IEEE Guide For The Protection of Network Transformers: IEEE Power Engineering Society

IEEE Std C37.
108™-2002
(Revision of C37.108-1989)
IEEE Standards
C37.108 TM
IEEE Guide for the Protection

of Network Transformers
IEEE Power Engineering Society

Sponsored by the
Power System Relaying Committee
Published by
The Institute of Electrical and Electronics Engineers, Inc.
3 Park Avenue, New York, NY 10016-5997, USA
Print: SH94988
19 September 2002 PDF: SS94988
Recognized as an IEEE Std C37.108™-2002
American National Standard (ANSI) (Revision of
IEEE Std C37.108-1989)

Sponsor
Power System Relaying Committee
of the
IEEE Power Engineering Society
Approved 21 March 2002

IEEE-SA Standards Board
Abstract: Guidelines for the application of network protectors are established. The use of network
transformers and protectors with distributed resources is addressed.
Keywords: distribution generation, master relay, network limiter, network protector, network relays,
network transformers, spot networks
The Institute of Electrical and Electronics Engineers, Inc.

3 Park Avenue, New York, NY 10016-5997, USA
Copyright © 2002 by the Institute of Electrical and Electronics Engineers, Inc.

All rights reserved. Published 19 September 2002. Printed in the United States of America.
National Electrical Code and NEC are both registered trademarks of the National Fire Protection Association, Inc.
National Electrical Safety Code and NESC are both registered trademarks and service marks of the Institute of Electrical
and Electronics Engineers, Inc.
IEEE Green Book and IEEE Buff Book are registered trademarks of the Institute of Electrical and Electronics Engineers,
Inc.
Print: ISBN 0-7381-3257-8 SH94988

PDF: ISBN 0-7381-3258-6 SS94988
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.
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
P.O. Box 1331
Piscataway, NJ 08855-1331
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 C37.108-2002, IEEE Guide for the Protection of Network Transformers.)
This guide was prepared by the Network Transformer Protection Working Group of the Substation
Protection Subcommittee of the IEEE Power System Relaying Committee. This guide is intended to aid in
the effective application of relays and other devices for the protection of power transformers in network
transformer vaults. New additions to this guide include clauses on distributed source generation and network
distribution Supervisory Control and Data Acquisition (SCADA).
Network distribution systems differ from radial distribution systems due to several factors. These factors
include: high-fault currents on the low-voltage side of the transformers, possible frequent operation of
network protectors, varying practices of providing dedicated high-voltage transformer protection, confined
spaces, backfeed issues, and proximity to dense public populations in city streets and office buildings.
Protection techniques presented in this guide will help minimize the effects and damage caused by network
faults. However, proper protection applications are only a supplement to other important issues related to
network power equipment. These issues, covered in other appropriate publications, include the following
items:
a) Vault and equipment design, construction, operation, and maintenance should conform with applica-
ble standards and regulations, including company, municipal, state, and country operating rules.
b) Manufacturers’ recommended operating and maintenance procedures should always be utilized.
Participants
At the time this guide was approved, the working group membership was as follows:
C. R. Sufana, Chair
J. J. Horwath, Vice Chair
C. H. Castro S. E. Grier A. P. Napikoski
W. T. Chew J. W. Hohn S. Phegley
S. P. Conrad H. D. Joham H. J. Pinto
C. J. Cook R. J. Landman R. V. Rebbapragada
D. J. Finley J. Moffat D. R. Smith
D. Fulton G. P. Moskos T. E. Wiedman
K. K. Mustaphi
Copyright © 2002 IEEE. All rights reserved. iii

The following members of the balloting committee voted on this standard. Balloters may have voted for
approval, disapproval, or abstention.
Hanna E. Abdallah Gary R. Hoffman Roger E. Ray
William J. Ackerman Jerry W. Hohn Radhakrishna V. Rebbapragada
Alexander P. Apostolov John J. Horwath Paulo F. Ribeiro
Thomas M. Barnes James D. Huddleston, III Jesus Martinez Rodriguez
George J. Bartok James W. Ingleson Miriam P. Sanders
Robert W. Beckwith George G. Karady David Shafer
Kenneth L. Black Mark J. Kempker Tarlochan Sidhu
Jack A. Buchsbaum Hermann Koch Mark S. Simon
John F. Burger Terry L. Krummrey Patrick Smith
Jeffrey A. Burnworth Robert Landman James E. Stephens
Simon R. Chano Lawrence M. Laskowski Peter G. Stewart
James F. Christensen William J. Marsh, Jr. James E. Stoner
Stephen P. Conrad J. E. McConnell William M. Strang
Ken L. Cooley Michael J. McDonald Charles R. Sufana
Robert W. Dempsey M. Meisinger Malcolm J. Swanson
Frank A. Denbrock A. P. Sakis Meliopoulos Richard P. Taylor
William K. Dick Gary L. Michel Bill Thompson
Clifford Downs Daleep C. Mohla James S. Thorp
Ahmed Elneweihi George Moskos Charles F. Todd
Markus E. Etter Charles J. Mozina Demetrios A. Tziouvaras
Dennis R. Falkenheim Brian Mugalian Joe T. Uchiyama
Joseph R. Fragola George R. Nail Eric A. Udren
Jeffrey G. Gilbert Philip R. Nannery Charles L. Wagner
Stephen E. Grier Benson P. Ng Ken D. White
E. A. Guro Michael W. Pate Thomas E. Wiedman
Robert W. Haas Shashi G. Patel Georg Wild
David L. Harris Carlos O. Peixoto David J. Zaprazny
Roy E. Hart Gary A. Petersen Janusz Zawadzki
Irwin O. Hasenwinkle Alan C. Pierce Karl Zimmerman
Roger A. Hedding Henry Pinto John A. Zipp
Charles F. Henville John M. Postforoosh John A. Zulaski
iv Copyright © 2002 IEEE. All rights reserved.

When the IEEE-SA Standards Board approved this standard on 21 March 2002, it had the following
membership:
James T. Carlo, Chair
James H. Gurney, Vice Chair
Judith Gorman, Secretary
Sid Bennett Toshio Fukuda Nader Mehravari
H. Stephen Berger Arnold M. Greenspan Daleep C. Mohla
Clyde R. Camp Raymond Hapeman Willaim J. Moylan
Richard DeBlasio Donald M. Heirman Malcolm V. Thaden
Harold E. Epstein Richard H. Hulett Geoffrey O. Thompson
Julian Forster* Lowell G. Johnson Howard L. Wolfman
Howard M. Frazier Joseph L. Koepfinger* Don Wright
Peter H. Lips
*Member Emeritus
Also included is the following nonvoting IEEE-SA Standards Board liaison:
Alan Cookson, NIST Representative

Satish K. Aggarwal, NRC Representative
Noelle D. Humenick
IEEE Standards Project Editor
Copyright © 2002 IEEE. All rights reserved. v

Contents
1. Overview.............................................................................................................................................. 1
2. References............................................................................................................................................ 2
3. Definitions............................................................................................................................................ 3
4. General background ............................................................................................................................. 4
4.1 Network development.................................................................................................................. 4

4.2 Application considerations of adding protection to spot-network vaults versus
grid-network vaults ...................................................................................................................... 6
5. Operation of equipment in a network transformer vault under fault conditions.................................. 6
5.1 Primary feeder faults.................................................................................................................... 6

5.2 Network equipment faults............................................................................................................ 7
6. Methods available to provide improved network transformer protection ......................................... 10
6.1 Electrical detection of faults ...................................................................................................... 10

6.2 Mechanical detection of faults ................................................................................................... 13
6.3 Thermal detection of faults ........................................................................................................ 14
6.4 Ultraviolet detectors................................................................................................................... 15
7. High-voltage, fault-interrupting devices............................................................................................ 15
7.1 Circuit breakers.......................................................................................................................... 15

7.2 Remotely located circuit breakers with transfer tripping........................................................... 15
7.3 Fault interrupters........................................................................................................................ 16
7.4 Power fuses and current-limiting fuses...................................................................................... 16
8. Low-voltage, fault-interrupting devices............................................................................................. 18
8.1 Network protector ...................................................................................................................... 18

8.2 Low-voltage fuses...................................................................................................................... 18
9. Distributed source generation ............................................................................................................ 18
10. Network distribution SCADA............................................................................................................ 20
Annex A (informative) Response of network relays to system faults ........................................................... 21
Annex B (informative) Example: 12.5 kV/480Y/277 V spot vault network transformer protection ............ 24
Annex C (normative) Other concepts being investigated .............................................................................. 29
Annex D (informative) Bibliography............................................................................................................. 30
vi Copyright © 2002 IEEE. All rights reserved.

1. Overview
Network transformer vaults are fire retardant enclosures normally within or adjacent to buildings or
underneath streets and alleys. They typically contain two or more power transformers. These transformers
are supplied from different subtransmission or distribution lines and are paralleled on their low-voltage side
through circuit interrupting devices called network protectors. Typically, high-voltage current-interrupting
devices have not been applied within the network vault. The low-voltage bus of a network vault may be
electrically tied to a number of other vaults to form a network secondary distribution system, which will be
called a low-voltage network grid in this guide, or each individual vault may stand alone as a spot network. A
one-transformer vault can be considered a network vault if connected to other such vaults via low-voltage
cables.
Low voltage in this guide implies 600 V or less, and high voltage implies 2400 to 34 500 V. Typical low
voltages are 208Y/120 V, 480Y/277 V, and 600Y/347 V. A typical high voltage is 12 500 V.
Low-voltage network systems have been used since the 1920s as a method of providing a highly reliable
source of electrical power to densely populated commercial areas, such as office buildings. Equipment
protection within the network vaults is typically limited. Historically, users have depended upon the physical
design of the vault to limit the risks of fault damage for faults within the vault. They have relied upon remote
detection and interruption for transformer faults and low-voltage devices, such as transformer fuse links and
low-voltage cable limiters, to provide a measure of low-voltage bus fault protection.
This guide is intended to aid those engineers who have reevaluated problems associated with faults within
network vaults, particularly for those network vaults located within or near high-rise buildings. It will also
identify currently available devices that are being used in network transformer protection schemes. These
devices should act to sense the fault and initiate fault interruption locally or remotely, thereby minimizing
damage and restoration time. These devices will be described as to their fault detecting capabilities.
Several annexes are included with this guide. Annex A describes network protector response for high-
voltage feeder faults. An example utilizing a number of protective schemes is presented in Annex B.
Annex C addresses other concepts that are currently being investigated. Annex D provides bibliographic
references, such as IEEE® Conference papers and IEEE Transaction papers, e.g., Anderson [B1].1, 2
1The mark IEEE is a registered trademark belonging to the Institute of Electrical and Electronics Engineers, Inc.
2The numbers in brackets correspond to those of the bibliography in Annex D.
Copyright © 2002 IEEE. All rights reserved. 1

IEEE
Std C37.108-2002 IEEE GUIDE FOR
2. References
This guide shall be used in conjunction with following publications. When the following standards are
superseded by an approved revision, the revision shall apply.
Accredited Standards Committee C2-2002, National Electrical Safety Code® (NESC®).3
ANSI C37.16-1997, American National Standard for Switchgear—Low-Voltage Power Circuit Breakers and
AC Power Circuit Protectors—Preferred Ratings, Related Requirements, and Application
Recommendations.4
ANSI C37.42-1996, American National Standard for Switchgear—Distribution Cutouts and Fuse Links—
Specifications.
ANSI C37.47-1981 (R1988), American National Standard Specifications for Distribution Fuse
Disconnecting Switches, Fuse Supports, and Current-Limiting Fuses.
ANSI C37.50-1989 (R2000), American National Standard for Switchgear—Low Voltage AC Power Circuit
Breakers Used in Enclosure—Test Procedures.
ANSI C37.51-1989 (R1995), American National Standard for Switchgear—Metal-Enclosed Low-Voltage

AC Power Circuit-Breaker Switchgear Assemblies—Conformance Test Procedures.
ANSI C57.12.40-2000, American National Standard for Network Transformers—Subway and Vault Types
(Liquid Immersed)—Requirements.
IEEE Std 80™-2000, IEEE Guide for Safety in AC Substation Grounding.5, 6
IEEE Std 142™-1991, IEEE Recommended Practice for Grounding of Industrial and Commercial Power
Systems, IEEE Green Book™.
IEEE Std 242™-2001, IEEE Recommended Practice for Protection and Coordination of Industrial and
Commercial Power Systems, IEEE Buff Book™.
IEEE Std C37.2™-1996, IEEE Standard Electrical Power System Device Function Numbers and Contact
Designations.
IEEE Std C37.13™-1990 (Reaff 1995), IEEE Standard for Low-Voltage AC Power Circuit Breakers Used in
Enclosures.
IEEE Std C37.20.1™-1993 (Reaff 1998), IEEE Standard for Metal-Enclosed Low-Voltage Power Circuit
Breaker Switchgear.
IEEE Std C37.29™-1981 (Reaff 1990), IEEE Standard for Low-Voltage AC Power Circuit Protectors Used
in Enclosures.
3The NESC is available from the Institute of Electrical and Electronics Engineers, 445 Hoes Lane, P.O. Box 1331, Piscataway, NJ
08855-1331, USA (http://standards.ieee.org/).
4ANSI 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/).
5IEEE publications are available from the Institute of Electrical and Electronics Engineers, 445 Hoes Lane, P.O. Box 1331, Piscataway,
NJ 08855-1331, USA (http://standards.ieee.org/).
6The IEEE standards referred to in Clause 2 are trademarks of the Institute of Electrical and Electronics Engineers, Inc.
2 Copyright © 2002 IEEE. All rights reserved.

IEEE
THE PROTECTION OF NETWORK TRANSFORMERS Std C37.108-2002
IEEE Std C37.41™-2000, IEEE Standard Design Tests for High-Voltage Fuses, Distribution Enclosed
Single-Pole Air Switches, Fuse Disconnecting Switches, and Accessories.
IEEE Std C37.60™-1981 (Reaff 1992), IEEE Standard Requirements for Overhead, Pad Mounted, Dry
Vault, and Submersible Automatic Circuit Reclosers and Fault Interrupters for AC Systems.
IEEE Std C37.91™-2000, IEEE Guide for Protective Relay Applications to Power Transformers.
IEEE Std C57.12.44™-2000, IEEE Standard Requirements for Secondary Network Protectors.
IEEE Std C57.109™-1993, IEEE Guide for Transformers Through-Fault-Current Duration.
NFPA 70-1999, National Electrical Code® (NEC®).7
3. Definitions
The following definitions are found in The Authoritative Dictionary of IEEE Standards Terms, Seventh
Edition [B9] and may aid in the understanding of this guide.
3.1 network feeder: A feeder that supplies energy to a network.
3.2 network limiter: An enclosed fuse for disconnecting a faulted cable from a low-voltage network distri-
bution system and for protecting the unfaulted portions of that cable against serious thermal damage.
3.3 network master relay: A relay that functions as a protective relay by opening a network protector when
power is backfed into the supply system and as a programming relay by closing the protector in conjunction
with the network phasing relay when polyphase voltage phasors are within prescribed limits.
3.4 network phase relay: A monitoring relay that has as its function to limit the operation of a network mas-
ter relay so the network protector may close only when the voltages on the two sides of the protector are in a
predetermined phasor relationship.
3.5 network protector (power and distribution transformers): An assembly comprising a circuit breaker
and its complete control equipment for automatically disconnecting a transformer from a secondary network
in response to predetermined electrical conditions on the primary feeder or transformer, and for connecting a
transformer to a secondary network either through manual control or automatic control responsive to prede-
termined electrical conditions on the feeder and the secondary network.
3.6 network protector fuse: A backup device for the network protector.
3.7 network secondary distribution system: A system of alternating current distribution in which the sec-
ondaries of the distribution transformers are connected to a common network for supplying light and power
directly to consumers’ services.
3.8 network transformer (power and distribution transformers): A transformer designed for use in a
vault to feed a variable capacity system of interconnected secondaries.
3.9 network tripping and reclosing equipment: A piece of equipment that automatically connects its asso-
ciated power transformer to an ac network when conditions are such that the transformer, when connected,
7The NEC is published by the National Fire Protection Association, Batterymarch Park, Quincy, MA 02269, USA (http://
www.nfpa.org/). Copies are also available from the Institute of Electrical and Electronics Engineers, 445 Hoes Lane, P.O. Box 1331,
Piscataway, NJ 08855-1331, USA (http://standards.ieee.org/).

IEEE
will supply power to the network and automatically disconnects the transformer from the network when
power flows from the network to the transformer.
3.10 spot network: A small network, usually at one location, consisting of two or more primary feeders,
with network units and one or more load service connections.
3.11 spot-network type: A unit substation, which has two step-down transformers, each connected to an
incoming high-voltage circuit. The outgoing side of each transformer is connected to a common bus through
circuit breakers equipped with relays that are arranged to trip the circuit breaker on reverse power flow to the
transformer and to reclose the circuit breaker upon the restoration of the correct voltage, phase angle, and
phase sequence at the transformer secondary.
3.12 vault: A structurally solid enclosure above or below ground with access limited to personnel qualified
to install, maintain, operate, or inspect the equipment or cable enclosed.
3.13 vault-type transformer (power and distribution transformers): A transformer that is constructed so
as to be suitable for occasional submerged operation in water under specified conditions of time and external
pressure.
4. General background
The physical and electrical characteristics of network transformer vaults have evolved over time, as have the
characteristics of the load that they serve. This clause gives a brief description of this evolution. The
application considerations of grid versus spot-network vaults and of 208Y/120 V versus 480Y/277 V
secondaries are also discussed.
4.1 Network development
The first low-voltage ac network system is reported to have been installed in Memphis, Tennessee, about
1907. The network transformers were supplied by primary feeders through distribution cutouts and were
connected to a solid grid of low-voltage cables that were protected with fuses. In Seattle, Washington, in
1921, improvements were made in the basic system by connecting the secondary terminals of the network
transformers to the solid cable grid through network protectors that would trip automatically upon reverse
power flow and were reclosed manually. In 1922, the first ac network system in which network protectors
were automatically tripped and closed by relays was placed in service in New York, NY by the United
Electric Light and Power Company. The cable grid was a three-phase, four-wire system and it operated at a
nominal voltage of 208Y/120 V. By 1925, this type of system became an accepted method of supplying
combined power and lighting load. Today’s 208Y/120 V network grid systems are very similar in
configuration and basic operation to the first systems. It is from these systems that the design, operating
practices, and overcurrent protection practices evolved for the first 480Y/277 V spot networks installed in
this country.
Grid networks served the needs of the commercial areas of many cities until the early 1950s, although it was
necessary to develop network protectors of higher continuous current ratings to carry the increasing loads.
By then, the loads in some commercial buildings were reaching the level where it was very difficult or
uneconomical for the utility to supply the load at 208Y/120 V from either a grid or spot network. These
higher loads were due mainly to air conditioning, higher lighting levels, and larger buildings. Furthermore,
the wiring costs of the building systems became very high as the loads increased at this voltage level.
Because of economics and equipment availability, higher utilization voltages, either 460Y/265 V or 480Y/
277 V, were adopted for service to large commercial buildings. Use of these higher voltages resulted in
significant reductions in wiring costs and in the cost of the utility system supplying the building load. Many

IEEE
utilities decided to serve these loads from 480Y/277 V spot networks. One utility made its first 480Y/277 V
installation in 1937, and by 1954 had eighteen 480Y/277 V vaults in service.
Figure 1 illustrates a typical network system. Many present-day network systems are installed without
network transformer high-side switches. The primary distribution substation bus may have several bus tie or
sectionalizing breakers in order to insure a reliable supply to the network. The bus tie breakers are normally
closed, but may be opened when there is an adequate source to each bus section to avoid voltage differences
between bus sections. A difference in bus voltage magnitude or angle may unload some feeders and cause
network protectors to open.
Figure 1—Typical network system

IEEE
4.2 Application considerations of adding protection to spot-network vaults versus

grid-network vaults
As noted earlier, the first network vault installations were operated primarily at 208Y/120 V and were
connected to other vaults on the low-voltage side. These vaults were generally located below grade and
external from buildings, for example, underneath sidewalks in business districts. Historical evidence
indicates that 208Y/120 V faults in these vaults were generally self-clearing. That is, faults were contained
within the vault enclosure and the locations of the vaults were sufficiently isolated from buildings to allow
for the fault to continue until enough conductive material melted, thereby extinguishing the fault-generated
arc without damaging the buildings.
As the loading in newly constructed high-rise buildings increased, the 208Y/120 V vaults were installed
inside them and dedicated to serving the load of the building. Due to the economics of scale, the building
vaults had enough capacity to supply the building load with at least a single contingency outage margin. Ties
between vaults were no longer needed or practical and the era of the spot-network vault was born.
Eventually, 480Y/277 V spot-network vaults replaced the 208Y/120 V spot networks, and during this
transition period, some small 480Y/277 V grid-network systems were installed.
These changes prompted consideration by some utilities to provide additional network vault protection.
Frequently, the damage due to arcing 480 V faults has caused lengthy power outages to buildings, more so
than at 208 V. Since many of these vaults are inside buildings, a higher incidence of damage caused by
smoke has occurred. This has prompted some utilities to provide additional network vault protection.
The principles of protection included in this guide apply equally to spot-network vaults or grid-network
vaults and at any low-voltage level. The need to consider additional protection and the associated high-
voltage interrupting devices should be evaluated on the basis of risk assessment and cost.
5. Operation of equipment in a network transformer vault under fault

conditions
A network system can be separated into three parts when considering operation under fault conditions:
a) The high-voltage primary supply circuits,

b) The network equipment, which consists of the network transformer, its associated primary-side
switch (disconnect and grounding), the low-voltage network protector, and the low-voltage bus,
c) Secondary voltage supply to the consumer and interconnecting cables to other transformer vaults, if
any.
Typical network transformer protective devices for faults in the vault and external to it are described in the
remainder of this clause.
5.1 Primary feeder faults
Referring to Figure 1, a short circuit on a high-voltage primary feeder circuit is cleared by tripping the
primary feeder breaker at the distribution substation and opening the network protectors of the transformers
connected to the faulted feeder. Primary feeder breaker tripping is usually controlled by nondirectional time
and instantaneous overcurrent relays sensing both phase and ground (residual) currents (devices 50/51 and
50N/51N). The network master relay (device 32) will trip its protector if the net three-phase power (watt)
flow is in the reverse direction, i.e., from the low-voltage bus back to the high-voltage feeder. The sensitively
set network master relay operates by sensing the reverse flow of real power caused by transformer
magnetizing and feeder load. If the feeder fault is single phase-to-ground and the network transformer is
delta connected on the high side, opening the feeder breaker will limit fault current flow to a value allowed

IEEE
by the primary system capacitance. If shunt reactors or shunt capacitors are connected to the primary
feeders, they will also influence the amount of current in the fault path. Some network transformer primaries
are connected wye grounded and will permit fault current to flow until its protector opens via the network
master relay. Some types of feeders, for example those with fused sections supplying the network unit,
require that a watt-var network master relay be applied. See Annex A for a description of the operation of the
network protector relays.
The feeder breaker and network protector may operate sequentially since the network master relay may not
operate until the feeder breaker is open. Thus, clearing time is equal to the sum of the feeder relay time,
feeder breaker time, network master relay time, protector opening time, and the arc interrupting time. Total
clearing time is usually 0.5 s or less for faults on the primary feeder excluding the case of a single, phase-to-
ground fault where the total clearing time can be higher due to the limited fault current flow after the feeder
breaker opens. Clearing feeder faults in this fashion does not result in an outage to the load served by the
network.
If the distribution voltage feeder circuit breaker fails to clear the fault, conventional breaker failure relaying
or other forms of backup relaying, such as power transformer overcurrent relays, operate to trip the feeder
breaker to isolate the fault. If the network protector fails to clear the fault, then the transformer low-voltage
current-limiting fuses or low-voltage fuse links are relied upon to melt due to the presence of line fault or
line load current feeding back through the network transformer.
Low-voltage current-limiting fuses are silver and sand devices that are designed primarily to quickly
interrupt high magnitude fault currents. The melted products are contained within a tubular enclosure where
they harden into a glass-like insulating substance called a fulgurite. Fusible links are copper or alloy devices
designed to carry high levels of transformer load current. They melt in the presence of transformer
overcurrent and can be sized to coordinate with the transformer through-fault thermal limit curve. Figure 1
shows the location of these devices in series with the network protector.
5.2 Network equipment faults
The protection of network transformers should be viewed in light of the physical and electrical environment
of the transformer. The majority of network transformers are located in vaults where space is at a premium.
This means that the usual zones of protection are not well defined. Faults in buses, protectors, and switches
may indirectly cause damage to the transformer itself.
5.2.1 Primary-voltage switch faults
Most network transformers are supplied by underground cables. These cables enter the transformer through
a compartment filled with insulating compound. On many transformers, the separate phases are attached to a
three-position disconnect switch in another insulating liquid-filled compartment. Both of these
compartments are usually welded to the network transformer tank with a cover plate bolted to the front of
the switch compartment.
Faults involving the high-voltage terminal compartment, disconnect switch, or the leads inside the network
transformer up to the primary coils will result in the same level of fault current as a primary feeder fault.
Such faults should be cleared by tripping the feeder breaker at the remote substation, opening the protector
on the faulted network unit, and opening all other weak sources to the line. These weak sources, which are
capable of sustaining voltage on the faulted line, are commonly called sources of backfeed. Sources of
backfeed are usually other transformer(s) electrically connected to a transformer connected to the faulted
line.
Faults in either compartment are generally detected quickly by the distribution substation source
instantaneous trip elements, and locally by network protector reverse power relays. Subsequent clearing,

IEEE
even by 0.25 cycle current-limiting fuses, will not insure that these relatively small compartments will not be
ruptured. Water leaking into the compartments is a common cause of these faults, as is low oil in the
compartment. Another cause of faults within the compartment is improper operator switching, such as
inadvertent grounding of the cable or opening of the transformer high-side disconnect switch under load.
5.2.2 Transformer faults
Transformer failure may result from winding-to-winding or winding-to-core faults, bushing failures, or other
conditions. Faults in the high-voltage winding of the network transformer will be cleared by the same
devices that clear faults in the high-voltage leads. However, the time required for the opening of the primary
feeder breaker will be longer for a winding fault, since a portion of the transformer impedance is in the
circuit.
Faults in the low-voltage winding of the transformer are cleared by opening the same devices. A fault in the
low-voltage winding may have to burn back into the primary winding before it is detected by the primary
feeder relays, which can result in prolonged fault clearing times. Furthermore, the ground relays for the
primary feeder will not detect faults in the secondary winding of the network transformer if it is connected
delta-wye. Experience has indicated that secondary winding faults are rare.
Transformer faults are cleared from the low-voltage bus by the operation of the network protector relays and
the network protector. In the event of a protector failure to clear a fault, the network protector fuse links are
relied upon to isolate the faulted transformer from the low-voltage bus and consumer load.
5.2.3 Protector faults
Faults within the protector itself are very difficult to clear without involving the transformer. One of the most
serious consequences of a protector fault would be damage to a network transformer tank and fracture of the
low-voltage porcelain bushings or epoxy bushings.
Arcing type faults involving the transformer low-voltage leads or the network protector may not be detected
by the primary feeder relays. Arcing faults in these locations should be detected by the network master relay,
which should initiate opening of the network protector. However, the network transformer may remain
energized from the primary side. If the fault is self-sustaining in these areas, it can cause extensive damage
to equipment until it burns to such an extent that the fault is detected by the primary feeder relays. By the
time this happens, the fault may involve the network low-voltage bus.
If a fault starts in the protector on the network side of the network protector current transformers, it may not
cause a reverse power flow in the protector, and the protector will not open. The primary feeder relays may
not detect this fault until there is considerable damage to the equipment. Sustained faults in the network
protector may be isolated from the low-voltage network by the network protector fuse links.
5.2.4 Low-voltage bus faults
The severity of low-voltage bus faults, as well as their frequency of occurrence, has proven to be dependent
on voltage level, exposure to adverse ambient conditions, type of construction, equipment age, and size of
transformers supplying the bus. All of these factors influence the self-sustaining nature of the electrical arc.
For systems operating at 208Y/120 V, there is less chance of damage or expansion of the fault into the
network transformer. Bus faults that occur on systems operated at this voltage normally clear by self-
extinguishing of the arc. This is especially true for open air constructed buses. Faults involving buses in
ducts tend to be less self-extinguishing.

IEEE
For buses operated at 480Y/277 V, arcing faults generally do not self-extinguish. They tend to generate high
arcing currents for short intervals, extinguish, and then reignite. The net result is that other equipment such
as transformers, protectors, and feeder cables eventually become involved in the fault.
The only protection for low-voltage bus faults provided by the standard network unit is the protector fuse
links. Some utilities have substituted silver and sand current-limiting fuses for these fuse links. (See 6.1 for a
description of these devices.) Both devices function best for high-current sustained faults. For example, a
sustained high-current primary feeder fault can be cleared by the fuse links before the transformer is
damaged, if the protector fails to operate. The through current can be ten to sixteen times full-load current,
depending on the number of transformers in the vault and their impedance. The drawback for this type of
protection is for arcing low-voltage bus faults. With fault current averaging not much more than load current,
these devices may operate only after tens of seconds or minutes have elapsed.
If the faulted low-voltage bus is a part of a low-voltage grid network, then the grid will also supply fault
current to the bus via intervault cables. Cable limiters located at the ends of the interconnecting cables may
melt or the fault must self-extinguish. These link-type devices are designed to coordinate with the through-
fault capabilities of the cables and as such only provide incidental bus protection. Cable limiters are most
frequently used with 208Y/120 V grid networks. Their primary function is to melt and isolate a faulted low-
voltage cable from the network grid. For faults on a primary feeder where a network protector fails to open,
the protector’s fuse links will usually melt before cable limiters in the 208Y/120 V system are damaged.
Similarly, for faults in a 208Y/120 V intervault cable, it is desired that the cable limiters operate before the
network protector fuses. For 208Y/120 V bus faults, coordination between cable limiters and protector fuse
links in other network vaults is desirable but not always possible.
5.2.5 Low-voltage faults in customer supply cables and bus
One of the more common locations of a fault is in the consumer’s switchgear or bus duct. Switch failures,
insulating board tracking, rodent and insect infestation, and water in bus ducts are common causes of these
types of faults. Damage to the consumer equipment rarely causes extensive damage to the utility vault. Low-
voltage current-limiting fuses at both ends of cables supplying the consumer switchgear limit the damage in
the cable section of the circuit. There may be four to eight cables per phase for each consumer supply
between protectors and collector bus. These devices should coordinate with the network protector fuse links
or current-limiting fuses.
Current-limiting fuses will generally operate to interrupt high-current consumer faults. However, for low-
current arcing faults, especially at 480Y/277 V networks, they have proven to be ineffective.
5.2.6 Other faults
Fuse links mounted externally to the transformer protector have been a location for arcing faults on 480 V
networks. Some fuse mounts are above the transformer protector and have a phenolic insulating board
common to all phases. Evidence has shown that tracking has occurred between phases on this type of board
due to water and contamination. Isolated phase construction of fuse mounts, or the use of interphase barriers,
greatly increases the tracking distances, and hence decreases the probability of flashovers between phases.
Alternatively, a barrier built above the fuse mounts would aid in reducing contamination. Enclosing the fuse
links may prevent arc extinction for protector faults and result in additional equipment damage.
Other causes of vault faults include metering current transformer failures, high-voltage cable failures,
insulation breakdown, and inadequate grounding. Insulation is exposed to various stress factors, such as
thermal aging, electrical, mechanical, and environmental stresses. Utility and consumer vault ground grids
should be properly connected to avoid cable and equipment damage when fault current seeks a path to return
to the transformer on 480 V networks. Refer to NESC, IEEE Std 80-2000, and IEEE Std 142-1991 for
guidance on proper grounding.

IEEE
5.2.7 Primary-voltage cable faults
Primary-voltage cable faults have been attributed to many different causes, including improperly installed
stress cones, where the primary-voltage cable comes into the high side of the transformer. Some failures
have been caused by bad cable splices and normal service deterioration factors. There have been failures
attributed to racking installation problems, where the cable has mechanically rubbed or chafed to the point
of completely removing the external insulation jacket. Failed bail are another source of cable failure in that a
retaining spring may corrode or weaken, allowing the bail to become partially or completely disengaged.
Common elbow failures include cross-threaded pins and corona resulting from unrelieved pressure from
improperly racked cables.
6. Methods available to provide improved network transformer protection
Clause 5 summarized the response of the typical network transformer protective devices to faults both
internal and external to the vault. This clause describes additional protection techniques that can be used to
promptly detect faults occurring within the vault and which can then be used to clear the faults by devices
such as those referred to in 7.1, 7.2, and 7.3 of this guide.
6.1 Electrical detection of faults
6.1.1 Transformer differential relays
Current differential relaying can be used to protect network transformers. The relays are connected to
current transformers on the high side and low side of the network transformer. The net operating current to
the relays is the difference between input and output currents to the network transformer zone of protection.
(See IEEE Std C37.91-2000 for details on relay types and current transformer connections.) Differential
relaying provides a clearly defined zone of protection and is capable of sensitively detecting both multiphase
and single phase-to-ground transformer faults.
Special protection concerns for protecting network transformers with differential relays are as follows:
a) There are generally no high-side current transformers on existing network transformers.

b) Limited vault space may make it difficult to add current transformers and/or high-side interrupting
devices.
6.1.2 Overcurrent relays and fuses
A fault external to a transformer can result in damage to the transformer. If the fault is not cleared promptly,
the resulting overcurrent on the transformer can cause severe overheating and failure. Overcurrent relays
and/or fuses may be used to detect and clear the transformer from the faulted bus or line before the
transformer exceeds the thermal or mechanical damage limits. On some transformers, overcurrent relays
and/or fuses may provide protection for internal transformer faults. (See IEEE Std C37.91-2000 for details
on applying time-overcurrent and instantaneous overcurrent relays.) Overcurrent relays, when applied to the
high-voltage side of the network transformer, are an economical means to provide detection of a transformer
fault. High-current faults can be detected by instantaneous overcurrent relays. Time-overcurrent devices may
protect for low-side faults and may provide backup protection for downstream devices provided that these
faults are not high-resistance arcing and low current. Overcurrent devices can be an alternative or
supplemental protection to other forms of protection.
Special protection concerns for protecting network transformers with overcurrent relays are as follows:
a) There are generally no current transformers on existing transformers.

IEEE
b) Limited space may make it difficult to add high-side interrupting devices.

c) Arcing low-side fault currents may not be of sufficient magnitude to operate the overcurrent relays.
However, there are several new high impedance/downed conductor/arcing detection relays being
developed with potential for application here.
d) Coordination should be maintained with the network master relay for primary feeder faults.
Special protection concerns for protecting network transformers with fuses are as follows:
a) Undetected operation of only one fuse will result in a single-phase supply to the network trans-
former. Other transformers feeding the secondary system will hide the single-phased condition of
the transformer since three-phase service is still provided to the load. The single-phase condition on
the network transformer leads to tank and core overheating due to flux imbalance, and this can lead
to transformer failure. In general, simple fuse protection of network transformers is not recom-
mended for this reason alone. However, fuses may be applied, with protection sensing enhancements
such as external means, to detect the single-phase condition with an interrupting device that can
interrupt all three phases when this condition exists.
b) Limited space may make it difficult to add high-side fuse interrupting devices, especially those with
schemes that can detect single-phase load conditions to operate a switch to open all three phases.
c) Concerns (3) and (4) for overcurrent relays apply to fuses also.
6.1.3 Ground-fault relays
Sensitive detection of ground faults can be obtained by differential relays or overcurrent relays specifically
applied for that purpose. Several schemes are practical, depending on transformer connections, availability
of current transformers, zero-sequence current source, system design, and operating practices. Annex B
provides examples of ground-fault relay applications to network transformer protection.
Referring to Figure B.1 in Annex B, high-voltage ground-fault protection for transformer number one is
accomplished with a sensitively set overcurrent device, 51N-T1, and an electronic fuse. The electronic fuse,
described in 7.4, provides instantaneous protection for high-current faults. (See Figure B.2 in Annex B for
the time versus current coordination curves for these devices.) If an electronic fuse or backup current-
limiting fuse (also referred to as a partial range current-limiting fuse, see ANSI C37.47-1981) is not used,
then an instantaneous relay, device 50N, should be used. A special protection concern for high-voltage
ground-fault protection is that these faults may involve all three phases before the protective relays can
operate.
Low-voltage ground-fault protection can be advantageous in network transformer protection. This is

particularly true on 480Y/277 V networks since ground faults are typically low-current, high-resistance
arcing faults. These faults can be at or below load current in magnitude and may be intermittent or restriking
in nature. The installation of current transformers (CTs) is easier on transformer neutral bushings than on the
transformer phases, and of course, fewer in number.
Low-voltage ground-fault devices can be set more sensitively than phase-overcurrent devices. A protective
relay connected to a transformer neutral, such as device 151N-T1 in Figure B.1, will detect all the ground-
fault current and the unbalanced load current returning to transformer one. It should be set above the
maximum unbalanced load that appears as neutral current. As the unbalanced current approaches full-load
current, the setting of this device approaches that of a phase overcurrent device, and therefore, becomes less
effective.
Another approach for ground-fault current detection is to sense current in the grounding conductor of the
transformer, as demonstrated by device 151G-T1 in the Annex B example. Ideally, the only grounds on the
low-voltage network system are at the transformers. If this is the case, and if all single-phase load is
connected phase-to-neutral, then device 151G-T1 can be set very sensitively since only fault current will be
detected. If the neutral is grounded at more than one point as shown in Figure B.1, which is a 1978 and

IEEE
beyond code requirement (NESC), then this is not as effective. First, a portion of unbalanced load current
may appear in the ground conductor, which requires that the device be set less sensitively. This unbalance
can only be reliably determined by test. Second, due to the shunting effect of the multigrounded neutrals, not
all of the downstream fault current will return on the transformer neutral grounding conductor. This
improves coordination with downstream devices, but also desensitizes the relay. However, for faults in the
utility vault, multigrounding of consumer neutrals should have minimal effect, i.e., the 151G-T devices will
detect almost all of the ground-fault current.
The last ground-fault detection method shown in Figure B.1 is a transformer case ground device, 64GF. This
protection can only be used if the transformer case can be grounded in such a way that all the fault current
can pass through the 64GF’s CT. This includes isolation from high-voltage cable sheaths, concrete
reinforcement bars, etc. Using this protection, all transformer internal faults to ground can be detected
quickly.
Special ground-fault protection concerns for network transformers are as follows:
a) Depending on the relay’s CT location, multiple ground paths between consumer and utility may
shunt some relay current.
b) Transformer neutrals must be isolated from the transformer tank.
c) Faults may involve all three phases before the relays can operate.
d) Relays may have to be set above the maximum unbalanced load current. As a result they may not
operate for arcing faults.
Faults on 480Y/277 V collector buses are especially troublesome because of the non-self-clearing nature of
faults at this voltage and the extreme insensitivity of the protection otherwise provided for these buses. Low-
magnitude burning faults on collector buses will not be detected by supply cable limiters or protector fuses
and will burn to destruction. Improved protection of collector buses can be achieved by ground-fault
schemes of two general types:
a) By electrical isolation and a single-point grounding of the bus enclosure. A current transformer in
the ground lead connects to a sensitive overcurrent scheme, the operation of which is wired to open
the protectors on all the transformers supplying the collector bus. This uses the remote trip contacts
available on the network protectors and is powered by the building service. Since no current should
ever flow in the grounding bond lead of the bus duct enclosure, this scheme can be set very sensi-
tively, responding to ground-fault currents of a few amps.
b) By installing current transformers in the leads of all network transformers and using the residual cur-
rent to operate an overcurrent relay for each transformer. The overcurrent relay for each transformer
operates its associated protector. This scheme is less sensitive than the isolated bus duct described
above, but it is easier to construct. It can also detect imbalanced load current and must not operate
for it. A setting of 1000 A per residual relay will coordinate with National Electrical Code® (NEC®)
required customer ground overcurrent trip devices and will afford improved protection for the col-
lector bus. For enhanced security of service and to avoid unwanted tripping, the overcurrent scheme
should operate independently for each transformer supplying the collector bus.
6.1.4 Distance relays
A distance relay that has a separately adjustable phase to phase unit (see Sonnemann, et al. [B22]) can be
used to supplement the overcurrent relaying on the substation feeder to supply protection for several network
transformers on that feeder. Because the phase-to-phase element exhibits immunity to balanced three-phase
load current, it may be possible to detect a bolted phase-to-ground fault on the secondary of the smallest
network transformer on the feeder after its network protector has opened.
Precautions to use with this very sensitive setting are as follows:

IEEE
a) Provide sufficient time delay to coordinate with downstream protective devices.

b) Protect for a loss of voltage that may operate the distance relay.
One utility uses a time-overvoltage relay connected phase-to-phase across an open delta voltage
transformer with its contact in series with the distance relay contact. Operation of the distance relay
energizes the time-overvoltage relay, which then operates to trip the feeder after a time delay. The voltage
tap on the time-overvoltage relay is selected so that an operation does not occur if either of the two voltage
transformer fuses has blown.
6.1.5 Network protector relays
The typical design for network protector relays has predominantly consisted of two mechanical induction-
cup devices. The two electromechanical relays are the network master relay and the network phasing relay.
The master relay contains both the trip and close functions. When the network protector is closed, the master
relay closes a trip contact after sensing reverse power flow. The phasing relay supplements the close function
of the master relay by only allowing closing of the protector when the transformer secondary voltage leads
the network voltage. The master and phasing relays work together to make sure the network voltages are of
the right magnitude and the correct phase relationship before allowing the network protector to close. The
main problem with this design was that they were not designed to be submersed in water. The most common
failure mode for these types of relays is due to humidity and direct water damage.
Another common problem with mechanical relays is that if the protector is not properly pressurized, then the
relay contacts may slightly oxidize to the point where the protector may stay closed. If the network protector
operates frequently, it is also possible for the contacts to become pitted.
The problems with the electromechanical relays led to a new generation of solid state analog network
protector relays. The solid state analog network protector relay is a single relay replacement for both the
master and the phasing relay and could be used in the applications where it was fully submersed in water.
The latest network protector relay design is a microprocessor controlled network protector relay. This type is
a single relay, three-phase device. The microprocessor design provides added features, such as both remote
monitoring capabilities and communications and control functions. The opening and closing of the protector
can now be controlled remotely by use of the utility’s Energy Management Center Supervisory Control and
Data Acquisition (SCADA).
6.2 Mechanical detection of faults
One method for detecting transformer faults other than by electric measurements is by the increase in tank
oil or gas pressures caused by internal transformer faults. Sudden-pressure relays using this method are
valuable for transformers that are not well suited to differential relaying. These relays may be more sensitive
for certain internal faults than relays that are dependent upon electrical quantities, and thus can be very
valuable in minimizing transformer damage due to internal faults. Sudden-pressure or rapid-pressure rise
relays do not require addition of CTs, and some types can be retrofitted on existing network transformers.
Some utilities use a separate lockout relay, Device 86, with the sudden-pressure relay to provide redundancy.
Special protection concerns for protecting network transformers with any of the above sudden-pressure
relays are as follows:
a) Some types of sudden-pressure relays may operate for close-in through faults. The use of sudden-
pressure inhibit schemes may alleviate this problem.
b) Sudden-pressure relays may be subject to misoperation during seismic disturbances. The use of seis-
mic qualified relays may be used to reduce this concern.

IEEE
c) Limited space may make it difficult to add high-side interrupting devices that the mechanical detec-
tion devices must operate to clear a faulted transformer off line.
6.2.1 Sudden-pressure relay (gas)
The sudden-pressure relay (gas) is applicable to all gas-cushioned, oil-immersed transformers and is
mounted in the region of the gas space. It consists of a pressure-actuated switch, housed in a hermetically
sealed case and isolated from the transformer gas space except for a pressure-equalizing orifice.
The relay operates on the difference between the pressure in the gas space of the transformer and the
pressure inside the relay. An equalizing orifice tends to equalize these two pressures for slow changes in
pressure due to loading and ambient temperature change. However, a more rapid rise in pressure in the gas
space of the transformer due to an internal fault results in operation of the relay. High energy arcs evolve a
large quantity of gas that operates the relay in a short time. The operating time is longer for low energy arcs.
6.2.2 Sudden-pressure relay (gas/oil)
An available design of the rapid-pressure rise relay utilizes two chambers and two control bellows along
with a single sensing bellows. All three bellows have a common interconnecting silicone oil passage with an
orifice and ambient temperature compensating assembly inserted at the entrance to one of the two control
bellows. Separate versions are available, one for oil pressure and one for gas pressure.
An increase in transformer pressure causes a contraction of the sensing bellows, thus forcing a portion of its
silicone oil into the two control bellows and expanding them. An orifice limits the flow of oil into one control
bellows to a fixed rate, while there is essentially no restriction to flow into the second control bellows. The
two control bellows expand at a uniform rate for gradual rate of rise in pressure, but during high rates of
transformer pressure rise, the orifice causes a slower rate of expansion in one bellows relative to the other.
The dissimilar expansion rate between the two control bellows will cause a mechanical linkage to actuate the
snap-action switch, which initiates proper tripping.
6.3 Thermal detection of faults
6.3.1 Heat detection
One method of detecting arcing faults, which can be very effective, is a heat-sensing system. Arcing faults
on the 208Y/120 V systems are of lesser concern because the arc at these voltages is generally self-
extinguishing. On the 480Y/277 V system, an arcing fault can be a major problem since the arc tends to
sustain itself. Arcing faults are difficult to sense because they are usually of low-magnitude current and are
not rapidly cleared by conventional overcurrent devices. The energy in an arcing fault does, however,
generate tremendous amounts of heat in an extremely short time. A heat-sensing system located inside
network protectors, near bus work, and near cabling, can be used effectively to sense arcing faults. A circuit
can be designed that can alarm and trip once an abnormal condition is detected, using heat probes, eutectic
tubes, etc. Staged fault tests by one utility, see Griffin, et al. [B7], have demonstrated that these devices
operate very quickly, comparable to differential relaying. A special protection concern requires that the heat
detection device be positioned close enough to the protected equipment to ensure quick operation for arcing
faults.
6.3.2 Smoke detection
Smoke detection can be used to alarm or trip once a situation becomes severe enough to cause combustion
and generate sufficient smoke. However, the protection concern of this type of detection system is that it may
be too slow to prevent major damage.

IEEE
6.4 Ultraviolet detectors
Ultraviolet detectors have been employed with some success by one utility as a sensitive means of detecting
arcing fault conditions within network vaults, see Roop, Vidonic [B17]. Modern ultraviolet detection
systems are highly reliable, fast, and flexible due to their use of fiber-optic cables for the signal
transmissions. The detectors may be strategically positioned within the vault, on the collector buses, and
within the network protector to sense ultraviolet light emanating from the arcing fault.
A typical ultraviolet detection scheme utilizes very sensitive optical detectors that operate for light with a
wavelength of between 185–245 nm, which corresponds to the arc and firelight emitted from a fault.
Selection of such an operating range provides security against misoperations due to normal switching arc
light, as well as sunlight and artificial sources of light, such as incandescent and fluorescent lights. This
range does not preclude operation of the scheme in the presence of ultraviolet light, such as that caused by
lightning, camera flashes, direct sunlight, and switching arcs in large apparatus.
The arc detector scheme can be made secure against such misoperation by sealing the vault against spurious
sources of ultraviolet light, adding time delay, or adding a ground-fault relay to the scheme. A major
protection concern of this scheme is that it will not detect faults internal to major equipment, such as
transformers and network protectors, until the fault becomes externally visible.
7. High-voltage, fault-interrupting devices
The function of a protective device applied on the primary side of a network transformer is, in general, to
provide system protection as well as transformer protection. With respect to system protection, the primary-
side protective device should interrupt a potentially damaging overcurrent condition and operate promptly to
isolate only the faulted segment, thereby minimizing the stresses on the remainder of the system and limiting
the extent of the service interruption. For transformer protection, the primary-side protective device should
operate promptly in response to a fault located between the protective device and the transformer. It should
further provide backup protection for the transformer in the event the low-voltage network protector or other
secondary-side protective device fails to operate properly. Some applicable devices are circuit breakers,
vacuum fault interrupters, and fuses. This clause will describe the presently available power equipment used
for fault current interruption and then discuss some application and coordination considerations of each type
of device. See Annex C for devices under development.
7.1 Circuit breakers
The conventional circuit breaker equipped with time and instantaneous overcurrent relays can be used to
provide phase and ground-fault detection. In addition, the breaker can be remotely tripped, thus providing
protection and isolation for externally detected faults (Clause 6). Applications of circuit breakers located at
the network transformers have been limited due to the relatively high cost of a breaker and its large size,
which frequently cannot be accommodated in a vault. The use of vacuum fault interrupters may overcome
the size problem.
7.2 Remotely located circuit breakers with transfer tripping
Circuit breakers remote from the network vault equipped with phase and ground overcurrent relays set to
coordinate with network transformer secondary faults are used successfully to isolate a transformer faulted
within its high-side switch compartment or primary winding. If secondary winding, protector, and low-
voltage bus faults require tripping of the remote circuit breaker, then transfer tripping is required.

IEEE
The installation of transfer trip equipment, transmitter and receiver, is not itself difficult or particularly
expensive. The installation of a communication link, however, can be a problem. Difficulties include
a) Insufficient duct space to pull all the needed communications cables to the remote substation(s)
b) Susceptibility of metallic pairs to induction problems caused by close proximity to the power cables.
One option may be to install fiber-optic cables. By multiplexing signals, the number of communication
cables that need to go back to the remote substation(s) would be reduced. Also, fiber-optic cable could be put
into the same ducts as the power cables, since it would not be affected by proximity to the power cables.
7.3 Fault interrupters
The development of the vacuum fault interrupter has led to the introduction of a new generation of fault-
interrupting devices. Interrupting fault current in a vacuum has found application not only in conventional
circuit breakers, but has led to development of the vacuum fault interrupter, which is typically a single-shot
device without automatic reclosing capabilities. Since automatic reclosing is normally not required for
transformer protection, this concept allows design of a protective device providing smaller size, lower cost,
and simpler operation. The smaller size and simplicity of operation allows a submersible type of
construction.
Vacuum fault interrupters utilize short operating strokes; thus a smaller operating mechanism is required.
The insulating medium surrounding the vacuum interrupter can be solid, air, SF6 gas, or conventional
transformer oil. It should also be noted that in some vacuum interrupter designs, a special external
encapsulation process is applied to enhance the external dielectric performance, particularly for applications
requiring high BIL ratings or operation at high altitude. Interruption in vacuum also results in a quiet
operation ideal for a vault-type installation. Vacuum fault interrupters are small in size, lightweight, and can
be mounted on the floor, wall, and even the ceiling of the transformer vault. Conventional phase and ground
overcurrent sensing and tripping can be provided by electromechanical relays or static overcurrent trip
devices. In addition, the vacuum fault interrupters can be remotely tripped from external protective devices
(Clause 6). Their fast operation (approximately 2 cycles) at the higher fault currents reduces damage.
Other interrupting devices that exhibit the same desirable qualities as vacuum fault interrupters are
becoming available. For example, SF6 fault interrupters are becoming available, which offer similar
advantages of small size, low cost, excellent switching and fault-interrupting performance, and oil less
construction. These automatic fault interrupters are designed and manufactured in accordance with
IEEE Std C37.60-1981.
7.4 Power fuses and current-limiting fuses
Power fuses are commonly used for protection of outdoor distribution and substation transformers, and they
can be applied for network transformer protection as well, because of their fast response characteristics at
high-fault currents and minimal maintenance. There are two basic types of power fuses: solid material fuses
and current-limiting fuses. Although solid material power fuses can typically be selected to have smaller
ampere ratings than similarly applied current-limiting fuses, they may not be appropriate for network
transformer protection application where current limitation to prevent tank rupture is desired. Current-
limiting fuses are advantageous in network transformer protection applications, since they clear high
magnitude faults completely within the first one-half cycle, limiting peak current and I2t let-through to less
than the circuit would deliver if either solid material power fuses or circuit breakers were used.
Another type of power fuse, the electronic fuse, combines many of the features and benefits of both types of
power fuses and even relays, see Glenn [B6]. The electronic fuse has a 600 A continuous rating and can
interrupt 40 kA. The electronic fuse consists of two separate components: an electronic control module that
provides the time-current characteristics and the energy to initiate tripping; and an interrupting module that

IEEE
interrupts the current when a fault occurs. These two modules, when joined together, fit in a suitable
mounting. A CT powers two types of logic circuits employed in the control module—one with instantaneous
tripping characteristics and one with time-delay tripping characteristics. These two circuits may be used
alone or in combination to provide a variety of time-current characteristics. When a fault occurs, the control
module triggers a high-speed gas generator that separates the main current path in the interrupting module,
transferring the current into the current-interrupting ribbon elements, which then melt and burn back.
The most important principle to be considered when selecting a transformer primary fuse is that it must
protect the transformer against damage from mechanical and thermal stresses resulting from through faults
that are not promptly interrupted. A properly selected fuse will clear such faults before the magnitude and
duration of the overcurrent exceeds the short-time loading limits recommended by the transformer
manufacturer. In the absence of specific information applicable to an individual transformer, the primary
fuse should be selected in accordance with recognized guidelines for the maximum permissible transformer
through-fault current duration limits. Curves representing these limits and information pertaining to their use
can be found in IEEE Std C57.109-1993 and IEEE Std 80-2000.
Besides selecting a transformer primary fuse to maximize protection for the transformer, it is also important
for the time-current characteristics of the transformer primary fuse to be coordinated with the time-current
characteristics of certain other overcurrent protective devices on both the secondary side and the primary
side of the transformer. Annex B illustrates typical coordinated network installations.
There are a number of items that need to be considered when applying fuses to network transformers:
a) Since fuses are typically sized to carry phase currents, including single contingency overloads, they
will probably not respond very quickly to low-magnitude ground faults. This lack of ground-fault
sensitivity may create difficulties when coordinating fuses with other protective devices that utilize
such sensing.
b) Since fuses are single-phase devices, operation of a single fuse can result in a single-phase condition
of the transformer. Relays are available, however, that can detect this condition and initiate remote
tripping of a three-phase switching device after a time delay measured in seconds. The device, 51/46
in Annex B, is such a relay.
c) In below-grade street vaults, care should be used to preserve the submersibility of these units.
d) Any form of localized time-overcurrent protection or fusing must coordinate with the network trans-
former low-side fuse links or current-limiting fuses.
e) If high-side current-limiting fuses are to be applied to reduce the risk of tank rupture, they must be
conservatively sized to account for all transformer vault emergency loading situations, cold load
pickup, and transformer inrush.
Generally, application of current-limiting fuses to network transformer protection should be restricted to

partial range (backup) fuses capable of current-limiting only for high-side, high-current faults. (See
IEEE Std 142-1991 for the definition of this device.) In this mode of operation, the user gains speed in high-
current clearing from the high-voltage feeder side of the fault. However, high current will continue to flow
from the low-voltage side after one fuse melts even for single, phase-to-ground faults with a delta high-side
connection. Watt-type protector relays may respond slowly or not at all due to the operation of a high-side
fuse for a transformer fault. As a result, a three-phase opening device on the primary should be used in
conjunction with partial range current-limiting fuses.
Fuses are designed and manufactured in accordance with IEEE Std C37.41-2000 and ANSI C37.42-1996.

IEEE
8. Low-voltage, fault-interrupting devices
8.1 Network protector
The switching device between each network transformer low-voltage bushing and the network is the network
protector. A secondary network protector consists of an electrically operated, low-voltage air circuit breaker
and network relays with associated equipment for automatic circuit breaker control. A network protector
may be flange mounted directly on the network transformer low-voltage terminals or may be separately
mounted.
Network protectors are available in submersible enclosures for installation in underground vaults or in
nonsubmersible enclosures where no possibility of flooding occurs. They are available in continuous current
ratings for use with network transformers ranging from 225 kVA to 2500 kVA and for network voltages of
208Y/120 V, 480Y/277 V, or 600Y/347 V. Network protectors are rated to interrupt low-voltage fault current
and are designed and manufactured according to IEEE Std C57.12.44-2000.
8.2 Low-voltage fuses
Fuses are usually installed at the output terminals of the network protector to provide backup protection for
the protector breaker. The fuses should not blow on a primary feeder fault before the network relays trip the
protector. If the protector breaker fails to open for a fault on the primary feeder cable, the protector fuse
should be coordinated to blow before the overcurrent relays on the other primary feeders supplying the
network can operate (assuming the remote feeder relays can see the fault). If this is not possible, then
additional protection should be considered.
Network protector fuses are commonly available in two styles: either a low-voltage fuse link or a low-
voltage current-limiting fuse. The fuse link is either a copper fuse or an alloy fuse and is located either
within the network protector enclosure, on the load side of the main contacts, or mounted externally above
the protector. If they are located within the protector enclosure, they can contribute significantly to the heat
within the enclosure and could limit the ability of the protector to carry emergency loads, such as would be
the case if one or more transformers supplying power to the bus were out of service. The alloy fuse link
characteristic varies widely with load and ambient temperature. There have been instances where these
devices have not melted for arcing faults within the protector.
The second type of fuse commonly used with network protectors is the low-voltage, current-limiting fuse. It
is especially applicable on 480 V networks because its arc products are completely confined within its tube.
It is sensitive to temperature and therefore is mounted in housing separate from the network protector
enclosure. At low levels of overcurrent, these devices may become damaged without complete melting.
Subsequent transformer loading can then cause them to overheat, open, or catch on fire. A new version of the
current-limiting fuse that is electronic is now available as a replacement for the Y-link device. It is
triggerable and thus could be initiated by the network protector lockout relay, which in turn was triggered by
the protector heat probes. The advantage of such a device is that the faulted protector could be isolated from
the 480 V bus without the need to trip any other transformer protectors supplying the bus. If the involved
transformer had a high-side interrupting device, then it could be completely isolated inside the vault for a
network protector fault without the need to drop network load.
9. Distributed source generation
A network protector device should not be used as a separation device between a constant frequency network
system and a distributed source of generation. Network protectors were not designed or tested to handle the
recovery voltages involved when interrupting load currents or fault currents between two points that are not

IEEE
locked together in synchronism. A review of IEEE Std C37.13-1990, ANSI C37.16-1997, IEEE Std
C37.20.1-1993, IEEE Std C37.29-1981, ANSI C37.50-1989, ANSI C37.51-1989, and IEEE Std C57.12.44-
2000 was made to check for wording related to generator usage. Only IEEE Std C37.13-1990 and
IEEE Std C37.29-1981 had wording specifically aimed at using generators with circuit breakers.
IEEE Std C57.12.44-2000 is specifically for network protectors and the case may be made that the network
protectors should be tested to the same standards as circuit breakers for generator operation. No evidence
indicates that to be the case. The power factors used for the test circuits are not the same.
Utility consumers with noninterruptible loads (e.g., broadcast stations, telephone companies, and hospitals)
have applied to utilities for permission to attach their own emergency generators to distribution networks by
using make-before-break (closed-transition) transfer switches. Such transfer switches would permit periodic
testing of emergency generators and their intended loads, without interrupting that load. Other utility
consumers have requested their peak shaving generators be allowed on a network protected system. This
testing duty or daily use would subject network protectors on the network to potential reverse power
conditions, for which they are not designed. Network protectors are designed to interrupt fault current flow
to a feeder fault from the low-voltage network, but are not designed to separate two dynamic systems—
utility- and consumer-owned generation. Network protectors with transient voltage rating sufficient to
withstand separation of two dynamic systems are generally not available.
Network protectors could be forced to operate more times than designed as the on-site generation responds
to the varying load. It is possible that the generator controls may not respond quickly enough to changing
load patterns and allow power to flow in a reverse direction out onto the power supply such that the protector
opens. Thus, it is possible that the protector may be standing open for long periods of time. Should a
generator fail during this time frame, then the consumer could be de-energized until the network protector
closes back in. Consumers with noninterruptible loads would thus be affected greatly.
Phase and voltage differences between utility network and consumer generation may cause network
protectors to open when they cannot interrupt current flow, or when large voltage or phase angle differences
can arise, resulting in failure. Such incidents have occurred.
Generally, utility practice is to disallow such closed-transition transfers, when network protectors may
operate as a result of the transfer.
Network protector testing and qualification is for network circuits with X/R ratios between 5 and 7, whereas
when applied to generator circuit applications, the circuit X/R ratio may typically be in excess of 20. This
has the effect of substantially increasing the dielectric stresses during circuit interruptions, which will
expose the network protector to voltages beyond its rated limits. Therefore, no transition should be allowed
unless the consumer’s generator output matches the utility voltage in both phase and magnitude, and even
then, the transition must be kept substantially shorter than the 0.05 s operating time of most network
protector relays. A manufacturer recommends against closed-transition transfers.
A battery-based uninterruptible power supply, if implemented to carry consumer load during an open-
transition transfer for periodic testing, would also be available to carry load in the event of failure of the
utility network and subsequent startup of consumer emergency generation, thus improving reliability of
consumer emergency generator backed-up power supply overall.
For cases where distributed generation is connected in parallel with a network system, appropriately rated
interrupting devices built per circuit breaker standards with full duty interrupting current and voltage ratings
should be utilized at all places between the distributed generation and the utility’s high-voltage cable. These
interrupting devices may require a relay and control battery or capacitor trip device. The duty of standard
network protectors is not adequate for placement in systems with energy sources connected on what is
normally their load side.

IEEE
10. Network distribution SCADA
As has been stated, underground grid and spot-network systems offer much greater reliability than other
commonly used distribution systems. Because a network system is designed for maximum service
reliability, maintaining power to consumers, even if one of its primary feeders or transformers fails, there is
no obvious indication either to the consumer or utility that there is a malfunction. There is a need to prevent
total network shutdown from secondary faults, to prevent a catastrophic failure in a building vault, and to
minimize damage to consumer and utility facilities. Until SCADA systems became practical, periodic visual
inspections were required, as well as preventive maintenance, to detect incipient failures. Practically
speaking, inspections cannot be frequent enough to detect defective equipment. Besides the relaying, sudden
pressure relays, and other automatic protection schemes described elsewhere, active remote monitoring of
the system is a method that should be considered. It has been demonstrated that SCADA systems can be
used to monitor networks to ensure that their inherent reliability does not become a significant risk factor
(see Landman, Louie [B13]).

IEEE
Annex A
(informative)
Response of network relays to system faults
A.1 Tripping characteristics
If a fault occurs on the primary feeder, and/or if there is a power reversal from the network to the primary
system, the network protector opens automatically. Figure A.1 shows that when the station breaker is opened
to isolate a feeder, the current from the network may be either exciting current or exciting current plus
charging current. The exciting current will be relatively small and will lag the reversed network voltage
phasor by an angle in the order of 65° to 80°. Cable charging currents may be quite large, particularly for the
higher feeder voltages, and the net current may lead the reversed network voltage phasor by an angle of 85°
or more. The watt relay characteristic is designed so that the relay will trip for either of these conditions.
Charging current should not exceed 250% of the protector rating, as the protector may not detect this high
magnitude leading current due to current transformer saturation.
Figure A.1—Typical tripping characteristics of master relay
A.2 Balanced faults
Primary feeder three-phase faults will produce short-circuit currents on each phase. Since this phasor falls
within the trip area of the master relay, the relay will open the network protector. Short-circuit currents

IEEE
caused by three-phase or single-phase faults on the secondary grid fall on the right side of the relay
characteristic and do not cause the protector to be tripped.
A.3 Tripping operations on unbalanced faults
If the primary winding of the transformer is connected in delta and if no other sources of zero-sequence
current are connected to the line, fault current will not be supplied from the network for a line-to-ground
fault on the high-voltage feeder after the feeder breaker opens. If the station breaker clears for such a fault,
the network then backfeeds into the transformer the exciting current of the transformer plus an unbalanced
charging current because of the grounded conductor. Net power flow is out of the network and the master
relay will operate to trip the protector. If the transformer primary should be connected in grounded wye, the
master relay will receive this fault current and will trip the protector, if net power flow is out of the network.
If high-side fuses are provided for the transformers of a grid network, high-current faults in the high-voltage
winding or cable on the load side of the fuses will usually cause the feeder to trip even though one or more
fuses may melt. The standard watt characteristic network master relay will then operate to open the network
protector from the transformer-exciting current or cable-charging current.
Spot networks may be served from nondedicated feeders, which also serve other loads. Fuses may be
installed at the primary terminals of the network transformers to prevent lockout of the feeder for a network
transformer fault and avoid interruption to the other loads. Where high-fault currents are available, the fuses
may clear severe faults on the transformer primary leads faster than the feeder breaker, limiting the energy to
such a fault. After the feeder breaker opens, the standard watt characteristic network master relay will open
the network protector due to the reverse power flow to the other loads (with one fuse blown) and exciting
current.
If, however, the fuse clears a single, phase-to-ground fault on the transformer leads or in the high-voltage
winding without tripping the feeder breaker, the unfaulted phases may still supply power to the network. The
net three-phase power flow in the network protector may not be in the reverse direction and may not operate
the watt-connected master relay. However, the reactive flow in the network protector will be in the reverse
direction. A protector master relay connected to provide watt-var characteristics, so that maximum torque
occurs when the current leads the network line-to-neutral voltage by 120°, will operate for this condition.
The tripping characteristic is shown in Figure A.1.
A second reason for recommending the watt-var relay for the spot-network protector when the transformer
primary is fused is to obtain faster protector operation for faults on the primary feeder. This may avoid
possible fuse blowing from the network backfeed. The watt-var characteristic network master relay develops
more torque and will operate faster than the watt-connected relay for the inductive flow to feeder faults.
It should be noted that the watt-var relay will not operate on the capacitive charging current of the primary
cable when the feeder breaker has been opened in the absence of a fault, but requires the inductive load flow
to the other connected loads to operate. If the spot-network nondedicated primary feeder is largely overhead,
the load flow would be primarily inductive unless the feeder was overcompensated with shunt capacitors for
power factor correction. Under fault conditions, either multi or single phase, the newer microprocessor
relays can trip successfully, whereas the older analog or electromechanical relays would not respond to the
highly capacitive backfeed circuits.
New developments include a master relay that exhibits watt characteristics at the low transformer-exciting or
cable-charging current levels and watt-var characteristics at the higher fault current levels.

IEEE
A.4 Reclosing characteristics
The master relay recloses an open protector when normal conditions return on the high-voltage feeder.
A typical closing characteristic of a network master relay is shown in Figure A.2. If the voltage difference
phasor terminates to the right of the master relay close characteristic, the relay will operate its closing
contacts and the network protector will close if the phasing relay close contacts are also closed. If the
transformer voltage is low enough to cause the difference phasor to terminate to the left of the relay
characteristic, the protector will remain open.
Figure A.2—Typical closing characteristics of master and phasing relays
The relay characteristic is normally spaced away from the tip of the network voltage phasor by a relay
setting of 1.5 V in 208Y/120 V systems, which is referred to as the master relay close setting. The
corresponding setting in a 480Y/277 V system is 3.4 V at 0°.
The characteristic shown is for one phase of a three-phase system. All three phases have identical
characteristics.
If the re-energized transformer voltage should lag the network voltage, reclosing would cause power flow
into the transformer and immediate tripping. To avoid this, a network phasing relay (device 60 in Figure 1) is
used with the master relay to permit reclosing only when the voltage conditions will cause power flow into
the secondary network.

IEEE
Annex B
(informative)
Example: 12.5 kV/480Y/277 V spot vault network

transformer protection
Engineers designing spot networks may choose to select from the following or other protection options
depending upon characteristics of their utility’s system and other aspects of their particular application. The
example vault shown in Figure B.1 is in a subbasement of a high-rise office building. For example purposes,
a number of protection devices are used. The vault contains 4–2500 kVA transformers connected delta on the
12.5 kV primary side and wye-grounded on the 480Y/277 V secondary, with 7% impedance. This vault is
fed from a major substation that supplies a portion of a central business district. Each spot-network
transformer is radially supplied from an underground 12.5 kV cable. Maximum three-phase fault current that
must be interrupted by any one 12.5 kV device in the vault is 8500 A (point 2 on Figure B.2). Bolted three-
phase 480 V bus fault current is 130 000 A. The transformers are networked on the low side through 3000 A
network protectors and Y-50 externally mounted fusible links. Conventional watt-type master relays and
phasing relays are used. On the high side of the transformer there is a three-position switch whose insulating
liquid-filled compartment is bolted to the transformer tank. This switch is used to interrupt transformer
magnetizing current, when the switch is opened to de-energize the transformer, to energize the transformer,
and to ground the 12.5 kV cable. For simplicity, Figure B.1 shows only two of the four transformers; all
transformers are protected identically. Figures B.2 and B.3 present the coordination of the current sensing
protective devices for transformer 1.
Protection of this vault starts with the addition of a three-phase vacuum fault interrupter. This device weighs
about 115 kg and is approximately 50 cm × 60 cm × 120 cm. The vacuum bottles are encapsulated in epoxy,
which allows for this rather compact design. The cable and transformer connections to the interrupter are to
be made using conventional 600 A elbow-type connectors. The interrupter has a 12 000 A (symmetrical)
interrupting capability and a minimum total clearing time of about 2 cycles. Line currents are sensed by
encapsulated bushing CTs. Control operation power is obtained from the CTs, so no additional power source
is required for tripping or for overcurrent relaying. The interrupter opens and closes by stored-spring energy.
Spring charging is done automatically via 120 V ac motor or manually after an opening operation. Standard
phase and ground overcurrent devices are used with inverse time and definite time characteristics,
respectively, and are designated as devices 51-T1 and 51N-T1 for transformer 1 in Figure B.1. These devices
trip transformer 1 vacuum interrupter directly, and a 52b contact of the interrupter then initiates a trip of the
protector. This provides backup clearing in the event of a control power failure or lockout relay failure.
In series with the three-phase interrupter are interrupting devices known as electronic fuses. These single-
phase devices interrupt fault currents above 2001 A in about 0.25 cycle. This feature limits energy and
thereby minimizes the possibility of tank rupture. In the nonlimiting region, above 400 A but less than 2001
A, it has a time-current tripping characteristic that coordinates with the feeder time inverse overcurrent
device; see Figure B.2. The _ cycle operating time of this device is too fast for the overcurrent elements to
detect and trip the three-phase interrupter, and therefore, it is possible that one or more fuses may be left
intact following the interruption of the major fault current for a transformer primary winding fault. Since this
condition may cause smoke and arcing to continue, an open phase detector, device 51/48, has been added.

IEEE
Figure B.1—Example of network transformer vault protection scheme

Referring to transformer 1 in Figure B.1, a sudden-pressure relay, device 63T1, is mounted on one of two
transformer access ports, which trips via the transformer lockout relay, device 86T1. A 450 °F bimetallic
heat probe, device 26T1, is inserted through the protector casing to detect protector faults. Heat probes are
also placed in close proximity to the 480 V utility bus for bus fault detection. Operation of a transformer
protector heat probe trips lockout relay device 86T1 and the bus lockout relay, device 86B. Operation of a
bus heat probe trips the bus lockout relay. Control dc is supplied by the parallel summation of three
individually charged capacitors, which in turn are supplied with half-wave rectified voltage from the 480Y/
277 V ac bus. One power supply is provided for each lockout relay. A loss of any two phases does not
constitute a loss of dc supply. An open or short on all three phases allows one lockout relay operation. Silver-
in-sand current-limiting fuses are placed in series with the 480 V supply cables to the consumer switchgear.
Each supply from the vault consists of 4–500 kcmil cables in parallel per phase. The low-voltage 500 kcmil
cable limiter fuse curve shown in Figure B.3 is assumed to be the sum of four identical cable limiter fuse
curves.

IEEE
Figure B.2—High-voltage coordination
For multiphase faults within the transformer high-side windings, the electronic fuse(s) will operate in about
0.25 cycle if the phase current is greater than 2001 A. If the phase current for a multiphase fault is less than
2001 A, then the phase overcurrent protection, device 51T1, will operate if properly coordinated before the
feeder relays, devices 51-L1. For phase-to-ground faults within the transformer windings, the electronic fuse
will operate for faults with the phase current greater than 2001 A and device 51N-T1 will operate before the
feeder relays, device 51N-L1. Note, the time-overcurrent portion of the electronic fuse provides backup for
devices 51-T1 and 51N-T1.

IEEE
Figure B.3—Low-voltage coordination
For example purposes, three overcurrent ground relays are used for transformer 1 (151N-T1, 151G-T1, and
64GF-T1 in Figure B.1) for low-voltage ground-fault protection. The 151N-T1 device must be set above the
maximum phase-to-neutral unbalanced load current. The 151G-T1 overcurrent device does not sense the
entire unbalanced load current, and therefore, can be set more sensitively than the 151N-T1 device. Test
measurements should be made to determine this division of load current. For any ground faults not involving
a neutral conductor or bus, the 151G device should see the same proportion of total ground-fault current
returning to its respective transformer as unbalanced load current. (See 6.1.3.) Note the suffix “G” is
preferred over “N” to denote that the purpose of the circuit is to detect fault current in a ground path. The
suffix “N” is preferred in the secondary neutral of current transformers or in the secondary of a CT whose
primary winding is located in the neutral of a power transformer or machine (ANSI C37.16-1997). See
Figure B.3 for the time-current coordination of these devices, including the consumer’s downstream ground

IEEE
sensor relay, device 151G-C1. To apply these relays securely, the transformer neutral bushing should be
isolated electrically from the transformer tank. The 64GF-T1 device is a case ground or point ground
protective device. If the transformer and protector can be insulated from ground, then this very sensitive
instantaneous acting relay can be used to detect only transformer and protector faults involving the case.
Coordination is not necessary with downstream devices, since the 64GF-T1 will not operate. Point 1 on
Figure B.3 represents the maximum low-voltage three-phase bus fault current contributed by one
transformer expressed in amps at 480 V. Point 2 represents a likely value for the maximum transformer fault
current in the presence of a 480 V arc. Each transformer is assumed to supply one-fourth of this total 480 V
bus fault current. See Dunki-Jacobs [B3] for an explanation of the magnitude of arcing current.
For faults located within a transformer low-voltage winding, between the low-voltage winding and its
protector, or in a protector, four units of current will flow—one from each of the unfaulted transformers as
well as one from the faulted transformer. The Y-50 fuse link for the faulted transformer will detect three of
these units of current. However, if the fault is high-resistance arcing and low in current magnitude, then even
three units of current will not be enough to properly melt the Y-50 links. For example, if this type of fault
was in the network protector, then heat detectors, device 26T1, are needed. The higher the current, the more
efficient the operation of the Y-link. The maximum current for a bolted fault would be three times 32 500 A,
or 97 500 A for this example. The Y-50 link would approach 0.2 s in clearing time.
Table B.1—NOTES for Figure B.1– Figure B.3
Device No.a Function (location)

(includes suffixes)
51-L1, L4 Time-overcurrent (line)
51N-L1, L4 Neutral time-overcurrent (line)
51-T1, T4 Time-overcurrent (transformer)
51N-T1, T4 Neutral time-overcurrent (transformer)
51/46-T1, T4 Open phase following high current (transformer)
86-T1, T4 Lockout relay (transformer)
63-T1, T4 Sudden gas pressure (transformer)
64GF-T1, T4 Case ground fault
151G-T1, T4 Low-voltage ground time-overcurrent (transformer)
151N-T1, T4 Low-voltage neutral time-overcurrent (transformer)
26-T1, T4 Heat detector (transformer)
32-T1, T4 Master (reverse power) (transformer network protector)
60-T1, T4 Phasing (voltage balance) (transformer network protector)
151G-C1, C2 Ground time-overcurrent (consumer)
E-F Electronic fuse (transformer)
Y-link Low-voltage fusible link (transformer)
CLF Low-voltage current-limiting fuse (consumer)
26-B Heat detector (bus)
86-B Lockout relay (bus)
aSuffix letters denote the location of the device in the circuit. See IEEE Std C37.2-1996 for a complete listing of
device function numbers.

IEEE
Annex C
(normative)
Other concepts being investigated
C.1 Low-voltage arcing fault detection
A utility and a manufacturer of network protectors are jointly working on a project to determine the
applicability of undervoltage and lower order harmonic voltage sensing of 480Y/277 V arcing network vault
faults. Field test data during normal vault operation and laboratory staged fault tests are being analyzed to
determine voltage relay applicability. It appears that an undervoltage device using conventional components
can be set to distinguish a 480 V fault from normal system operation. Lower order harmonics appear to be
the most practical component of bus voltage to detect during an arcing fault, other than the fundamental
frequency component. Both devices should be easy to install and maintain. Both devices are nonselective
and would, therefore, require time delay to coordinate with consumer protection, network protectors, and
fuse links.
C.2 Microprocessor-based current differential relays
Future current differential relays may be developed that overcome the lack of local high-side current
transformers by utilizing fiber-optic channels. It would be possible to use magnetic optic current transducers
located at the source stations in conjunction with the local low-side current transformers if such a relay
existed that allowed optical or traditional wiring for the current transformer connections. The use of the
fiber-optic channel would also allow transfer trip signals to be sent to the source station circuit breakers.
C.3 Current-limiting fuses and three-phase load-break switch
For those network installations where current limitation is important (usually because of very high available
fault currents), another device is being developed that combines current-limiting fuses and a gang-operated,
three-phase load-break switch.
For all fault currents above the interrupting rating of the load-break switch, the current-limiting fuses will
operate to clear the one-, two-, or three-phase fault, resulting in a minimal level of energy into the fault arc.
This is especially important where the network transformers use flammable conventional transformer oil.
Depending on the actual fuses used, I2t can be limited to values as low as 500 000 A2 s or less.
To eliminate any possibility of ferroresonance if only one or two fuses operate, the fuses are equipped with a
striker mechanism that will operate to trip the three-phase load-break switch whenever any fuse operates.
This mode of operation can eliminate the unnecessary blowing of fuses on unfaulted phases and provide the
benefit of three-phase disconnection of the transformer for any fault. The switch can be equipped with phase
and ground overcurrent electronic controls for detection of low-magnitude faults and can be tripped by
remote signals, such as from sudden-pressure relays or heat probes located elsewhere in the network vault.
The fuse and switch will be enclosed in SF6 gas for insulation. Replacement of the fuse(s) will most likely
not be an on-site procedure.

IEEE
Annex D
(informative)
Bibliography
[B1] Anderson, M. W., “A Utility Applies Dual Protection to Spot Networks,” IEEE Summer 1969 Power
Meeting, Dallas, TX, Conference Paper No. 69CP669-PWR, June 1969.
[B2] Cranos, J. C., and Gilligan, S. R., “Spot Networks and Connected Building Systems,” IEEE Industrial
and General Applications, vol. IGA-6, pp. 598–606, Nov./Dec.1970.
[B3] Dunki-Jacobs, J. R., “The Effects of Arcing Ground Faults on Low-Voltage System Design,” IEEE
Transactions on Industry Applications, vol. IA-8, no. 3, pp. 223–230, May/June 1972.
[B4] Erickson, I. B., “Ground Fault Protection That Works,” Electrical Systems Design, pp. 28–37, Nov./
Dec. 1986.
[B5] Fisher, L. E., “Resistance of Low-Voltage AC Arc,” IEEE Transactions on Industry and General
Applications, vol. IGA-6, pp. 607–616, Nov./Dec. 1970.
[B6] Glenn, D. J., “A New Fault-Interrupting Device for Improved Medium-Voltage System and Equipment
Protection,” IEEE/IAS Pulp and Paper Conference, Toronto, Ontario, Canada, Paper 440-T52, June 1984.
[B7] Griffin, T. R., Byrd, G. L., and Whitter, D. A., “Fault Tests on 480Y/277 V Spot Networks,” IEEE/PES
Conference, New Orleans, Paper 66-485, July 1966.
[B8] Heller, F., and Matthysse, I., “Limiters, Their Design Characteristics and Applications,” AIEE
Transactions, vol. 74, pp. 913–915, Oct. 1955.
[B9] IEEE 100, The Authoritative Dictionary of IEEE Standards Terms, Seventh Edition.8
[B10] Johnston, W. F., “How Do You Protect Your Spot Networks,” Electric Light and Power, pp. 71–73, Jan.
1970.
[B11] Kaufmann, R. H., and Page, J. C., “Arcing Fault Protection for Low-Voltage Power Distribution
Systems—Nature of the Problem,” AIEE Transactions on Power Systems Applications, vol. PAS-79,
pp. 160–167, June 1960.
[B12] Kischefsky, J. A., “Design and Operating Experience of Protectors for Network Transformers,” EEI
Working Group on 277/480 Volt Secondary Fault Protection Systems Meeting, Duluth, MN, Oct. 1986.
[B13] Landman, R. J., and Louie, B., “Fiberoptic SCADA Safeguards Underground Distribution Network,”
IEEE Computer Applications in Power, vol. 5, no. 2, pp. 39–44, April 1992.
[B14] Leinback, E. L., and Brookes, A. S., “Coordination of Secondary Network Protection,” AIEE
Transactions, vol. 74, pp. 924–930, Oct. 1955.
[B15] Peach, N., “Protect Low-Voltage Systems from Arcing-Fault Damage,” Power, vol. 108, pp. 61–65,
Apr. 1964.
8The IEEE products referred to in Annex D are trademarks of the Institute of Electrical and Electronics Engineers, Inc.

IEEE
[B16] “Polychlorinated Biphenyls in Electrical Transformers,” US Code of Federal Regulations, US EPA

Final Rule, Federal Register, 40 CFR Part 761.
[B17] Roop, D. W., and Vidonic, N. G., “Arcing Fault Protection on VEPCO’s 480Y/277 V Secondary Spot
Networks,” IEEE Transactions on Power Apparatus and Systems, vol. PAS-102, no. 2, pp. 364–372, Feb.
1983.
[B18] Schwab, R. L., and Stohr, E. W., “Secondary Network Equipment for 250–600 Volt Systems,” AIEE
Transactions on Power Applications and Systems, vol. PAS-73, pp. 1531–1536, Dec. 1954.
[B19] Smith, D. R., “Another Look at 480 Volt Spot Network Protection,” Electric Light and Power, T/D
Edition, pp. 60–63, Mar. 1971.
[B20] Smith, D. R., Bishop, M. T., and Rackliffe, G. B., “Research and New Protection Concepts for 480
Volt Spot Networks,” EEI Working Group on 277/480 Volt Secondary Fault Protection Systems Meeting,
Duluth, MN, Oct. 1986.
[B21] Smith, D. R., and Matty, P. W., “Spot Networks Can Improve Service Reliability to Suburban Load
Centers,” Westinghouse Engineer, May 1969.
[B22] Sonnemann, W. K., et al., “Compensator Distance Relaying—Part I, II, and III,” AIEE Transactions
on Power Apparatus and Systems, vol. PAS-77, pp. 372–402, 1958.
[B23] Wagner, C. F., and Fountain, L. L., “Arcing Fault Currents in Low-Voltage AC Circuits,” AIEE
Transactions, vol. 67, pp. 166–174, 1948.
[B24] Xenis, C. P., “The Limiter—Its Basic Functions in Network Distribution Systems,” AIEE
Transactions, vol. PAS-74, pp. 922–924, Oct. 1955.

IEEE Guide For The Protection of Network Transformers: IEEE Power Engineering Society

Uploaded by

Copyright:

Available Formats

IEEE Guide For The Protection of Network Transformers: IEEE Power Engineering Society

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 Guide For The Protection of Network Transformers: IEEE Power Engineering Society

Uploaded by

Copyright:

Available Formats

IEEE Std C37.

IEEE Guide for the Protection

IEEE Power Engineering Society

IEEE Guide for the Protection

Approved 21 March 2002

The Institute of Electrical and Electronics Engineers, Inc.

Copyright © 2002 by the Institute of Electrical and Electronics Engineers, Inc.

Print: ISBN 0-7381-3257-8 SH94988

Secretary, IEEE-SA Standards Board

Copyright © 2002 IEEE. All rights reserved. iii

iv Copyright © 2002 IEEE. All rights reserved.

Also included is the following nonvoting IEEE-SA Standards Board liaison:

Alan Cookson, NIST Representative

Copyright © 2002 IEEE. All rights reserved. v

4. General background ............................................................................................................................. 4

4.1 Network development.................................................................................................................. 4

5. Operation of equipment in a network transformer vault under fault conditions.................................. 6

5.1 Primary feeder faults.................................................................................................................... 6

6. Methods available to provide improved network transformer protection ......................................... 10

6.1 Electrical detection of faults ...................................................................................................... 10

7. High-voltage, fault-interrupting devices............................................................................................ 15

7.1 Circuit breakers.......................................................................................................................... 15

8. Low-voltage, fault-interrupting devices............................................................................................. 18

8.1 Network protector ...................................................................................................................... 18

9. Distributed source generation ............................................................................................................ 18

10. Network distribution SCADA............................................................................................................ 20

Annex A (informative) Response of network relays to system faults ........................................................... 21

Annex C (normative) Other concepts being investigated .............................................................................. 29

Annex D (informative) Bibliography............................................................................................................. 30

vi Copyright © 2002 IEEE. All rights reserved.

Copyright © 2002 IEEE. All rights reserved. 1

Accredited Standards Committee C2-2002, National Electrical Safety Code® (NESC®).3

ANSI C37.51-1989 (R1995), American National Standard for Switchgear—Metal-Enclosed Low-Voltage

IEEE Std 80™-2000, IEEE Guide for Safety in AC Substation Grounding.5, 6

2 Copyright © 2002 IEEE. All rights reserved.

IEEE Std C57.109™-1993, IEEE Guide for Transformers Through-Fault-Current Duration.

NFPA 70-1999, National Electrical Code® (NEC®).7

3.1 network feeder: A feeder that supplies energy to a network.

Copyright © 2002 IEEE. All rights reserved. 3

4.1 Network development

4 Copyright © 2002 IEEE. All rights reserved.

Figure 1—Typical network system

Copyright © 2002 IEEE. All rights reserved. 5

4.2 Application considerations of adding protection to spot-network vaults versus

5. Operation of equipment in a network transformer vault under fault

a) The high-voltage primary supply circuits,

5.1 Primary feeder faults

6 Copyright © 2002 IEEE. All rights reserved.

5.2 Network equipment faults

5.2.1 Primary-voltage switch faults

Copyright © 2002 IEEE. All rights reserved. 7

5.2.2 Transformer faults

5.2.3 Protector faults

5.2.4 Low-voltage bus faults

8 Copyright © 2002 IEEE. All rights reserved.

5.2.5 Low-voltage faults in customer supply cables and bus

5.2.6 Other faults

Copyright © 2002 IEEE. All rights reserved. 9

5.2.7 Primary-voltage cable faults