a cable system that includes cables
and accessories.
2
IEEE Std 400.3-2006
2. Normative references
The following referenced documents are indispensable for the application of this document. For dated
references, only the edition cited applies. For undated references, the latest edition of the referenced
document (including any amendments or corrigenda) applies.
Accredited Standards Committee C-2, National Electrical Safety Code® (NESC®).2
ICEA T-24-380-1994, Guide for Partial-Discharge Test Procedure.3
IEC 60270, High-voltage test techniques—Partial discharge measurements.4
IEC 60885-2, 1987, Electrical test methods for electric cables—Part 2: Partial discharge tests.
IEC 60885-3, 1987, Electrical test methods for electric cables—Part 3: Test methods for partial discharge
measurements on lengths of extruded power cables.
IEEE Std 400TM-2001, IEEE Guide for Field Testing and Evaluation of the Insulation of Shielded Power
Cable Systems.5, 6
IEEE Std 400.2TM-2004, IEEE Guide for Field Testing of Shielded Power Cable Systems Using Very Low
Frequency (VLF).
IEEE Std 510TM, IEEE Recommended Practices for Safety in High-Voltage and High Power Testing.7
NFPA-70E, Standard for Electrical Safety Requirements for Employee Workplaces.8
3. Definitions, acronyms, and abbreviations
3.1 Definitions
For the purposes of this guide, the following terms and definitions apply. The Authoritative Dictionary of
IEEE Standards Terms [B18]9 should be referenced for terms not defined in this clause.
2
ANSI publications are available from the Sales Department, American National Standards Institute, 25 West 43rd Street, 4th Floor,
New York, NY 10036, USA (http://www.ansi.org/).
3
ICEA publications are available from the Insulated Cable Engineers Association, P.O. Box 20048, Minneapolis, MN 55420, USA
(http://www.icea.org/).
4
IEC publications are available from the Sales Department of the International Electrotechnical Commission, Case Postale 131, 3, rue
de Varembé, CH-1211, Genève 20, Switzerland/Suisse (http://www.iec.ch/). IEC publications are also available in the United States
from the Sales Department, American National Standards Institute, 11 West 42nd Street, 13th Floor, New York, NY 10036, USA.
5
IEEE publications are available from the Institute of Electrical and Electronics Engineers, 445 Hoes Lane, Piscataway, NJ 08855,
USA (http://standards.ieee.org/).
6
The IEEE standards or products referred to in this clause are trademarks of the Institute of Electrical and Electronics Engineers, Inc.
7
IEEE Std 510-1983 has been withdrawn; however, copies can be obtained from Global Engineering, 15 Inverness Way East,
Englewood, CO 80112-5704, USA, tel. (303) 792-2181 (http://global.ihs.com/).
8
NFPA publications are available from Publication Sales, National Fire Protection Association, 1 Batterymarch Park, P.O. Box 9101,
Quincy, MA 02269-9101, USA (http://nfpa.org/codes/index.html).
9
The numbers in brackets correspond to those of the bibliography in Annex D.
3
IEEE Std 400.3-2006
3.1.1 electrical trees: Tree-like growths consisting of non-solid or carbonized microchannels, which can
occur at electric field enhancements such as protrusions, contaminants, voids, or water trees subjected to
electrical stress for extended time periods. Partial discharges are responsible for electrical tree growth.
3.1.2 extruded insulation: Insulation such as high-molecular-weight polyethylene (HMWPE), crosslinked

polyethylene (XLPE), tree-retardant XLPE (TRXLPE), ethylene propylene rubber (EPR), etc. applied using
an extrusion process.
3.1.3 laminated insulation: Insulation formed in layers typically from tapes of either cellulose paper or
polypropylene or a combination of the two. An example is the PILC (paper insulated lead covered) cable
design.
3.1.4 partial discharge (PD) pulse: A current or voltage pulse that results from a partial discharge. In a
cable, the pulse propagates away from the PD source in both directions.
3.1.5 pulse sequence: The occurrence of partial discharge (PD) pulses as a function of time that can be
used to determine statistical correlations such as the amplitude of one pulse and the time to the next pulse.
3.1.6 water trees: A tree-shaped collection of water-filled micro voids that are connected by oxidized
tracks. Water trees can occur at electric field enhancements such as protrusions, contaminants, or voids in
polymeric materials subjected to electrical stress and always in the presence of water.
3.2 Acronyms and abbreviations
CAIFI customer average interruption frequency index

DAC damped ac voltage
Ed stress in the insulation adjacent to the cavity
EPR ethylene propylene rubber
FTRC frequency tuned resonant circuit
HMWPE high-molecular-weight polyethylene
ITRC inductively tuned resonant circuits
MAIFI momentary average interruption frequency index
n partial discharge pulse repetition rate
PD partial discharge
PDEV partial discharge extinction voltage, Ve
PDIV partial discharge inception voltage, Vi
PILC paper insulated lead covered
PPE personal protective equipment
q apparent discharge magnitude
SAIDI system average interruption frequency index
SOV system operating voltage
T period of the test voltage
ti time measured from the preceding positive going transition of the sinusoidal test voltage
through zero to the PD pulse
TDR time domain reflectometry
TRXLPE tree-retardant crosslinked polyethylene
Vo nominal operating voltage to ground
VLF very low frequency
XLPE crosslinked polyethylene
Φ phase angle of PD pulse given by Φi = 360 (ti/T)
4
IEEE Std 400.3-2006
4. Partial discharge data interpretation

Partial discharges are a consequence of local breakdown either as a result of (a) an electric field
enhancement within or on the surface of the insulation, or (b) a region of low breakdown field. PDs appear
as individual events of very short duration; are always accompanied by emissions of light, sound and heat,
as well as electromagnetic pulses; and often result in chemical reactions.
The PD parameters that are usually measured during tests on installed cable systems are as follows:
⎯ PD inception voltage, PDIV (off-line tests).
⎯ PD extinction voltage, PDEV (off-line tests).
⎯ PD location.
⎯ PD magnitude (q).
⎯ PD repetition rate (n).
⎯ PD density—the density of discharges measured per unit of time and per unit of length (pC/m.s)
(laminated cable only).
⎯ Phase angle of PD pulse (Φ) given by Φi = 360 (ti/T), where ti is the time measured from the
preceding positive-going transition of the sinusoidal test voltage through zero to the PD pulse,
and T is the period of the test voltage.
⎯ Phase resolved PD plot (n vs. Φ vs. q).
⎯ PD magnitude vs. voltage plot (q vs. V), (off-line tests).
The characteristics of the PD parameters depend on

⎯ Type and location of defects, i.e., PD sources, in the insulation system.
⎯ Insulating material.
⎯ Operating conditions such as applied voltage, load, and time.
⎯ On-line tests can measure q, n, and Φ at operating temperature, whereas off-line tests are
performed on cable systems that have cooled down.
Each of these characteristics affects the interpretation of PD data and will be discussed separately in 4.1,
4.2, and 4.3.
Water treeing is an important form of degradation that can afflict older HMWPE and XLPE extruded
cables. At the site of a water tree, the insulation is degraded, i.e., has a higher dielectric constant and lower
dielectric strength than the original insulation. The water tree growth under service conditions is a very
slow process, and usually it takes many years to completely penetrate the insulation. Water trees do not
generate partial discharge. However, water trees can lead to electrical trees when subjected to high
electrical stresses as a result of a lightning impulse, a switching or dc overvoltage, a high ac voltage, or
when the tip of the water tree approaches a conductor or insulation shields. There is no evidence of water
treeing being an important issue with EPR or TRXLPE cables.
A more comprehensive treatment of the subject of water trees and PD is provided in Annex A.
The existence of PD does not indicate the likelihood of the PD to cause failure without additional
information concerning the source of the PD. Although discharge-free cables manufactured since 1994
5
IEEE Std 400.3-2006
have less than 5 pC at 8 kV/mm (200 volts/mil), cables made prior to 1987 may have partial discharge
above 5 pC as allowed by industry specifications (AEIC) at the time of manufacturing. Measurement of
partial discharge in the field above 1 Vo, where Vo is the nominal operating voltage to ground of the cable
system, does not necessarily indicate that these cables have developed partial discharge due to in-service
aging.
The detected apparent discharge magnitude, q, may or may not play a significant role in determining the
severity of the defect, particularly when comparing different defects. For example, a few picoCoulombs
detected from an electrical tree in an extruded insulation may require an immediate repair or replacement;
however, thousands of picoCoulombs of partial discharge between the cable neutral and ground shield is
tolerable in that same extruded insulation. If the defect type is known, the PD magnitude, along with other
parameters, may give a general indication of the condition of a defect in both extruded and fluid
impregnated cables. There is usually a statistical variation in the PD magnitudes when measurements are
made over a short time for a particular defect site. Variations of PD magnitude of up to 100% for a
particular discharge site are neither uncommon nor usually significant.
Numerous studies have attempted to correlate PD characteristics with various defect types in extruded
cables and accessories to determine whether a particular type of defect has unique PD characteristics
(Kreuger [B20], Suzuki and Endoh [B29], Gulski [B13]). Earlier studies tried to relate quantities such as
discharge magnitude and discharge power to damage caused by PD without success. However, phase-
resolved PD plots, which provide more complete characterization of the PD, have only become available in
the last five to ten years. Much of this later work is proprietary as it is used by PD testing companies to
determine the type of defect, the amount of damage caused by PD at the defect, and to assess the condition
of the cable system. Experience developed by correlating PD data with physical examination or dissection
of cable PD sites forms the basis of the know-how used to make PD severity assessments. The voltage Vi at
which a discharge is initiated (PDIV) and the voltage Ve at which a discharge is extinguished (PDEV) are
critical values for extruded cables. The PDEV can be significantly less than the PDIV. In practice, the
extinction voltage Ve is often the lowest voltage at which the magnitude of a PD pulse becomes equal to or
less than the sensitivity of the measuring equipment, and the inception voltage Vi is often the lowest
voltage at which the magnitude of a PD pulse quantity becomes equal to or exceeds the sensitivity of the
measuring equipment..
For laminated cables and their accessories, where PD can be tolerated for a long time due to the high PD
resistance of impregnated paper, the use of so-called guidelines (van Schaik et al. [B31]) can be an
effective tool to improve data interpretation. These guidelines describe the typical behavior of materials or
a combination of different materials under a discharge regime. For some components, 100 pC or 200 pC
may already be a dangerous level, for example, if carbonized tracks have developed, whereas for others,
104 pC or more will probably not affect the future performance. These guidelines form the basis of
“objective” recommendation criteria (van Schaik et al. [B31]).
There is a significant level of uncertainty in the interpretation of PD data with respect to predicting the
future performance of cable systems due to the statistical behavior of PD:
⎯ Some data are obtained from accelerated aging tests in the laboratory. There is always a
question of the validity of accelerated aging test data to predict the performance under normal
service conditions. This introduces some uncertainty in the application of accelerated aging data
to field test data.
⎯ Aging tests are usually performed on a limited number of test cables and accessories, with or
without artificial defects, so that not all types of cables, accessories, and defects are tested.
⎯ There is usually a significant variation in the PD characteristics among similar test cables and
accessories, with or without artificial defects, which have been subjected to the same aging test
conditions. For example, it is not unusual in tests on five or more identical test samples to have
an order-of-magnitude variation in the time to failure.
6
IEEE Std 400.3-2006
Apart from the scatter in the PD characteristics for “identical” defects in the same cable system, there is
further variation in the PD characteristics with time under voltage.
The precision in interpretation of PD data varies with the defect type; for example, it is usually easier to
predict the damage and assess the condition of a cable for an electrical tree than for a cavity or void.
As a general rule, the accuracy in interpretation is good when testing “very good” or “very bad” cable
systems.
⎯ Very good cable systems will exhibit low levels of PD activity; for example, the PDIV will be
typically greater than 2 Vo. This, of course, assumes that the background noise is low and that
the sensitivity of the measurements is good (better than 5 pC or equivalent mV reading). It
should be noted that a cable system with a measured PDIV above 2 Vo does not guarantee a
long life. For example, a lightning surge could initiate an electrical tree and cause rapid failure
or it could initiate an otherwise intermittent PD that was extinguished at the time of the test. The
probability of these conditions occurring is small but not zero.
⎯ Very bad cable systems will typically exhibit a low PDIV with well-defined PD characteristics
such as easily recognizable phase-resolved PD patterns similar to those known to occur at
defects that degrade the cable system (e.g., improperly installed accessories). It should be noted
that recognizing phase-resolved patterns is not always accurate and the wrong identification of
defects can sometimes occur. However, these occurrences should decrease as more experience
is gained.
The accuracy in interpretation is lower for cables between the “very good” and “very bad” conditions. The
accuracy is often affected by the sensitivity of the measurement, which varies from location to location and
also with time. For example “false positives” can occur when the PD detected is incorrectly identified as
occurring within the cable system when it is actually originating from an external source, or the defect
causing the PD has been incorrectly identified due to too high or too low measurement sensitivity and/or
high background noise. “False negatives” can occur if PD signals were missed, for example, by insufficient
sensitivity of the detection system, high background noise, operator error, or high pulse attenuation along
the cable system. As a PD pulse propagates along a cable, it suffers increasing attenuation with frequency,
which results in an increasing width of the PD pulse and decreasing amplitude as a function of the distance
propagated away from the PD source. However, the total area under the pulse, which is proportional to
charge, remains relatively constant. Repeating the measurements on the same cable circuits can sometimes
reduce the risk of incorrect interpretation.
The accuracy in interpretation can be increased if the following criteria are met:
⎯ All relevant information about the cable system is known. This includes the age and type of
cable(s) under test (insulation type, presence of jacket, cable design, etc.), the number and types
of accessories in the circuit of the cables, and the operating conditions (duct, direct buried, wet,
dry or both, incidence of surges, load conditions, previous failure behavior, etc.).
⎯ Cable dissections are carried out on failed cable. This will help in the correlation between
possible failure mechanisms and the measured PD characteristics.
⎯ Additional testing and eventual dissection are performed on cables and accessories removed
from service to establish correlations between PD characteristics and cable system performance
such as time to failure.
⎯ As more data are collected on cable systems in service, better correlations between PD
measurements and future performance may be established.
⎯ Measurements are repeated periodically to obtain trends in the data. In this way the rates of
change of the PD characteristics can be determined, for example, changes in the phase-resolved
PD patterns, discharge magnitudes, PDIV, and so on.
7
IEEE Std 400.3-2006
⎯ Standardized test and analysis procedures are developed. When standardized test and analysis
procedures have been developed, these will aid the comparison of databases of different PD
service providers and also from different utilities.
Some PD testing companies store their field test data in a data bank that classifies PD characteristics
according to the different defects in order to establish “knowledge rules.” As more tests are performed,
more data are accumulated and can be compared with the future performance of the cable system. This
builds up a library of data and should gradually improve the accuracy in the interpretation of PD
measurements.
Thus, with the current level of knowledge of the interpretation of PD data, the accuracy in estimating the
future performance of a cable system varies according to the cable condition. As mentioned, the accuracy is
reasonably good for very good or very bad cables, but it is lower for cable systems between these
conditions. The accuracy can be further improved as more data are accumulated. Greater confidence in the
accuracy in the data interpretation on particular cable circuits can be gained if periodic measurements are
made so that trends in the data can be established. When properly implemented PD testing is very useful to
rank cables, i.e., to determine whether one cable will perform better than another and to prioritize cable
replacement.
4.1 Type and location of defects
The type and location of defects affect the interpretation of PD data. It is well known that defects in cable
systems can result in premature failure, so it is important to be able to locate the defects and to determine
the type of defect. Defects may be inadvertently introduced into cables or accessories during manufacture,
storage, handling, transportation, and installation, or they may develop as the cable system ages in service
while exposed to electrical, mechanical, thermal, and environmental stresses. Some defects do not produce
partial discharges. Typically, defects will cause PD only if the local electric stress at the defect exceeds the
inception stress.
4.1.1 Typical partial discharge sources in extruded cables
Typical defects in extruded cable systems that can be sources of PD are as follows:
⎯ Voids or cavities within the insulation or at interfaces between the insulation and the
semiconducting shields. This includes knife cuts, gouges, or cracks in the insulation shield.
⎯ Interfacial cavities in cable and accessory interfaces.
⎯ High-resistance insulation shield or broken neutral.
⎯ Electrical trees initiated from protrusions, voids, or water trees.
Several PD characteristics are usually used to assess the severity of a PD source. These characteristics
include, but are not limited to, detailed measurements of its statistical properties (discharge magnitude,
repetition rate, phase, etc.) and inception and extinction voltages.
4.1.1.1 Voids/cavities within the insulation or at interfaces between the insulation and the
semiconducting shields
The PDIV of a void or cavity will depend on the shape, size, and position of the cavity within the
insulation, and the type of gas and its pressure inside the cavity. Annex B shows examples of how the shape
of a cavity affects the magnitude of the electrical stress in the cavity. The PDIV for a spherical cavity in
XLPE is about twice that for a flat cavity, for example, at a loose insulation shield and almost three times
the value for the same cavity shapes in EPR insulation. If the flat cavity is aligned in the direction of the
applied electrical stress, i.e., usually radially, the PDIV will be between 20% and 30% greater than that for
a spherical cavity of similar size and location.
8
IEEE Std 400.3-2006
Both the PD magnitude and the repetition rate increase with cavity surface area. Multiple discharges can
occur within the same cavity if the surface area is large; with time, however, as the conductivity of the
cavity surfaces increase, the multiple discharges can merge into a smaller number of larger magnitude
pulses. The PD magnitude will also increase with cavity depth. Thus, any thermal expansion and
contraction with cable loading that affects the cavity dimensions will affect the PD characteristics (e.g.,
PDIV, PDEV, q, and n). In smaller diameter spherical cavities, the discharge may cease as the walls of the
cavity become conductive or if the pressure of the gas in the cavity increases. Raising the test voltage will
re-ignite the discharges. PDs from cavities are identified through phase angle, pulse sequence, and pulse
height distribution data.
As the insulation thickness increases for cables with larger voltage ratings, the discharge magnitude in a
cavity of the same depth will decrease at the PDIV for a particular cable. For example, a 0.5 mm spherical
cavity at the conductor shield in a 15 kV cable will have a magnitude of about 10 pC, but about 5 pC in a
35 kV cable. Typical magnitudes in voids in extruded insulation systems are less than 100 pC for test
voltages up to twice the operating voltage.
Cavities, usually flat, can also occur at interfaces, such as the insulation/semiconducting shields. At
voltages well above the PDIV, discharge in such flat cavities may consist of a large number of spatially
distributed discharges. With wideband detection in such a situation, the individual discharges can be seen to
occur almost simultaneously. With conventional low-frequency, narrow-bandwidth PD detectors, the
individual discharges cannot be distinguished, and pulse superposition may take place, resulting in a much
greater measured PD magnitude and much lower PD repetition rate than actually occurs. Cavities adjacent
to a semiconducting shield or a conductor usually have larger magnitude discharges in one half cycle of the
applied voltage than the other.
The usual progression of damage in cavities due to PD is the gradual erosion of the surfaces, the creation of
pits, and the eventual formation of electrical trees, which grow relatively rapidly to complete the failure.
The approximate size or the stage of progression can sometimes be determined from the PD magnitude and
repetition.
4.1.1.2 Cavities between cable and accessory interfaces (interfacial cavities)
Cavities between the insulation interfaces of cables and their accessories (i.e., interfacial cavities) may
occur if there is bad workmanship while removing the insulation shielding, insufficient pressure to maintain
good contact, or if silicone grease has migrated from the interface. The cable loading could affect the
pressure and thus the cavity dimensions and PD characteristics (PDIV, PDEV, q, and n).
Interfacial discharges or tracking in cavities located at insulation interfaces between the cable and its
accessories are often identified from their pulse height distribution, which tends to progress through three
stages. At early stages, both PD magnitude and repetition rate are small. As the tracking progresses, both
the magnitude and the repetition rate increase. During the final pre-failure stage, the PD magnitude will
decrease while the PD repetition rate increases significantly. Therefore, the severity of the tracking is
determined by examining the data of both the PD magnitude and the repetition rate.
4.1.1.3 High-resistance insulation shield or broken neutral
The resistivity of the semiconducting insulation shield of an unjacketed extruded cable can increase
significantly if the shield becomes contaminated with an organic liquid such as transformer oil, gasoline, or
the preservatives used in some wood utility poles. Such contamination is most likely to occur at the base of
a utility pole, where the cable enters the ground or in industrial chemical plants. A high resistivity
insulation shield causes an increase in the electric stress at the neutral wires/tapes, so that PD can occur
between the neutral wires/tapes and the outer insulation semiconducting shield. This type of PD is
identified through the pulse repetition rate, phase angle, dominant polarity, pulse sequence, and pulse
height distribution with magnitudes often in the range of hundreds to thousands of pC. Such discharges,
although of large magnitudes, seldom result in cable failure. Therefore, trending is beneficial for this type
9
IEEE Std 400.3-2006
of defect. External PD can also occur if the neutral wires are broken due to corrosion; in which case, a
continuity test of the cable neutral may also be performed. Damage to the cable due to external discharges
of this type has been reported (Abdolall [B1]). However, such PD is not likely to occur for a jacketed cable,
unless the integrity of the jacket has been compromised.
4.1.1.4 Electrical trees initiated from protrusions, voids, or water trees
Protrusions or contaminants within the insulation or at the insulation/semicon interface produce local
increases in the electrical stress. If the stress becomes sufficiently large, an electrical tree can be initiated.
At the site of an electrical tree, the insulation is damaged irreversibly, partial discharge may be present, and
complete insulation breakdown is only a question of time for non-laminated insulation. In general,
electrical trees are more difficult to initiate than to grow, so that an electrical tree, once initiated, tends to
grow to failure by partial discharge whenever the PD inception voltage is exceeded. The PD magnitudes
from an electrical tree depend mainly on the length of the tree in the direction of the electric field and the
magnitude of the applied electric field. In some electrical trees the PD magnitudes increase gradually with
the increasing length of the tree channels. These discharge magnitudes are larger than those found in
spherical cavities. In other electrical trees the magnitudes stay relatively constant but the repetition rate
changes as the tree grows.
Cavities generally generate electrical trees prior to failure. An electrical tree, once initiated, often grows
rapidly if partial discharges persist, leading to a failure in a relatively short time ranging from minutes to
some weeks, depending on the insulation material and operating conditions. A positive detection of an
electrical tree should require immediate replacement or repair.
Although partial discharges do not occur during the initiation and growth of water trees, partial discharges
do occur when a water tree leads to an electrical tree. The likelihood of causing a water tree to lead to an
electrical tree during a field PD test increases with the test voltage magnitude and the test duration. Any
partial discharge at a water tree implies the existence of one or more electrical trees at that water tree. An
electrical tree initiated from a water tree may be either from an overvoltage such as lightning, a switching
surge, or a field test. Alternatively, at operating voltage, the properties of the insulation surrounding the
water tree may be such that it leads to the initiation of an electrical tree. Immediate failure of the cable
circuit may not result, but failure can occur several hours to several months after electrical tree initiation.
4.1.2 Typical partial discharge sources in fluid-impregnated cables
Fluid-impregnated [paper insulated lead covered (PILC)] cables are more resistant to PD than XLPE
cables. Typical sources of PD are as follows:
⎯ Fluid-deficient butt gaps, soft areas of the insulation, and voids or cavities due to poor
impregnation.
⎯ Dry, brittle, and cracked paper.
⎯ Waxing of the fluid.(wax formation due to degradation hinders fluid flow that can lead to void
formation).
⎯ Sites of carbonized tracks (e.g., treeing).
⎯ Water in the insulation.
⎯ Leakage of fluid at gaskets or due to holes in the sheath.
Cavities in fluid-impregnated (laminated) insulation can be formed by expansion of the lead sheath during
repeated loading, thereby increasing the volume of the cable, waxing of the fluid due to aging, holes in the
sheath that allow the egress of fluid, or leakage of fluid at gaskets. PD in gas-filled bubbles will generate
additional gases changing the shape of the bubble, which in turn affect the PD characteristics. Thus, it is not
unusual for PD in laminated insulation to move from one location to another due to the transient behavior
10
IEEE Std 400.3-2006
of gas bubbles. Repeated discharges in one location will cause carbonization of the paper and the initiation
of tracking. The tracking may follow the interfaces between paper layers, cross the butt gaps, and thus
extend axially along the cable.
In belted cables, mainly for the lower class voltages, where the cores are not individually screened as they
are in the higher voltage cables, the area between the three cores becomes a high risk for deterioration due
to PD. This happens particularly during heavy loading, which may cause core separation. Another problem
with this type of belted cable at higher voltages is the high tangential stress causing discharge activity
between the layers and finally resulting in a failure. Typically, the application of individual screening of the
cores (Hochstaedter layer or H-type cable), which transforms the electric field into a completely radial
field, solves this problem in an effective way.
The discharge magnitudes are significantly larger in laminated insulations ranging up to more than
10 000 pC. Finally it should be noted that there is no reference PD data on new laminated cables as PD
testing is not required as part of routine factory testing. Newly produced laminated cable, contrary to
extruded cable, is never tested for PD behavior, because discharge activity has not been considered to be a
relevant quality standard for new laminated cable, as it certainly is for extruded cable. Although newly
produced laminated cable is subjected to dissipation factor (tan delta) testing, this test is not as sensitive a
tool for measuring PD activity as a conventional PD testing procedure. The use of PD testing as a
diagnostic tool for laminated cable is relatively recent. As there is no obvious standard, the interpretation of
test results has to be developed in an empirical way.
4.1.3 Typical partial discharge sources in accessories
Less than ideal conditions during installation can lead to the inclusion of defects of accessories in the field.
Typical defects that can result in PD are as follows:
⎯ Voids in molded products due to improper shrinkage of accessory components or poorly shrunk
layers in heat-shrink and cold-shrink products.
⎯ Loose insulation-joint interfaces (tracking), especially in the absence or migration of silicone
grease, and at indentations made into the insulation during joint installation.
⎯ Incorrectly assembled joints, e.g., misalignment, improper positioning, installation scratches,
contamination, and voids.
⎯ Contamination such as moisture, metal oxide, and so on may also leak into interfaces.
⎯ Poorly installed terminations (incorrect positioning of stress relief devices), surface tracking,
cuts into the insulation at the edge of the cable semiconducting shield, or corona from terminal
hardware.
⎯ Knife cuts or bruises produced by splicing tools, especially at the edges of semiconducting
shield cut-backs, and voids in the insulation, mostly along knit lines or at interfaces.
⎯ No contact made between the connector and the semicon or if the neutral wire is not properly
connected to the outer semicon of the splice body.
⎯ Terminal connector hardware and contaminated surfaces.
Differences in thermal expansion of the cable and accessory components may cause the generation of
cavities at high temperatures.
Discharges resulting from a discontinuity in the connector or shield of a splice or a cable often resemble a
corona discharge. This type of discharge is often identified from its pulse height distribution. As the
problem progresses, the PD magnitude often remains unchanged while the PD repetition rate increases.
11
IEEE Std 400.3-2006
The assessment of the condition of a splice with PD is very complex and often inconclusive, as it depends
on the particular type of splice. Phase-resolved PD displays may be able to differentiate between tracking-
type defect and void-type defect.
However, more experience in interpreting PD results is needed by carefully testing, dissecting, or leaving in
service until failure large numbers of splices of different designs. Past research in this area has not been
successful. In general, any tracking-type PD that occurs continuously at operating voltage should be
considered as severe and warranting early repair, whereas PD sources that extinguish at a voltage above
operating level need not be considered as candidates for immediate repair. Periodic monitoring can help
establish trends and help to better assess the severity.
4.2 Insulating materials
It is important to know the materials being tested to better interpret PD data as the resistance to damage by
PD depends on the insulating material. The order of PD resistance is XLPE << EPR << Laminated (i.e.,
fluid-impregnated paper). Cable accessories, often made with filled rubber, may have a high endurance to
PD activity, provided this does not occur adjacent to extruded cable insulation.
Shielded distribution cables fall into two classes: PD-free and PD-resistant. PD-resistant cable can sustain
substantial amounts of PD over long periods of time without failure. PD-free cable can be formulated with
a range of dielectrics, having low PD resistivity. However, for both types of cable, certain forms of PD will
eventually cause failure, whereas other forms of PD can continue almost indefinitely without failure.
Knowledge rules, which give objective guidance for the interpretation of measured data for the different
materials in use, are necessary.
4.3 Operating conditions
Operating conditions also influence PD characteristics. The electric stress to initiate PD in cable systems
comes from the operating voltage and transient overvoltages such as lightning or switching surges. The
occurrence rate and the magnitude of voltage transients depend on the structure of the electrical system and
its geographic location. A cable operating in Florida is exposed to more severe lightning surges than a cable
operating in California. A system where frequent switching of capacitor banks is required tends to be
exposed to more severe switching transients than a system without capacitor banks. The quality of surge
protection affects the transient voltage impressed on a cable during a lightning storm. Selection of adequate
arrester characteristics, proper grounding, and judicious placement of surge arresters affect the quality of
lightning protection. These, in turn, bear an influence on the rate at which PD sources may be created or
existing PD sources may deteriorate in distribution cable. The maximum test voltage selected is influenced
by the quality of insulation coordination. Cables with defects are, therefore, exposed to PD events with
probabilities that depend on the operating environment.
Another operating environment affecting the relative severity of a PD causing defect is the cable loading
during service. If an XLPE insulated cable is operated at very high temperatures, significantly higher than
the crystalline melting temperature, mechanical and structural changes could occur in the insulation. These,
in turn, may affect the PD characteristics. The size of some defects varies with cable loading so that the PD
magnitude varies with the cable loading. PD in some cases might disappear (extinguish) as load conditions
change. For a laminated cable, long-term operation at high temperatures may cause the insulating fluid to
migrate. The extent of this depends on the viscosity of the impregnating fluid. This can affect the PD-
producing defect, which may either improve or further deteriorate depending on the cable topography. This
type of operation may also produce unexpected failure due to thermal instability.
In testing three-core cables, care must be taken to identify the correct phase in which the PD is occurring.
PD may occur in more than one phase, and cross-talk between phases may give misleading results.
12
IEEE Std 400.3-2006
When a cable circuit is taken out of service for off-line PD testing, the PD activity is extinguished when the
voltage is removed. To re-initiate PD that can continue at normal operating voltage, a voltage of up to ~1.5
Vo may have to be applied for some minutes, as the PD inception voltage (PDIV) may be greater than the
voltage required to maintain PD (PDEV) once it is initiated. A test duration of some minutes is required to
ensure that there is an electron to initiate the discharge. Although in theory the PDIV may be two times the
PDEV, in practice it is usually about 1.3 to 1.5 times. Thus, in off-line PD testing, the test voltage should
be raised above normal operating voltage to initiate PD that was probably already active before the cable
was taken out of service. If the cable is tested on-line without removing the voltage prior to testing, then
those PD sources that are likely to be active during normal service are probably already in PD at the time of
the test, although intermittent PD can occur if the PDEV is close to the service voltage.
Defects that have both the PDIV and the PDEV below the system operating voltage (SOV) will produce PD
once the cable system is energized. The PD will be maintained during operation of the cable. Both on-line
and off-line methods will detect these defects provided the PD detection system has the required sensitivity.
As there is a high risk of failure when there is continuous PD in an electrical tree in an extruded cable, it is
unlikely that PD observed at the operating voltage is coming from an electrical tree. It is not unusual to
have partial discharges up to thousands of pC in laminated cable, particularly during changes in
temperature and pressure.
Defects that have their PDIV > SOV and their PDEV < SOV will not produce PDs at normal operating
voltage unless they are triggered by transient overvoltages. However, once triggered, they may be self-
sustaining until the voltage is removed, or they could become intermittent or be extinguished completely. If
the PD activity does not induce further damage at the defect, these sites will require a system overvoltage to
initiate PD each time system voltage is restored. Both on-line and off-line methods will pick up these
defects provided that, for the on-line test, a surge has already initiated the PD.
Defects that have both their PDIV and their PDEV above the SOV can initiate PD by transient
overvoltages, but the PD will usually extinguish quickly after some cycles of the AC voltage. PD could also
be generated by off-line tests, particularly if the test voltage is >2 Vo (see 7.4). These defects are normally
harmless during normal system operation, but there is a possibility that an electrical tree could be initiated
having PD that could persist below the SOV, i.e., PDEV < SOV due to the high electric field at the tips of
the trees.
For most PD sources, both the magnitude and the repetition rate increase as the excitation voltage is
increased. If the PD source generates low magnitude pulses, for example, in spherical cavities or in cables
with thicker insulation, high sensitivity detection needs to be used and the PDIV may then depend on the
detection sensitivity.
In a complex cable system such as network feeders, the network can consist of a mixture of cables with
different insulations, constructions, and load capabilities. Discharge can occur at the locations where the
ground system of two cable sections operates at different ground potentials. In three-conductor cables,
discharge can also occur between the shields of the individual phases and the overall cable shield when the
shield is not connected properly. This often occurs when two cable sections with different constructions are
spliced together. The voltage build up in the ground system is induced by the current being conducted and
is therefore a strong function of the cable loading. In some cases, if an inadequate grounding is employed,
the ground potential difference is caused by imbalance between the cable phases. The imbalance between
the currents carried by the cable phases is more pronounced as the cable loading increases. For this type of
discharge, testing under heavily loaded conditions is essential.
5. Partial discharge detection

PD pulses are very short, typically 1 ns to 5 ns wide, and they can have significant frequency components
up to 1 GHz at its source. Two general approaches are available to detect PD pulses in installed cables,
which is referred to as off-line and on-line detection (Ahmed and Srinivas [B3]). Off-line testing is
normally carried out using a separate voltage source after the cable has been removed from service (see
13
IEEE Std 400.3-2006
Mashikian [B22] and CIGRE WG D1.33 Task Force 05 [B10]). On-line (in-service) testing is carried out
during normal operation of the cable system (Ahmed and Srinivas [B2]).
Some advantages of off-line PD testing are as follows:

⎯ PDIV and PDEV can be measured if a variable voltage source is used.
⎯ PD characteristics can be obtained at different voltages, which can aid in the identification of
certain types of defects.
Some advantages of on-line PD testing are as follows:
⎯ PD characteristics can be obtained under different load conditions, which can aid in the
identification of certain types of defects.
⎯ Tests can be performed without having to take an outage.
Some PD test equipment and methods are designed to only detect the existence of partial discharges and
not locate their sites. This type of equipment is used to test lumped components such as switchgear or
capacitors. Other PD test equipment and methods can detect and locate PD sites. Only the latter are covered
in this guide.
Two important factors in PD testing are detection sensitivity and condition assessment from the
interpretation of the data. PD test methods vary significantly in their detection sensitivity, and the PD test
providers use different criteria to assess the insulation condition posed by the source of the detected PD.
Cable users often rely on economic considerations when they select a test method. The following
considerations are recommended to aid cable users in weighing the advantages and disadvantages of the
available test methods:
⎯ The detection sensitivity should be given a high priority. Poor sensitivity may result in fewer
problems being detected.
⎯ The test methods should be capable of providing pattern recognition to identify the types of PD.
⎯ The operating conditions of the cables, for example, the level of surge protection and the load
history, as well as the planned load in the future should be taken into consideration when
assessing the risk associated with the detected PD.
⎯ The test method should not induce or aggravate degradation of the system.
5.1 General test setup
An off-line test is generally conducted according to the test setup shown in Figure 1. The cable is
disconnected from the network at both ends and properly isolated. A voltage source and a coupling device,
or sensor, are connected at one of the ends (near end), whereas the remote end is left open. The coupling
device could be capacitive or inductive. The coupling device is connected to the PD detecting and
processing systems. Variations of this setup include a measuring system with sensors at both ends and
means to communicate the far end data to the near end processing devices or, in the case of a branched
system, sensors placed at the end of each branch. Multi-terminal testing also has the advantage of greater
sensitivity in the PD testing of very long cable lengths as the pulse travel distances are considerably shorter
and consequently the related attenuation of pulse amplitude will be less.
14
IEEE Std 400.3-2006
Figure 1 —Typical test setup for off-line PD testing
In an on-line measuring system (Figure 2), the cable remains in service with both ends connected to the
system. Coupling devices of types 2, 3, and 4 are generally used as illustrated in Figure 1.
NOTE—The termination is connected to the power system.

Figure 2 —Typical test setup for on-line PD testing
5.2 Partial discharge detection methods
For PD assessment to be a useful diagnostic tool, the PD detection system should be capable of measuring a
wide range of PD characteristics, including PD location, required to identify the types of defects that
produce PD in the cable system.
In general, broad categories of detection are electromagnetic and acoustic. Acoustic detection is sometimes
used in compressed gas (SF6) filled lines, transformers, and on cable accessories to determine the exact
location of PD activity. The prevalent detection method is electromagnetic. This is accomplished by
coupling the cable to the detection instrument either capacitively or inductively, as illustrated in Figure 1
and Figure 2. Each method has advantages and limitations, which are beyond the scope of this guide to
discuss. The user of such a test method should be concerned about the overall sensitivity and PD location
resolution of the test method.
15
IEEE Std 400.3-2006
5.3 Partial discharge detection sensitivity
The detection sensitivity in any field test method is dominated by external noise in the operating
environment. Noise suppression through analog techniques, e.g., frequency filtering, seldom results in
detection sensitivity better than a few tens of picoCoulombs. Such sensitivity is acceptable for laminated
cable but not for extruded dielectric cables for which a higher sensitivity is preferred. However, frequency
filtering by modern digital filters can improve the sensitivity.
With regard to extruded dielectric cables having PD activity in the range of several picoCoulombs,
inadequate detection sensitivity may mask the existence of serious defects with low PD magnitudes.
Inability to detect low levels of PD may result in “false-negative” situations that are expected to lead to
unexpected post-testing service failures. In addition incorrectly identified PD may lead to “false-positive”
situations leading to unnecessary cable replacement.
The detection sensitivity is usually expressed by the minimum magnitude of apparent partial discharge, in
picoCoulombs or the equivalent millivolts (equivalence to be provided by the tester), which the instrument
is capable of resolving under existing field conditions. As the cable length and the environmental noise
conditions increase, the sensitivity tends to decrease. Cable construction, such as size, type, and the
condition of the neutral conductor, and the properties of the semiconducting shields, have a significant
influence on PD pulse propagation velocity, and attenuation. The latter, in turn, affects the detection
sensitivity. Internally generated noise can also limit the sensitivity of an instrument. High sensitivity is
crucial when testing extruded cables where detrimental PD magnitudes are known to be relatively low.
Detection sensitivity and PD location accuracy must not be confused with each other, although the
effectiveness of noise mitigation can influence both quantities. The following three cable conditions may
arise in a test situation:
a) A circuit under test may contain cables made of the same material, but different conductor cross
sections, connected in series. In such a case, the estimated cable length, the attenuation, and the
calculated discharge magnitude may be slightly affected.
b) A circuit under test may contain cables using different insulating materials or constructions
(mixed cable systems). In such a case, both the detection sensitivity and the PD location
accuracy may be adversely affected.
c) A circuit under test may contain new and service aged cables with different water content and
different wave propagation characteristics. In such a case, the PD location accuracy may be
adversely affected.
If the cables are tested in sections, the above conditions do not apply.
The instrument sensitivity can be checked by injecting a calibration pulse, such as 10 pC, 20 pC, 50 pC, or
100 pC, into the cable and determining the ability of the instrument to resolve the response of the cable to
the smallest pulse. Calibration is discussed in Annex C.
5.4 Partial discharge location
The location of PD sites is accomplished either in the time domain or in the frequency domain. Each
category will be described in turn. PD measuring systems capable of good location resolution on installed
cables usually have bandwidths of several megahertz, typically in the 5–20 megahertz range. Note that the
bandwidth in IEC 60270 is defined by an upper, f2, and a lower frequency, f1, limit at which the transfer
impedance has fallen by 6 dB from the peak pass-band value. Note also that the frequency range defined in
IEC 60270 for “wideband” PD detection systems is 30 kHz to 500 kHz as lower and upper frequency
limits, respectively, with a Δf (=f2 – f1) of 100 kHz to 400 kHz.
16
IEEE Std 400.3-2006
5.4.1 Time-domain partial discharge detection and location
Time-domain testing is usually performed off-line, although on-line with multiple sensors is also possible.
Individual pulses are measured if a wide bandwidth detector of at least several tens of megahertz is used.
Lower bandwidth detectors may result in the superposition of pulses. When using low bandwidth detectors,
for example, as per IEC 60270, the superposition of pulses can be checked by using a double pulse
calibrator with an adjustable pulse sequence from 1 μs to 100 μs according to IEC 60885-3. The phase
positions, magnitudes, and repetition rates of the pulses can be measured.
A common method of estimating the location of a PD in a cable uses the principle of reflectometry, as
illustrated in Figure 3 (Mashikian et al. [B21]). The excitation and measurement are implemented from one
cable end (near end). A PD signal splits into two equal signals that travel in opposite directions. The direct
signal traveling toward the near end is recorded first as pulse A. The horizontal axis represents time. The
signal traveling in the opposite direction is reflected at the remote end and travels back to the near end
where it is recorded as pulse B. Using the difference t between the arrival times of these two pulses and the
velocity of pulse propagation, the PD location can be estimated. In this figure, the third pulse C represents
the recording of pulse A after it has undergone successive reflections at the near and remote cable ends. Its
time difference with respect to pulse A is the cable round-trip time.
Figure 3 —Reflectogram illustrating PD location in the time domain
A simultaneous measurement of pulses at the near and remote ends is possible if a PD detector is installed
at each end, and the detectors are provided with means to synchronize their recordings and transmit them to
a measuring station where they are processed. The PD location is determined as a function of the PD pulse
arrival time at the various sensors. This method is useful for PD measurements on branched circuits and for
very long cable lengths. Technical and economic considerations may limit the number of branches that can
be tested simultaneously (van Schaik et al. [B30]).
For direct buried cables, an estimate of the PD site is not sufficient to locate it exactly. Several instruments
are commercially available to match the above-ground location with the actual PD location along the cable,
as estimated by the methods described previously. Once the cable is unearthed, it is possible to pinpoint the
exact PD site by means of specialized instruments, provided the external metal shields or sheaths are
removed. For cables buried in ducts, these steps are unnecessary, as an entire section between two
manholes will have to be pulled out.
Termination discharges can be identified by time-domain location methods and by means of special probes,
although some instruments have been developed to isolate these discharges by means of very high-
frequency amplifiers/filters. These would detect signals emanating only from within short distances, as
cables rapidly attenuate very high-frequency signals.
17
IEEE Std 400.3-2006
In high-voltage cable applications, PD detection in joints and terminations can also be accomplished by
means of sensors capacitively coupled to individual joints and connected to very high-frequency bandpass
filters.
5.4.2 Frequency-domain partial discharge detection and location
Frequency-domain testing, which can be performed either off-line or on-line, is based on measuring the
frequency components of the PD pulses (Ahmed and Srinivas [B2]). The frequency spectrum of the time-
varying PD pulses can be obtained either digitally (mathematical conversion) or through the use of analog
swept filters as in a spectrum analyzer. For both analog and digital frequency-domain PD detection, the
outcome depends on the setting of the measuring device used (parameters such as sample rate in the case of
digital conversion, filter bandwidth, and sweep rate in the case of analog conversion).
In general, the detected frequency components of a PD signal are in the range of a few hundreds kilohertz
to up to 1 GHz depending on the location of the PD with respect to the measuring point. However, a
frequency range of a few hundred kilohertz to 300 MHz is more practical due to limitations imposed by the
frequency bandwidth of the PD detection sensors. Frequency-domain testing normally is conducted in both
full and zero-span modes. In the full span mode, the frequency range scanned can be adjusted to examine
signals in narrow-frequency bands as well as wide-frequency bands. Measurement made using narrow-
frequency bands significantly enhances the resolution in locating the PD sites. PD energy versus frequency
is a strong function of distance from the PD source. The zero-span mode is used to examine PD pulses in a
time domain as measured in a relatively narrow bandwidth at a center frequency, which provides good
detection sensitivity (i.e., within a frequency range where little interference occurs).
The zero-span mode is used to find PD pulses occurring at one or more cycles of the operating voltage of a
specific frequency. The signal from the zero-span mode is often coupled to a pulse phase analyzer. The
pulse-phase analyzer is capable of recording PD pulses sorted by their phase angle and magnitude relative
to the power frequency excitation.
The frequency-domain technique does not usually provide a direct basis for calibration of the resulting data
in terms of apparent charge (pC). PD sources are recognized on the basis of their spectral characteristics
and through the use of the zero-span mode of the spectrum analyzer to provide a correlation between the
signal in the selected spectral region and power frequency. For such systems, the relationship between the
measured signal and the “severity” of PD source is based on subjective information, such as experience,
and the results of laboratory testing.
Unlike the time-domain methods, the frequency-domain methods are more immune to interference from
external noise, provided means are procured to prevent stray signals from being coupled into the cable
tested, when testing is performed on-line. However, a skilled operator is needed to collect and analyze the
data. A high sensitivity can be achieved with adequate sensors and if the distance between the test points is
kept below 150 m (500 ft).
PD location is estimated by measuring the energy versus frequency of the PD pulses. The accuracy of the
location depends on two factors:
⎯ The frequency range of the pick up sensors (greater bandwidth results in improved accuracy).
⎯ The distance between two consecutive test points (shorter distance improves accuracy).
PD location is often judged from the frequency content of the PD-induced signal in combination with
knowledge of the cable type (i.e., the high-frequency attenuation of the cable). If the PD-induced signal is
sufficiently large, the location can be determined using time-domain approaches.
18
IEEE Std 400.3-2006
5.4.3 Location resolution
The resolution with which PD can be located is affected by the bandwidth and internal noise of the
measuring instrument, the attenuation of the cable, the soundness of cable terminations, the number and
quality of splices, the physical obstacles encountered at the cable termination sites which interfere with
proper connections, the analysis methods and procedures used to assess PD site location, and very
importantly, the background noise and its mitigation (Zhifang et al. [B33]). As the instrument bandwidth
decreases or the cable attenuation and length increase, the PD location resolution decreases. In time-domain
PD testing, the exact time of arrival of a pulse cannot be measured accurately either by locating the
beginning or the peak of a pulse, so that sophisticated methods are required to perform a measurement with
high sensitivity and high resolution under adverse test conditions. In frequency-domain testing, the location
resolution is mainly influenced by the distance between two consecutive test points (shorter distance
improves accuracy). Annex B discusses in greater detail the conditions affecting PD site location results.
5.4.4 Splice location
Prior to performing high-voltage PD testing, the cable splices need to be identified and located in order to
differentiate between PD signals emanating from a cable and PD signals from a splice. This is necessary
because of the differences in the insulating materials of cables and splices and, therefore, their relative
resistance to PD degradation.
A low-voltage pulse is sent from the near end of a cable and all its reflections are recorded. A mismatch
between the characteristic impedances of the cable and splices causes these reflections. An impedance
mismatch may also be due to impedance changes caused by extensive physical deformations, corroded
neutrals, or heavy moisture absorption. The principle of time-domain reflectometry (TDR) is used to locate
the sites of abrupt impedance changes. In on-line testing, the same technique is employed with the
exception that the low-voltage pulse is injected into the energized cable via a current transformer.
5.5 Test limitations
PD testing limitations may be caused by the following:

⎯ Lack of detection or location sensitivity (for example, taped insulation shields in older cables
can greatly attenuate PD pulses).
⎯ Defective neutral conductor (excessively corroded, cracked wire/tapes, etc.).
⎯ Inability to energize a long length of cable for off-line testing.
⎯ Inability to properly locate the source of PD in a long buried cable section for frequency-domain
testing.
⎯ Presence of dominant extraneous sources of partial discharge that may be coupled to the
measuring system.
⎯ Poorly trained test operators.
⎯ Inaccessible cable or its components.
Limitations associated with each type of PD test method or voltage source have already been discussed. An
assessment of these limitations and their consequences on the test results should be made preferably before
initiating the test; however, if the limitations are discovered during the tests, they should be reported in the
test report. Of course, a correct interpretation of the results is essential. Under certain conditions, the exact
level of PD severity may not be assessable. This should be clearly pointed out in a test report.
19
IEEE Std 400.3-2006
6. Voltage sources
The purpose of the test voltage is to produce partial discharges at locations where there are defects in the
cable, terminations, and splices. On-line testing uses the system voltage of a constant fixed magnitude, The
desired characteristics of off-line voltage sources for field partial discharge measurements are as follows:
⎯ The applied voltage should cause partial discharges in the cable, terminations, and splices that
have characteristics close, if not identical, to those that occur when the cable system is in
service.
⎯ It should cause no appreciable damage to the cable system during the time required to perform
the measurements.
⎯ In the case of off-line testing, the maximum voltage applied should be variable.
⎯ The size and weight of the equipment to produce the voltage should facilitate field
transportation.
Voltage sources that are used for commercially available field partial discharge measurement systems fall
into the general categories of power frequency and alternative voltage sources. A summary of these
voltages sources as well as some of their advantages and disadvantages are contained in the following
sections.
6.1 Sinusoidal power-frequency voltage sources
Constant amplitude, power-frequency, sinusoidal voltage sources have the advantage of duplicating the
cable system operating voltage. This type of voltage source also has the advantage that the partial discharge
measurement results are directly comparable with partial discharge measurement results from routine
production tests for extruded dielectric cable systems as long as the PD detection equipment comply with
industry standards for measurement and calibration as used in the factory. For the purpose of this guide,
power-frequency voltage sources refer to sinusoidal voltages with a frequency ranging from 20 Hz to
300 Hz.
One special case of this category of voltage source is on-line testing, where the power system provides the
rated 50 Hz or 60 Hz voltage during the partial discharge measurements. On-line partial discharge
measurements require a minimum of equipment to perform the measurements and may eliminate cable
system downtime; however, it is not possible to vary the magnitude of the test voltage to determine the
PDIV and PDEV. On-line partial discharge measurements may reduce the possibility of extinguishing
partial discharges during removal of the operating voltage to perform subsequent off-line measurements.
However, this temporary extinction of partial discharges to prepare for off-line measurements generally is
not a problem if the off-line test voltage is significantly higher than rated cable system voltage. On-line
measurements do not cause any damage to the cable system during the partial discharge measurements.
To obtain the added flexibility of performing partial discharge measurements at voltage magnitudes above
and below rated voltage, field transportable power frequency test equipment has been developed for
commercial applications. To keep the size and weight suitable for field transportation, resonant test sets are
generally used. Two types of power frequency resonant test systems are commercially available for field
partial discharge measurements:
a) Frequency tuned resonant circuits (FTRCs).
b) Inductively tuned resonant circuits (ITRCs).
20
IEEE Std 400.3-2006
6.1.1 Frequency tuned resonant circuits
FTRC systems have the advantage (compared with ITRC systems) of having no moving parts, which
results in less susceptibility to damage during transportation.
A disadvantage of FTRC systems is that they employ high-power electronic converters that lead to the
introduction of noise pulses. It is necessary to filter the noise pulses out of any PD measuring system. As
the noise pulses may coincide with PD current pulses in the cable, it may be difficult to detect certain cable
defects.
6.1.2 Inductively tuned resonant circuits
An advantage of ITRC systems over FTRC systems is that they use variable auto-transformers that do not
produce interference pulses that affect partial discharge measurements. Voltage variation is gradual and
smooth and facilitates an accurate determination of PDIV.
ITRC systems contain moving parts, which require periodic maintenance. Both the FTRC and the ITRC
test sets are generally heavier than equipment required by other alternative voltage sources.
6.2 Alternative voltage sources
For the purpose of this guide, alternative voltage sources refer to sources used for field partial discharge
measurements that are non-sinusoidal and/or have frequencies other than power-frequency voltages.
6.2.1 Very low-frequency voltage
Commercially available VLF voltage sources are low in weight and low in capacitive power demand
compared with power-frequency excitation sources. Unlike dc voltage, VLF test voltage is less likely to
produce harmful space charge because of its continuous polarity changes (Steennis et al. [B28]). Some
commercially available VLF voltage sources employ mechanical switches in the high-voltage circuit and
thus are not PD free in these portions of the voltage cycle.
Depending on the type of defect, VLF voltage sources, usually 0.1 Hz, for extruded-dielectric cable
systems may require a higher test voltage to generate the same partial discharge level compared with tests
performed with power-frequency voltages. For example, the conductivity of the surface of a cavity that has
been exposed to PD increases, which allows any charges deposited on the surface by PD to leak away and
thus lowers the electric field in the cavity. As more charge can leak away between polarity reversals at VLF
than at power frequency, the PDIV at VLF will be larger than that at power frequency. If there has been no
previous PD activity to increase the conductivity of the cavity surface, the PDIV at VLF and power
frequency will be similar. Also, as more charge on electrical tree channels can leak away between polarity
reversals under VLF than at power frequency, the tree growth rates at the two frequencies will be different.
It is not possible to determine the change in surface conductivity and thus the difference in the PDIV
between power frequency and VLF. In addition, the PDIV and PD patterns produced with VLF excitation
may deviate from those produced with power-frequency voltage sources.
At low frequency, PD pulse reflection patterns are better separated, which makes site location of the PD
easier for the test engineer when testing cables producing discharges with high repetition rates, for
example, laminated cables and its accessories (Hetzel and MacKinlay [B15]).
It has been reported that the growth rate of electrical trees undergoing partial discharge can be higher for
0.1 Hz than for power frequency. Once an electrical tree has been initiated during VLF testing, there is a
higher likelihood of failure; however, there is no evidence that VLF testing has a lower electrical tree
initiation voltage. It should be noted that this concern does not apply to laminated cable systems. For more
details on VLF testing, refer to IEEE Std 400.2-2004.
21
IEEE Std 400.3-2006
6.2.2 Damped alternating voltage
The generation of a damped AC voltage (DAC) can also be used as an alternative to other methods of
generating an ac test voltage on site (Gulski et al. [B14]). The voltage sources are low in weight and have
low power requirements compared with power-frequency excitation sources.
A typical test circuit consists of a direct voltage source, which charges the cable capacitance with
continuously increasing voltage (no steady state situation). After reaching the desired peak voltage, a
switch short-circuits the cable through a wave-shaping circuit. The wave-shaping circuit can be adjusted to
produce the desired test frequency. The frequency of the damped oscillation varies from some tens of hertz
to a few kilohertz depending on the type of wave-shaping circuit used.
Test systems using damped alternating voltage usually require repeated applications of the test voltage. The
short dwell time at peak voltage level of the test voltage may be viewed as both an advantage and a
possible disadvantage. The small number of cycles reduces the risk of damage to the cable system when the
peaks of the test voltage are higher than rated voltage. However, discharges associated with certain types of
cable system defects may require a longer voltage application time to initiate.
With damped alternating voltage in the range of kilohertz, the PD inception and occurrence may differ from
those for power-frequency voltage.
6.2.3 Impulse voltage
An impulse voltage may also be used as a voltage source for field partial discharge measurements. The use
of impulse voltages with a very fast rate of rise and a decay rate equivalent to power frequency has been
reported. Its main advantage is its lighter weight.
A disadvantage of the impulse voltage test is the difficulty in accurately determining the PDIV. As cables
tend to highly attenuate high-frequency voltage pulses, the test voltage pulse tends to be attenuated and
distorted as a function of cable length. Therefore, the voltage stresses in the cable system may differ
significantly with distance from the voltage source.
It is also difficult to relate partial discharge values measured with this type of voltage to partial discharge
values from routine factory tests.
7. Practical testing guidelines
7.1 Pre-test owner information
To facilitate the testing process, detailed information needs to be obtained from the cable owner. This
information includes the following.
7.1.1 Cable characteristics
These include conductor and neutral sizes, type and condition of neutral, rated voltage, insulation type,
cable age and operating history, cable dimensions or total capacitance, cable length, and type and location
of joints. If multiple cable types exist within a given test span, the above information should be provided
for each type of cable in the span, along with the locations at which the different cable types are spliced
together.
22
IEEE Std 400.3-2006
7.1.2 Cable configuration
This includes information on whether the system is single or three-phase (is each phase individually
shielded or is there one single shield over all three phases), whether the system is radial or looped, whether
the system is direct buried or in duct, and other information pertaining to the cable routing, particularly
with branched circuits.
7.1.3 Termination type and accessibility
Information is needed on whether the termination is live or dead-front, in a switching cabinet or substation
housing, on a pole-top, in SF6 or other. It is also necessary to know whether terminations are accessible,
whether arresters and transformers can be disconnected, and whether any switch permanently connected to
the cable can take the maximum test voltage without discharging or flashing over.
7.2 Preliminary inspection and/or testing
A physical inspection by the testing organization of the system to be tested is helpful and is highly
recommended. A relatively quick and inexpensive preliminary test is recommended, including a sensitivity
test, to determine whether (a) the cables to be tested are insulated with certain old-vintage rubber
compounds that cannot be tested; (b) the cable construction is such that very high attenuation is expected;
or (c) the neutral is suspected of not being continuous or of being ineffective. The preliminary test can be
effectively performed by means of low-voltage reflectometry. If the attenuation is so high that the reflected
pulse is hardly visible or if the neutral is not continuous, a high-voltage PD test may not be warranted, as
PD signals, especially those with low magnitude, may go undetected. Such a preliminary screening test can
be technically and economically justified.
7.3 Pre-test conference
Prior to PD testing, a conference with the cable operator is useful and recommended. During this
conference, the following items should be discussed.
7.3.1 Test protocol
The PD test protocol and the results of any preliminary screening tests shall be reviewed. Specifically, the
test voltage level shall be discussed and an agreement reached for its maximum allowable value and the
duration of its application. The advantages and limitations of the particular PD test method used must be
clearly conveyed to the cable owner. The cable owner must approve the test protocol and clearly state any
conditions that would require its revision.
7.3.2 Safety
The safety procedures, as described in Clause 9, should be reviewed and agreed upon. This includes safety
equipment, switching and blocking practices, grounding, responsibility for control of the cable under test,
and safety of the public.
7.4 Test conditions (voltage levels and duration)
On-line tests use the system voltage. The duration of the test should be until sufficient data are collected
(see 9.2), up to a maximum of 15 min.
Off-line tests can be carried out using different voltage sources; see Clause 6. There is a solid technical
rationale of testing up to 1.5 Vo to 2 Vo to ensure that the PDEV of the cable system is greater than 1 Vo.
23
IEEE Std 400.3-2006
There is an increased risk of initiating damage at defects in aged cable systems that are innocuous at
operating voltage if testing is carried out at voltages greater than 2 Vo. There is also an increased risk of
failures during the PD testing. However, some utilities will request testing up to a maximum of 3 Vo on
new cables, either on the reel or newly installed, to ensure that there was no damage during transportation
or installation. In addition, some utilities will test up to 3 Vo even though there is a significantly higher
probability of failure during the testing of the following:
⎯ Cable circuits with generic defects that may cause high failure rates, e.g., some silane-cured
cables can cause severe corrosion of aluminum conductors.
⎯ Cable circuits that are being considered for silicone injection, the rationale being that all cables
with electrical trees will fail at higher test voltages. The higher test voltages could also initiate
new electrical trees.
⎯ Cable circuits that may have suspect accessories and/or cables to ensure operation during high
load periods, e.g., during the summer months in some urban areas.
The voltage in power frequency tests may be applied for up to a maximum of 15 min to ensure that
electrons are available in cavities to initiate PD. However, once PDs are detected, the voltage should be
applied long enough to collect sufficient data up to a maximum of 15 s. Some PD testing organizations will
decrease the voltage very soon after the onset of steady PD when testing extruded dielectric cable circuits.
In summary it is not possible to standardize a specific test protocol at the current time for either on-line or
off-line tests. This may become possible as more data are collected. For off-line tests, the amplitude of the
test voltage can be varied. For heavily aged systems, a maximum test voltage of 2 Vo is suggested. As the
anticipated condition of the cable improves, the test voltage may be increased to as much as 2.5 Vo. New
cables, either on the reel or newly installed, may be tested to a maximum of 3 Vo at the concurrence of the
cable owner and cable manufacturer. The test duration should be long enough to allow the availability of
electrons to initiate PD, but once PDs are detected, the voltage should be applied long enough to collect
sufficient PD data.
8. Test results and recommendations

Regardless of the test method or voltage source utilized, the following results are expected:
a) Detection of the existence, if any, of any significant PD above a specified detection sensitivity.
b) Identification of the location of each PD site.
c) Assessment of the severity of the defect associated with each PD site.
When comparing PD test results obtained with different voltage sources but equivalent detection sensitivity
and the same applied voltage level, the lack of identical results is not unusual and should not be surprising.
The PD inception voltage levels of certain defects may be significantly affected by the type of voltage
source.
8.1 Interpretation of test results—recommendations and implications
Interpretations of PD data from both extruded and laminated cable systems are already available; however
interpretation is still developing and knowledge is continuing to be accumulated (Boone et al. [B9]). Many
organizations and companies have data banks of PD characteristics for different defects and have created
knowledge rules to improve data interpretation. New test data are compared with these stored data to
identify the type, location, and severity of the defect. The measured PD characteristics, for example, phase-
resolved PD patterns (pulse repetition rate vs. phase angle vs. magnitude), can be analyzed in various ways
based on experience and knowledge of the types of defects anticipated in the cable and accessories to
provide insight into the condition of the cable system.
24
IEEE Std 400.3-2006
As examples, for extruded cable systems, PD characteristics are being obtained for internal cavities in the
insulation, interfacial cavities between the shields and the insulation, broken shields, electrical trees from
cavities, protrusions or water trees, knife cuts, contaminants, and poorly installed accessories. Similarly for
laminated cable systems, PD data are being obtained on fluid-deficient butt gaps, soft areas of the
insulation with poor impregnation, dry and brittle paper, carbonized tracking, moisture in the insulation,
and the leakage of fluid from gaskets or holes in the sheath. The PD characteristics are dependent on the
cable design, the materials used, and the operating conditions. For example, cavities can change size or
migrate with temperature and thus the load on the cable. This could result in significant changes in the PD
activity and thus the measured PD characteristics.
Data interpretation for both extruded and laminated cable systems is difficult and is discussed more fully in
Clause 4. The PD characteristics produced by most defects usually exhibit significant variations over both
the short and the long term. The magnitudes of the variations depend on the type and location of the
defects, the type of cable system, and the operating conditions. Low magnitude discharges may not be
measured if there is high background noise during the PD measurement.
In general the accuracy in interpretation of PD data is good when testing “very good” (low levels of PD
activity) or “very bad” cable systems with, for example, low PDIV with well-defined PD characteristics.
The accuracy in interpretation is less when testing cable systems between “very good” and “very bad”
conditions. Cable attenuation and background noise affect the PD detection and measurement sensitivity of
every circuit. Thus, there is a risk of not being able to detect PD pulses or wrongly identifying pulses as PD
pulses, according to the test conditions. This could lead to the risk of an incorrect assessment of the cable
circuit. This risk must be recognized by everyone involved in the cable testing, the cable owner, and the test
provider.
The accuracy in condition assessment may improve as

⎯ More data are collected and are compared with actual cable system performance.
⎯ Data are compared with information from the results of dissections of cable parts or accessories,
which were recommended to be replaced.
⎯ Additional testing is carried out on cables and their accessories that were recommended to be
replaced.
⎯ Periodic measurements are made on the same circuits, i.e., trending,. When available, data from
previous PD tests on the same circuit will be very helpful in diagnosing the implications of
detected PD sources.
⎯ All relevant information about the cable system is known, such as the age type and design of
cables and accessories, operating conditions, and so on.
⎯ Standardized test and analysis procedures are developed. This will aid the comparison of data
bases of different PD service providers and from different utilities.
Based on the interpretation of the PD test data, the cable user has to make decisions regarding re-use,
replacement, repair, or refurbishment or retesting within a specified period to be determined. The reliability
impact of this decision depends on the criticality of the affected circuit—the potential for personal injury,
damage to facilities/property, and/or impact on reliability performance metrics such as System Average
Interruption Frequency Index (SAIDI), Customer Average Interruption Frequency Index (CAIFI), and
Momentary Average Interruption Frequency Index (MAIFI). The economic impact of this decision depends
in part on the construction (direct buried vs. conduit system), the importance of the circuit, and the potential
lost revenues or customer claims. In a conduit system, for instance, if a termination shows unacceptable
discharge, it may be cheaper to replace the termination and the first section of cable rather than just the
termination. Likewise, if a cable is direct buried, it may be more cost-effective to replace a short segment of
the cable rather than the entire section.
25
IEEE Std 400.3-2006
8.2 Partial discharge test documentation
The PD test provider should provide cable users with a report of the cables tested and the PD test results,
and in a quotation to provide such services or in standard marketing documentation, the PD test service
provider should make clear the type of documentation that is provided. The PD test provider should give
the cable user recommendations on possible corrective action to be taken. The report of the test results
should include the value of PD detection sensitivity and a reference to the method used in obtaining this
value. The PD site location results must also be provided with an assessment of the accuracy limits within
which these results can be interpreted under the conditions of the specific test. This becomes critical where
the location is at or near a splice.
Details to be included in the report are as follows.
8.2.1 Cable system identification
The following information is desirable:

⎯ Name of cable manufacturer.
⎯ Cable section identification (i.e., substation name, from switch No. to switch No.).
⎯ Cable voltage class.
⎯ Cable insulation (if mixed, specify).
⎯ Operating voltage (phase-to-neutral).
⎯ Conductor type and size (if mixed, specify).
⎯ Cable length.
⎯ Location of splices.
⎯ Cable vintage or year placed in service.
⎯ Neutral type, for example, concentric wires, metal tapes, or flat strap, and size, for example, full
or one third.
⎯ Type of construction, i.e., direct buried, duct, aerial, jacketed, unjacketed, and so on.
⎯ Splice type, if available.
⎯ Termination type, i.e., pole-top, switching cabinet live-front/dead front, premolded, heat-shrink,
and so on.
8.2.2 Partial discharge test results
The following information is recommended:

⎯ Test date.
⎯ Date of the most recent previous test.
⎯ Estimated cable length.
⎯ Splice location.
⎯ Background noise level.
⎯ Minimum resolvable PD signal pC magnitude (sensitivity) and how it was determined. If the
sensitivity is lower than expected, provide the reason(s).
⎯ Test voltage levels.
26
IEEE Std 400.3-2006
⎯ At each test voltage level, the location of each PD site, along with the limits of accuracy.
⎯ At each voltage and site location, the number of PD events per second or per cycle of a
sinusoidal excitation voltage.
⎯ At each voltage and site location, a phase-resolved PD representation (pC versus phase angle for
each PD event recorded), provided the excitation voltage is sinusoidal. Specify the number of
cycles included in the phase-resolved diagram.
⎯ For a frequency-domain measurement, describe the spectral characteristics and the estimated
location for each PD site. Specify the limits of accuracy.
⎯ Any other diagnostic results pertinent to the test method used.
⎯ An indication of the severity of the PD behavior, if PDs are detected, and recommendations on
possible corrective action to be taken.
The format of data reporting may vary. For instance, some prefer reporting individual PD events in a three-
dimensional (3-D) form with location, picoCoulomb level, and phase angle at which each PD is initiated.
Variations of this three-dimensional representation are also possible. Others prefer a set of two-dimensional
representations, showing PD location with PDIV, and apparent charge (pC) versus phase angle for each PD
site, at each voltage level, and PD repetition rate for each PD site at each voltage level.
8.2.3 Time-domain partial discharge tests with sinusoidal voltage
For a PD test based on a sinusoidal voltage and time domain PD pulse measurements, documentation
should, ideally, be based on a 3-D plot, or equivalent representation, showing PD repetition rate (or
Number of PD Pulses over a stated time) vs. PD Magnitude and PD Phase relative to the applied voltage
for each PD site detected, or a set of two-dimensional representations, showing PD location with PDIV, and
pC versus phase angle for each PD site, at each voltage level, and PD repetition rate for each PD site at
each voltage level. If the test provider and user agree to limit the test duration once PD has been detected to
below that for which the data necessary for a useful 3-D plot can be obtained, a plot of the number of PD
pulses vs. PD magnitude may be appropriate. These data will be very useful to the system owner as a basis
for comparison with future PD tests on the same circuit, which might be carried out by a different PD test
provider. For this reason, a measurement that provides the industry standard “apparent PD charge” to
within reasonable accuracy is important. In all cases, the documentation should include an interpretation of
the data, including accuracy, in terms of implications for system reliability and the need to replace or repair
the circuits tested.
8.2.4 Time-domain partial discharge tests with non-sinusoidal voltage
Some field PD test apparatus use a low-frequency ac or unipolar test waveform with a rapid change in
voltage that simulates the time-rate of voltage variation (dV/dt) of power frequency voltage. For such test
waveforms, PD is usually only measured during the rapid discharge phase of the waveform, and reported
results should include, at least, Number of PD vs. Magnitude of PD for various locations at which PD is
detected, along with an appropriate interpretation of the data in terms of implications for system reliability
and the need to replace or repair the circuits tested.
8.2.5 Frequency-domain partial discharge tests
Reports of PD testing based on frequency-domain PD tests should provide the user with useful
documentation of the measured data along with meaningful interpretation, including accuracy, in terms of
implications for system reliability and the need to replace or repair the circuits tested. This may involve
presentation of a radio-frequency spectrum for each measured PD source along with an interpretation of
that spectrum or similar objective documentation as well as appropriate interpretation of the data in terms
of implications for system reliability and the need to replace or repair the circuits tested.
27
IEEE Std 400.3-2006
9. Safety
The use of safety procedures is a routine practice whenever high-voltage testing is performed [refer to the
National Electrical Safety Code® (NESC®) (Accredited Standards Committee C-2) and IEEE Std 510].
High-voltage withstand testing, such as dc hi-pot, is commonly practiced by cable owners whose staff is
trained to observe the necessary safety procedures. Partial discharge testing is typically performed by the
service provider. The safety practices of the two organizations must, therefore, be critically reviewed and a
common set of safety procedures agreed upon.
While testing, one or more cable ends will be remote from the testing site, therefore, before testing is
begun:
a) Cable ends under test must be cleared and guarded.

b) Cables must be de-energized and grounded.
The specific topics to be discussed and agreed upon include, but are not limited to, personal protective
equipment, switching and blocking practices, grounding and responsibility for control of the cable, public
safety, and daily meetings.
9.1 Personal protective equipment
The use of basic personal protective equipment (PPE), such as hard hats, safety glasses, hard-toe shoes, or
flame or arc-resistant clothing, is normally required by employer work rules and regulations such as NFPA-
70E. PPE requirements must be discussed and agreed upon before testing begins, and they must be
complied with rigorously during work around energized equipment.
9.2 Switching and blocking practices
Prior to undertaking the testing, the cable owner and the service provider shall review the owner’s
switching and blocking practices in order to ensure the safety of the test personnel and the test equipment.
9.3 Grounding and responsibility for control of cable
Grounding and other safety-related procedures at utility locations are governed by the NESC and local
utility rules, and they must be adhered to. All parties must familiarize themselves with these rules and
requirements prior to proceeding with a testing program. In particular, the general steps involved in passing
control of the cable during the test are spelled out as follows:
The owner and the testing organization must carefully coordinate the control of the cable under test. For
off-line testing, the owner removes the cable from service and grounds it, following generally accepted
safety procedures. The control of the cable is, then, turned over to the testing organization, making sure that
the transfer of responsibility is acknowledged. The testing organization connects the necessary test leads to
the cable and, only prior to energizing the cable for testing, the grounds are removed. The initiation of the
test must be clearly announced and a visible indicator light illuminated to indicate that the cable is
energized for testing. After the test is completed, the grounds are reapplied to the cable, the test leads are
removed, and the responsibility of the cable returned to the owner. No one is allowed to touch a test lead
until the grounds are reapplied. The procedures and tools used by the testing organization to disconnect and
reconnect the grounds must be acceptable to the cable owner.
28
IEEE Std 400.3-2006
9.4 Safety of the public
Prior to starting a test, a visit to the test site by representatives of the cable owner and the test provider
takes place. The placement of any test vehicle, the test leads, and the power source are discussed and
agreed upon. Consideration is given to vehicular and pedestrian traffic, and means are provided to ensure
the safety of everyone involved. Test leads may be a hazard both with respect to tripping and exposure to
voltage. Safety cones and ribbons are, therefore, used to prevent any such hazard. Whenever portable
generator sets are used, the impact of noise should be seriously considered and mitigated.
When tests involve the maneuvering of vehicles within an outdoor substation or near other electrical
equipment, inspection of the route in and out of the test site is performed to ensure adequate clearance
between any test vehicle and substation equipment, especially any overhead conductors or pipe ways.
9.5 Daily safety meeting
Holding a daily safety meeting is recommended. During the meeting, the safety performance of each day is
reviewed, potential weaknesses corrected, and any special conditions that may be encountered during the
next test day discussed.
10. Conclusions
Partial discharges, which can occur at some types of defects in high-voltage cable systems, can lead to
premature failures. This guide provides information on partial discharge testing and information on the
interpretation of test data to assess the condition of the cable.
There are two types of partial discharge testing, on-line and off-line. Both types have advantages and
disadvantages. There are also different types of partial discharge detection methods, one using the time
domain and the other, the frequency domain. Both types can locate the source of the partial discharge.
The risk of failure during or after the field test increases with the test voltage. For heavily aged systems, a
maximum test voltage of 2 Vo is suggested. As the anticipated condition of the cable improves, the test
voltage may be increased to as much as 2.5 Vo. New cables, either on the reel or newly installed, may be
tested to a maximum of 3 Vo at the concurrence of the cable owner and cable manufacturer.
The ability of partial discharge testing to assess the future performance of a cable system is continuously
improving.
The best accuracy in assessing the future performance is achieved on “very good” or “very bad” cable
systems.
Further improvements may occur as more data are collected, test and analysis procedures are standardized,
and if measurements on particular cable circuits are repeated on a periodic basis, i.e., trending.
29
IEEE Std 400.3-2006
Annex A
(informative)
Partial discharge and water trees
Water treeing is the most important form of degradation that may afflict older XLPE and high-molecular-
weight polyethylene extruded cables. As a result, the phenomenon of water treeing has been studied
extensively, including means by which the degree of water tree-induced degradation can be assessed. Water
treeing can be described as a self-propagating dendritic pattern of electro-oxidation, which reduces the ac
and impulse breakdown strengths of extruded insulation and is the primary mechanism of degradation of
extruded medium-voltage distribution cables. Although studied extensively, the initiation and growth
mechanisms of water treeing are not clearly understood; they are not a single mechanism but complex
interactions of chemical, electrical, and mechanical phenomena that depend on the material and applied
stresses (see Crine [B11], Ross [B26], and Zeller [B32]). The visible manifestation of water treeing is
strings of water-filled microcavities. The water-filled microcavities are connected by electro-oxidized
tracks, which are usually less than 0.1 μm in diameter, which is too small to see (see Crine [B11], Moreau
[B23], and Zeller [B32]).
The detection of water trees has been an important issue for some time, and attempts have been made to
relate dielectric loss and partial discharge characteristics to water treeing in both laboratory and field tests
on cable insulation (see Bahder et al. [B4] and Hvidsten et al. [B17]). Greater partial discharge detection
sensitivity can usually be achieved in laboratory tests with the result that many tests have been performed
on cables removed from the field or on insulation samples molded and aged in the laboratory. Numerous
researchers have attempted to detect partial discharge from water trees using both electrical and optical
techniques. The laboratory electrical partial discharge detection systems often had sensitivities in the range
of 0.01 pC to 0.1 pC (see Bahder et al. [B4]). Measurements have also been made on cables removed from
service after 5 to 13 years (see Kirkland et al. [B19]). In no case has partial discharge been detected from
water trees, either using electrical or optical detection (see Bahder et al. [B4], Bamji et al. [B5], Hvidsten et
al. [B17], Kirkland et al. [B19], Nitta [B24], and Steennis and Kreuger [B27]), unless an electrical tree
formed from the water tree (see Rasikawan et al. [B25]).
The above results indicate that despite extensive efforts to find partial discharges from growing water trees,
none has been found. This is not surprising given the relatively low electrical field necessary for water tree
initiation and growth, typically 1 kV/mm to 2 kV/mm (25 V/mil to 50 V/mil). Even allowing for some
stress enhancement at the tip of a water tree, the electric stress at operating voltage will not be sufficient to
sustain partial discharges in the microcavities found in typical water trees [<50 µm, (<2 mils)].
Water trees do not generate partial discharge. However water trees can lead to electrical trees as a result of
a lightning impulse (see Boggs et al. [B7] and Hopkinson [B16]) or as a result of AC voltage (see Bamji et
al. [B5] and Steennis and Kreuger [B27]). The likelihood of causing a pre-existing water tree to lead to an
electrical tree during a field PD test increases with the test voltage and the test duration. In general,
electrical trees are more difficult to initiate than to grow, so that an electrical tree, once initiated, tends to
grow to failure by partial discharges. Thus one can conclude that growing water trees do not generate
partial discharge signals, unless they give rise to an electrical tree. Any partial discharges at a water tree
imply the existence of one or more electrical trees at that water tree.
30
IEEE Std 400.3-2006
Annex B
(informative)
Effect of cavity shape on partial discharge inception voltage
The PD inception voltage (PDIV) of a void or cavity will depend on the shape, size, and position of the
cavity within the insulation and the type of gas and its pressure inside the cavity.
Table B.1 shows examples of how the shape of a cavity affects the magnitude of the electrical stress in the
cavity.
Table B.1—Values of electric stress and PDIV for different cavity shapes and locations in
15 kV, 250 mm2 (500 kcmil) cable
XLPE PDIV in kV for 0.5 mm (20 mil) cavity
Cavity shape Electric stress in Cavity at conductor Cavity in middle of Cavity at insulation
cavity shield insulation shield
Spherical 1.23Eda 11.7 13.8 16.0
Flat longitudinal 2.3Ed 6.3 7.4 8.6
Flat radial Ed 14.0 17.0 20.0
EPR PDIV in kV for 0.5 mm (20 mil) cavity
Cavity shape Electric stress in Cavity at conductor Cavity in middle of Cavity at insulation
cavity shield insulation shield
Spherical 1.31Ed 11.0 13.0 15.0
Flat longitudinal 3.5Ed 4.1 4.8 5.6
Flat radial Ed 14.0 17.0 20.0
a
Ed is the stress in the insulation adjacent to the cavity.
For example, the electrical stress inside a spherical cavity in XLPE insulation will be about half that in a
flat cavity such as a loose insulation shield. Thus, the PDIV for a spherical cavity in XLPE is about twice
that for a flat cavity and almost three times the value for the corresponding flat longitudinal cavities in EPR
insulation. If the flat cavity is aligned in the direction of the applied electrical stress, i.e., usually radially,
the stress inside the cavity will be the same as that in the adjacent insulation so that the PDIV will be 20%
to 30% greater than that for a spherical cavity of similar size and location.
31
IEEE Std 400.3-2006
Annex C
(normative)
Partial discharge calibration and location accuracy
C.1 Introduction
This guide covers devices and techniques to measure and to locate partial discharge activity. There are
many types of PD detection and location equipment, generally categorized as “wide band” or “narrow
band.” Some types indicate values in the time domain, and some types indicate values in the frequency
domain. Due to the differences in measuring technique, there can be great differences in sensitivity of
partial discharge detection.
Partial discharge measurements are not absolute in the sense that the measured PD magnitude, PDIV,
PDEV, and other properties are a function of defect size, position of the defect within the sample, and so
on. For example, the measured PD magnitude is generally inversely proportional to sample size. Thus, the
purpose of calibrating PD measurements is not to provide absolute accuracy but to assure that PD
measurements yield the same result under the same circumstances. Calibration is also a convenient method
to check the detection sensitivity, which is an important parameter in PD measurements. It can also be used
to verify that the test equipment is functioning normally or has changed during a test or series of tests.
C.2 Partial discharge calibration
Cable standards, ICEAT24-380, IEC 60270, IEC 60885-2, and IEC 60885-3, contain calibration methods
that are used in factory tests. The standards stipulate maximum acceptable PD magnitudes in new cables;
these specified maximum magnitudes have decreased over time so that a cable installed 30 years ago might
not meet the current standards. The calibration methods described in the standards are valid for low- or
narrow-bandwidth measurements, but they do not address wide-band measurements. In general, if the
measurement of partial discharge activity is to be compared with values obtained in the factory where tests
are done according to ICEA or IEC norms, a standardized calibration method should be used. If other
methods are used to calibrate PD measuring systems, they should be documented in the field test reports.
The calibration will be affected by attenuation in the cable, which is usually larger in aged cables and is
dependent on cable length.
For off-line PD testing in the field, the following calibration procedure may be used that follows the
narrow-band measurement procedure described in the standards. A voltage pulse of short duration (in the
order of several tens of nanoseconds) and known pC magnitude is applied across the cable insulation at the
remote cable end. This procedure is not trivial as stray capacitance effects can influence the actual pulse
magnitude applied to the cable. The pulse received at the near end is integrated with respect to time. For
analog integration, the gain is adjusted until the pC value read is equal to that of the calibration pulse. For
digital integration, a constant multiplier is adjusted until the pC obtained is the same as that of the
calibration pulse. The PD test instrument is now calibrated to measure with relatively good accuracy the
apparent charge of the PD pulse. Once calibrated, the calibration remains valid for future tests as long as
the cable characteristic impedance and the reflection coefficients at the accessories remain unchanged, and
the characteristics of the detection and measuring circuits have not changed. Periodic calibration is
recommended to ensure that all equipment is working satisfactorily.
In some on-line PD testing, the relative magnitude of the PD pulse is provided in units of voltage rather
than in pC. Some HV cable joints are built with extra electrodes where calibration signals can be injected
during on-line PD measurements in the joints. During on-line testing, a sensitivity of about 1 pC is
achievable if the distance between two consecutive test points is kept less than 150 m (500 ft).
32
IEEE Std 400.3-2006
A note of caution needs to be made concerning the pC values provided by limited bandwidth (30–400 kHz)
instruments, such as those commonly used in cable manufacturing plants, and the values obtained by large
bandwidth (several megahertz) instruments, such as those used in field PD location. Because of the
different measurement and calibration methods used in these two applications, the pC values reported may
be different.
C.3 Partial discharge location accuracy
This subclause discusses the major conditions affecting the accuracy of PD location.
PD site location is normally performed using wide-band systems that have good distance accuracy and
lower background noise susceptibility. The major site-specific parameters affecting PD location include
wave-propagation characteristics, cable length, number, quality and spacing of splices, quality of
terminations, and noise level. The major non-site-specific parameters include PD instrument performance,
noise mitigation schemes, analysis methods and procedures, and operator competence.
The wave-propagation characteristics of a cable, namely attenuation and wave velocity, are strongly
affected by cable neutral size and configuration, semiconductor shield characteristics, and cable
homogeneity. Attenuation tends to reduce the amplitude and increase the duration (width) of a PD pulse as
it travels along a cable. Amplitude reduction adversely affects the ability of the instrument to detect the
existence of a PD, whereas the broadening of the pulse adversely affects its location accuracy. A neutral
completely surrounding a cable, such as a continuous metal tape or housing, yields low attenuation and
offers an ideal condition for accurate PD location, provided its electrical conductivity is not impaired by
corrosion. In the case of concentric neutral wires, attenuation increases as the percent neutral is decreased.
The minimum percent neutral size recommended with respect to acceptable attenuation is 25%. Highly
reduced neutral size, such as 10%, produces unreasonably high attenuation and lower than expected
velocity. For the same neutral size, a flat strap configuration tends to produce less attenuation than a round
conductor. In frequency-domain PD location, high frequencies tend to attenuate over relatively shorter
distances with reduced neutral size. The role of cable construction in location accuracy becomes, therefore,
more important than in time-domain PD location.
Cable neutral size and configuration can act in combination with the electrical characteristics of the
semiconducting shields to play an important role in wave velocity and attenuation and, therefore, in PD
location accuracy. If the length of a cable is known, the propagation velocity has to be adjusted
accordingly. Otherwise, the assumed velocity should take into account the cable construction. Non-
homogeneous cable systems exist when cable constructions are mixed (i.e., mixed extruded and laminated,
or mixed conductor sizes or insulation thickness). Non-homogeneous situations also arise when a cable
section contains new and service-aged cables with different water content and non-uniform wave
propagation characteristics. The foregoing conditions tend to decrease the PD location accuracy.
For equivalent propagation characteristics, long cables attenuate more than short cables and tend to produce
less accurate PD location results. The presence of a large number of splices, especially when they present
abrupt impedance changes, adversely affects attenuation and spatial resolution. Closely spaced splices tend
to create pulses resulting from multiple reflections, distorting the original PD pulse and causing an apparent
shift in location.
Ambient (background) noise can mask PD pulses totally or create distorted pulses with resulting errors in
location assessment. A robust PD locating instrument, with effective noise mitigation measures, should
circumvent this undesirable situation. PD analysis can be totally automated, partially automated or placed
totally under human control. The effectiveness of each of these schemes significantly affects the PD
location accuracy. The bandwidth of a PD locating instrument influences the location accuracy. As the
bandwidth increases, the PD pulses tend to be sharper and crisper, tending to improve accuracy. However,
frequency-dependent noise may increase, and interfere with this accuracy, unless the noise mitigation
measures prevent it.
33
IEEE Std 400.3-2006
Annex D
(informative)
Bibliography
[B1] Abdolall K., Halldorson G., and Green D., “Condition assessment and failure modes of solid
dielectric cables in perspective,” IEEE Transactions on Power Delivery, vol. 17, no. 1, pp. 18–24,
2002.
[B2] Ahmed, N. H. and Srinivas, N. N., “On-line partial discharge detection in cables,” IEEE
Transactions on Dielectrics and Electrical Insulation, vol. 5, no. 2, pp. 181–188, 1998.
[B3] Ahmed, N. H. and Srinivas, N. N., “On-line versus off-line partial discharge testing in power
cables,” 2001 IEEE Transmission and Distribution Conference and Exposition, vol. 2, pp. 865–870,
Nov. 2001.
[B4] Bahder, G., Katz, C., Lawson, J., and Vahlstrom, W., “Electrical and electrochemical treeing
effect in polyethylene and cross-linked polyethylene cables,” IEEE Transactions on Power Apparatus
and Systems, vol. PAS-93, no. 3, pp. 979–990, 1974.
[B5] Bamji, S., Bulinski, A., Densley, J., Garton, A., and Shimizu, N., “Water treeing in polymeric
insulation,” Conference Internationale des Grandes Reseaux Electriques Haute Tension (CIGRE),
1984, Paper 15-07.
[B6] Bamji, S., Bulinski, A. T., and Densley, J., “Final breakdown mechanism of water treeing,”
1991 Annual Report of the Conference on Electrical Insulation and Dielectric Phenomena, pp. 298–
305.
[B7] Boggs, S. A., Densley, R. J., and Kuang J., “Mechanism for Conversion of Water Trees to
Electrical Trees under Impulse Conditions,” IEEE Transactions on Power Delivery, vol. PD-13, no. 2,
pp. 310–315, 1998.
[B8] Boggs, S. and Densley, R. J., “Fundamentals of partial discharge in the context of field cable
testing,” IEEE Electrical Insulation Magazine, vol. 16, no. 5, pp. 13–18, 2000.
[B9] Boone, W., van Schaik, N., and Schacker, M. S., “Reliability improvement of underground
cable circuits by condition based maintenance programs,” NETA Conference, San Antonio, TX, March
15–16, 2004.
[B10] CIGRE WG D1.33 Task Force 05, “Experiences in partial discharge detection of distribution
power cable systems,” Electra, vol. 208, pp. 35–43, June 2003.
[B11] Crine, J.-P., “Electrical, chemical and mechanical processes in water treeing,” IEEE
Transactions on Dielectrics and Electrical Insulation, vol. 5, no. 5, pp. 681–694, 1998.
[B12] Dorris, D. L., Pace, M. O., Blaleck, T. V., and Alexeff, I., “Current pulses during water
treeing, procedures and results,” IEEE Transactions on Dielectrics and Electrical Insulation, vol. DEI-
3, no. 4, pp. 523–528, 1986.
[B13] Gulski, E., “Digital analysis of partial discharges,” IEEE Transactions on Dielectrics and
Electrical Insulation, vol. 2, no. 5, pp. 822–837, 1995.
34
IEEE Std 400.3-2006
[B14] Gulski, E., Wester, F. J., Smit, J. J., Seitz, P. N., and Turner, M., “Advanced partial discharge
diagnostic of MV power cable system using oscillating wave test system, IEEE Electrical Insulation
Magazine, vol. 16, no. 2, pp. 17–25, 2000.
[B15] Hetzel, E. and MacKinlay, R., “Diagnostic field testing of PILC MV cable,” Jicable, pp. 472–
475, 1995.
[B16] Hopkinson, R. H., “Better surge protection extends cable life,” IEEE Transactions on Power
Apparatus and Systems, vol. PAS-103, pp. 2827–2834, 1984.
[B17] Hvidsten, S., Ildstad, E., and Sletbak, J., “Understanding water treeing mechanisms in the
development of diagnostic test methods,” IEEE Transactions on Dielectrics and Electrical Insulation,
vol. 5, no. 5, pp. 754–760, 1998.
[B18] IEEE 100, The Authoritative Dictionary of IEEE Standards Terms, Seventh Edition, New
York, Institute of Electrical and Electronics Engineers, Inc.10
[B19] Kirkland, J. W., Thiede, R. S., and Reitz, R. A., “Evaluating the service degradation of
insulated power cables,” IEEE Transactions on Power Apparatus and Systems, vol. PAS-101, no. 7,
pp. 2128–2136, 1982.
[B20] Kreuger, F. H., Gulski, E, and Krivda, A., “Classification of partial discharges,” IEEE
Transactions on Electrical Insulation, vol. 28, no. 6, pp, 917–931, 1993.
[B21] Mashikian, M. S., Luther, R., McIver, J. C., Jurcisin Jr., J., and Spencer, P. W., “Evaluation of
field-aged crosslinked polyethylene cables by partial discharge location,” IEEE Transactions on
Power Delivery, vol. 9, no. 2, pp. 620–628, 1994.
[B22] Mashikian, M. S., “Preventative maintenance testing of shielded power cable systems,” IEEE
Transactions on Industry Applications, vol. 38, no. 3, pp. 736–743, 2002.
[B23] Moreau, E., Mayoux, C., Laurent, C., and Boudet, A., “The structure characteristics of water
trees in power cables and laboratory specimens,” IEEE Transactions on Electrical Insulation, vol. EI-
28, no. 1, pp. 54–64, 1993.
[B24] Nitta, Y., “A possible mechanism of water trees from water electrodes,” IEEE Transactions
on Electrical Insulation, vol. EI-9, no. 3, pp. 109–112, 1974.
[B25] Rasikawan, S., Ishihara, H., and Shimizu, N., “Comparison between water-treed and
deteriorated regions. Electrical tree precursor,” IEEE Transactions on Dielectrics and Electrical
Insulation, vol. DEI-1, no. 4, pp. 597–603, 1994.
[B26] Ross, R., “Inception and propagation mechanisms of water treeing,” IEEE Transactions on
Dielectrics and Electrical Insulation, vol. 5, no. 5, pp. 660–680, 1998.
[B27] Steennis, E. F., and Kreuger, F. H., “Water treeing in polyethylene cables,” IEEE
Transactions on Electrical Insulation, vol. 25, no. 5, pp. 989–1028, 1990.
[B28] Steennis, E. F., Hetzel, E., and Verhoeven, C. W. J., “Diagnostic medium voltage cable test at
0.1 Hz,” Jicable, pp. 408–414, 1991.
10
IEEE publications are available from the Institute of Electrical and Electronics Engineers Inc., 445 Hoes Lane, Piscataway, NJ
08854, USA (http://standards.ieee.org/).
35
IEEE Std 400.3-2006
[B29] Suzuki, H. and Endoh, T., “Pattern recognition of partial discharges in XLPE cables using a
neural network,” IEEE Transactions on Electrical Insulation, vol. 27, pp. 550–556, 1992.
[B30] van Schaik, N., Steennis, E. F., Boone, W., and van Aartrijk, M., “Partial discharge
diagnostics on very long and branched cable circuits,” Nordic Insulation Symposium, Tampere, June
11–13, 2003.
[B31] van Schaik, N., Steennis, E. F., and Kruithof, M. G., “Partial discharge diagnostics as part of
condition based maintenance on MV power cables: yes or no?” Jicable, pp. 241–246, 2003.
[B32] Zeller, H. R., “Noninsulation properties of insulation materials,” 1991 Annual Report of the
Conference on Electrical Insulation and Dielectric Phenomena, pp. 19–47.
[B33] Zhifang D., Willet, P. K., and Mashikian, M. S., “Performance limits of PD location based on
time-domain reflectometry,” IEEE Transactions on Dielectrics and Electrical Insulation, vol. 4, no. 2,
pp. 182–188, 1997.
36

IEEE STD 4000 (DP) PDF

Uploaded by

Copyright:

Available Formats

IEEE STD 4000 (DP) PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

IEEE STD 4000 (DP) PDF

Uploaded by

Copyright:

Available Formats

What is the purpose of partial discharge testing?

What is the purpose of partial discharge testing?

What types of cable systems can partial discharge testing be used on?

What types of cable systems can partial discharge testing be used on?

IEEE Guide for Partial Discharge

Testing of Shielded Power Cable

IEEE Power Engineering Society

IEEE Guide for Partial Discharge

Approved 15 September 2006

The Institute of Electrical and Electronics Engineers, Inc.

Copyright © 2007 by the Institute of Electrical and Electronics Engineers, Inc.

Print: ISBN 0-7381-5221-8 SH95571

Secretary, IEEE-SA Standards Board

Matthew S. Mashikian, Chair

Nezar Ahmed Craig Goodwin Ralph E. Patterson

Roy W. Alexander Edgar O. Galyon William M. McDermid

Mark D. Bowman William B. Hopf T. W. Olsen

Satish K. Aggarwal, NRC Representative

1.1 Scope ................................................................................................................................................... 2

3. Definitions, acronyms, and abbreviations .................................................................................................. 3

3.1 Definitions ........................................................................................................................................... 3

4. Partial discharge data interpretation ........................................................................................................... 5

4.1 Type and location of defects................................................................................................................ 8

5. Partial discharge detection........................................................................................................................ 13

5.1 General test setup............................................................................................................................... 14

6. Voltage sources ........................................................................................................................................ 20

6.1 Sinusoidal power-frequency voltage sources..................................................................................... 20

7. Practical testing guidelines ....................................................................................................................... 22

7.1 Pre-test owner information ................................................................................................................ 22

8. Test results and recommendations............................................................................................................ 24

8.1 Interpretation of test results—recommendations and implications.................................................... 24

9.1 Personal protective equipment........................................................................................................... 28

10. Conclusions ............................................................................................................................................ 29

Annex A (informative) Partial discharge and water trees............................................................................. 30

Annex C (normative) Partial discharge calibration and location accuracy................................................... 32

Annex D (informative) Bibliography ........................................................................................................... 34

Accredited Standards Committee C-2, National Electrical Safety Code® (NESC®).2

ICEA T-24-380-1994, Guide for Partial-Discharge Test Procedure.3

IEC 60270, High-voltage test techniques—Partial discharge measurements.4

NFPA-70E, Standard for Electrical Safety Requirements for Employee Workplaces.8

3. Definitions, acronyms, and abbreviations

3.1.2 extruded insulation: Insulation such as high-molecular-weight polyethylene (HMWPE), crosslinked

3.2 Acronyms and abbreviations

CAIFI customer average interruption frequency index

4. Partial discharge data interpretation

The characteristics of the PD parameters depend on

4.1 Type and location of defects

4.1.1 Typical partial discharge sources in extruded cables

4.1.1.2 Cavities between cable and accessory interfaces (interfacial cavities)

4.1.1.3 High-resistance insulation shield or broken neutral

4.1.1.4 Electrical trees initiated from protrusions, voids, or water trees

4.1.2 Typical partial discharge sources in fluid-impregnated cables

4.1.3 Typical partial discharge sources in accessories

4.2 Insulating materials

4.3 Operating conditions

5. Partial discharge detection

Some advantages of off-line PD testing are as follows:

5.1 General test setup

Figure 1 —Typical test setup for off-line PD testing