Kalam, Akhtar - Kothari, Dwarkadas Pralhaddas - Power System Protection and Communications-New Age Science LTD (2010)

Power System
Protection and
Communications
AKHTAR KALAM
D P KOTHARI
Preface
Electrical protection is as fundamental to the generation, transmission and

distribution of electricity as generators, transformers and transmission lines
themselves. It is an interesting and complex area of power engineering
and many engineers have devoted their life’s work to the subject. There is
always a need to provide education in the theory and practice of protection
engineering for engineers and technical personnel, because of its
importance in design and operation of the power system.
Victoria University (VU) has conducted courses in protection for
many years and over the past five years, has presented a three day course
as part of the Electricity Supply Association of Australia (ESAA)/Australian
Power Institute (API) Continuing Professional Development Program. Over
700 engineers and senior technical staff have participated in these courses
and have indicated that they found the experience of great value. The text
is designed to give participants the basic skills to commence practice in
the basic protection system principles.
About two decades ago microprocessor based relays replaced the
traditional electromechanical ones and it found increasing usage in the
various types of protection relays and other Intelligent Electronic Devices
(IEDs). This was due to the fact that there were many components in an
IED that are similar and in some cases identical between substation
protection, control, monitoring and recording devices. Microprocessor
technology was extensively used by manufacturers of devices required by
the different domains in the substation. Protection relay manufacturers
started implementing protection functions by emulating the principles of
electromechanical or solid state relays using numerical methods.
Both vendors and users realized that microprocessor based
technology allows the development and application of functions that were
impossible in the world of electromechanical devices. This process occurred
simultaneously from several directions. Protection relay manufacturers
vi PREFACE
started adding other features like fault locators, measurements and

recording capabilities. Advancements in communications and acceptance
of standard international or industry protocols resulted in the widespread
of substation automation systems and further expanded the functionality
of substation IEDs.
The second part of the book highlights the new international
standard for substation communications (IEC 61850) which will completely
change the ways protection, control, monitoring and recording has been
traditionally done in the substation. This standard for communication
networks and systems in substations allows the development of high-speed
peer-to-peer communications based distributed protection applications that
result in significant changes in the ways protection functions are
implemented. This replacement of functions implemented in a single device
with equivalents using exchange of analogue and status information over
the substation local area network has revolutionized the conventional
power system protection.
We are indebted to the industry experts who have given valuable
time to set up the new IEC standard 61850. This book will try and give not
only the basic understanding of the basic protection principles but also the
numerous protocols as developed by the communication engineers and
which is now going to be extensively used by the 21st century power
industry. This book describes in detail the principles of different IEC 61850
distributed functions and analyses the factors that will affect their
performance. The definitions of the individual components of distributed
functions are presented in detail, including the different possible allocations
of sub-functions and functional elements in physical devices.
We are thankful to our colleagues at Victoria University, Melbourne
and VIT University, Vellore, who have helped us directly or indirectly in
completing the book. First author is grateful to Dr. Rushan Lloyd
Muttucumaru, Postdoctoral Fellow and Dr. Cagil Ozansoy and
Dr. Amanullah Maung Than Oo for their help in preparing the latter part
of the work. The second author is thankful to Dr. G. Viswanathan,
Chancellor, VIT University, Vellore for his constant encouragement for
completing this project. We thank our families for their patience and
encouragement shown while we were working on this project and Mr.
Prabhakar Kartikeyan for his help in preparing the manuscript.
Akhtar Kalam
D.P. Kothari
List of Acronyms
RTU — Remote Terminal Units

PLC — Programmable Logic Controllers
EMS — Energy Management System
IED — Intelligent Electronic Devices
I&C — Instrumentation and Control
SCADA — Supervisory Control And Data Acquisition
ISO — International Standards Organization
OSI — Open System Interconnection
SER — Sequential Events Recorder
MTBF — Mean Time Between Failures
HMI — Human Machine Interface
LAN — Local Area Network
WAN — Wide Area Network
UCA — Utility Communication Architecture
RS485 — The Electronics Industry Association, Recommended
Standard 485
RS232 — The Electronics Industry Association, Recommended
Standard 232
ACSI — Abstract Communication Service Interface
CORBA — Common Object Request Broker Architecture
IEC 61850 — Is a framework for substation automation that addresses
more of what is required for interoperability of Intelligent
Electronic Devices.
SA — Substation Automation
MMS — Machine Monitoring Systems
DCOM — Distributed Component Object Model
SCSM — Specific Communication Service Mapping
viii LIST OF ACRONYMS
GOOSE — Generic Object Oriented Substation Event

EPRI — Electric Power Research Institute
MODBUS — A popular protocol with industrial users, popular in
substations
Profibus — Protocol
TCP/IP — Transmission Control Protocol/Internet Protocol
VTs — Voltage Transformers
CTs — Current Transformers
RTDs — Resistance Thermal Detectors
VAR — Voltage Ampere Reactive Controllers
Controllers
CIGRE — International Council on Large Electric Systems
IEEE — Institute of Electrical and Electronic Engineers
DMS — Distribution Management System
DA — Distribution Automation
CVTs — Capacitor Voltage Transformers
PQ — Priority Queuing
CQ — Custom Queuing
WFQ — Weighted Fair Queuing
WRED — Weighted Random Early Detection
CAR — Committed Access Rate
RSVP — Resource Reservation Protocol
QoS — Quality of Service
Tspeck — Traffic Specification
RESV — Reservation Request
OMA — Object Management Architecture
IDL — Interface Definition Language
CASM — Common Applications Service Models
MMS — Manufacturing Message Specification
IEC TC — International Electrotechnical Committee Technical
Committee
IEC — International Engineering Consortium
NIS — Network Integrated System
IEPS — Information Embedded Power System
SNMP — Simple Network Management Protocol
NTP — Network Time Protocol
FACTS — Flexible AC Transmission System
VBR — Various Bit Rates
MPCS — Massively Parallel Computing Systems OR (Mathematics,
Physics and Computer Science)
LIST OF ACRONYMS ix
XML — Extensible Markup Language

JSP — Java Server Pages
DTD — Document Type Definition
DOM — Document Object Model
CSV — Comma Separated Variable
OASIS — Open Access Sametime Information Systems
XSL — Extensible Stylesheet Language
OXC — Optical Cross-Connects
OADM — Optical Add-Drop Multiplexer
FDDI — Fiber Distributed Data Interface
UDP — User Datagram Protocol
IP — Internet Protocol
NIS — Network Integrated System
SQL — Structured Query Language
DAType — Data Attribute Type
ToS — Type of Service
IEE — Institute of Electrical Engineers
VOIP — Voice Over IP
ATM — Asynchronous Transfer Mode
TDM — Time-Division Multiplexing
This page
intentionally left
blank
Contents
Preface ............................................................................................................. v
List of Acronyms .......................................................................................... vii
Recommended Protection Terminology ........................................................ xv
Part A
POWER SYTSEM PROTECTION
Chapter 1 Basic Principles 3–21
1.1 Introduction to Protective Relaying 3
1.2 Power System Plant and Layout 5
1.3 Switching Arrangements 8
1.4 The Function of Protective Relaying 13
1.5 Principles of Protective Relaying 14
1.6 Unit and Non-unit Schemes 17
1.7 Zones of Protection 19
1.8 Common Terminologies 20
Chapter 2 Network Analysis and Fault Calculations 22–50

2.1 Faults on Power System 22
2.2 Faults Type 23
2.3 Fault Level Calculations 24
2.4 Limiting Short-circuit Levels 26
2.5 Transients During a Balanced Fault 27
2.6 Sequence Networks for Calculation of Unbalanced Faults 30
2.7 Calculation of Voltages in the Network 38
2.8 Short-circuit Fault Calculations 39
xii CONTENTS
Chapter 3 Earth Fault and Interferences 51–79

3.1 Introduction 51
3.2 Power System Arrangements and Construction Features 52
3.3 Earth Potential Rise 57
3.4 Safety Considerations 63
3.5 Application of Safety Criteria 64
3.6 Demands on Protection Arising from Safety, Reliability
and Interference Considerations 65
3.7 Interference on Supplied and other Systems 75
Chapter 4 Relaying Transducers 80–100

4.1 Introduction 80
4.2 Voltage Transformers 80
4.3 Current Transformers (CTs) 85
4.4 Guidance in Application of CTs 92
Chapter 5 Overcurrent Protection 101–125

5.1 Introduction 101
5.2 Types and Construction 102
5.3 System Analysis 106
5.4 Settings of IDMT Relays 107
5.5 Relay Discrimination 109
5.6 Grading Margin 121
5.7 Earth Fault Protection 122
Chapter 6 Fuses 126–166

6.2 Categories of Fuses 128
6.3 Fuse Operating Oscillograms 145
6.4 Time vs Current Characteristic 148
6.5 Discrimination 155
6.6 Testing of Fuses 161
6.7 Future Developments 166
Chapter 7 Distance/Impedance Protection 167–208

7.1 Overview of Distance Protection 167
7.2 Principle of Distance Protection 167
7.3 Analog Amplitude and Phase Comparison 173
CONTENTS xiii
7.4 Relay Types and Their Application 176

7.5 Derivation of Signals for Distance Protection 183
7.6 Methods of Realising Comparators 185
7.7 Signal S2 193
7.8 Pre-filters Used in Power System Protective Relaying 194
7.9 Effect of the Ratio Source Impedance to Line Impedance
(ZS/ZL) 196
7.10 Time Grading of Distance Relays 197
7.11 Requirements of Definite-distance Schemes 199
7.12 Reach of Distance Relay 200
7.13 Digital Computation by a Microprocessor 202
Chapter 8 Differential Protection 209–218

8.2 Circulating Current Differential (Low Impedance) 209
8.3 Biased Differential Protection 212
8.4 High Impedance Current Differential 213
8.5 Pilot Wire Protection 215
8.6 Phase Comparison Protection 217
Chapter 9 Unit, Remote and Back Up Protection 219–239

9.1 Unit and Non-unit Protection Schemes 219
9.2 Remote and Local Back Up Protection 224
9.3 Example Demonstrating a Method to Determine Current
Settings of Circuit Breaker Failure Schemes 237
Part B
POWER SYTSEM COMMUNICATIONS
Chapter 1 Introduction 243–245

1.1 International Electrotechnical Committee (IEC)
Technical Committee (TC) 57 243
1.2 Electric Power Research Institute (EPRI) 243
1.3 Innovative Integration Developments (IEDs) 244
Chapter 2 Communication Principle 246–259

2.1 Terminologies 246
2.2 Architectures 251
xiv CONTENTS
Chapter 3 Protocols 260–278

3.1 Introduction to Power System Communication 260
3.2 Protocols in General 264
3.3 Expand on DNP 265
3.4 Standardisation Developments 267
Chapter 4 Middleware 279–315
4.2 Automation Systems and Communication Needs 280
4.3 Middleware Requirements for Protection Applications 285
4.4 Middleware Architectures 288
4.5 Publish/Subscribe Middleware 294
4.6 Corba and its Features 302
4.7 Common Architectures for Communication Devices 310
Chapter 5 Information Embedded Power Systems 316–322
5.1 Overview of Communication Networking Requirements 316
5.2 Information Embedded Power System via LAN/WAN 317
5.3 The Benefits of using IEPS-LAN/WAN Technology 321
Chapter 6 Fiber Optic Network Infrastructure 323–334
as Next Generation Power System
Communications
6.1 Background 323
6.2 Current Power System Data Communication Media 324
6.3 Networks and Information Technology 326
6.4 Wide Area Communication Infrastructure 329
6.5 Local Area Substation Network Design 331
6.6 Time Data Communication and Exchange 332
Chapter 7 Conclusions 335–342
• Conclusion to Chapter Three on Protocols 335
• Overall Conclusion to IEC61850 and DNP-3 336
• Conclusion to Chapter Four on Middleware 338
• Conclusion to Chapter Five on Information Embedded Power
System 340
• Conclusion to Chapter Six on Fiber Optic Network
Infrastructure as Next Generation Power System
Communications 341
References 343–354
Index 355–357
Recommended Protection
Terminology
ESAA COMMITTEE NO. 2.14 – PROTECTION
1. INTRODUCTION
The purpose of this study was to compile a list of recommended terms to
enable authorities of ESAA to discuss protective schemes, with mutual
understanding of the terms used to describe these schemes and their
performance parameters.
No attempt has been made to categories the various terms, which
are set out in alphabetical order alone. Where possible, standard items of
terminology from either IEC or BS have been adopted in toto and the
appropriate reference appears in parentheses below the term.
Terminology of a purely telecommunications nature is not included
herein; reference for such terminology should be made to ESAA Publication
D (b) 7-1968: Communication Terms for Power System Telecommunications.
Terms associated with auto-reclosing applications, as contained in
ESAA Publication D(b) 12-1971: Guide to the Application of Auto Reclosing
to Radial Overhead Lines Supplying Urban and Rural Areas, are not
repeated herein.
The terms as listed below are commonly used by the larger supply
authorities in Australia.
xvi POWER SYSTEM PROTECTION AND COMMUNICATIONS
2. REFERENCES
The basic references used in this study were as follows:
BS 142-1966 Electrical Protective Relays
BS 3950-1965 Electrical Protective Systems for AC Plant
(withdrawn and not replaced October 1981)
IEC 50(16)-1956 Protective Relays
IEC 255.4-1976 Appx.E. Impulse Voltage Withstand and High
Frequency Disturbance Tests.
3. RECOMMENDED TERMINOLOGY
Term Definition Remarks

Acceleration A term applied to describe the e.g. reducing the operating
action of speeding up the time for a fault within Zone 2
operation of a distance relay at one end of a line by an ‘end-
which would otherwise be time to-end’ signal from the Zone 1
delayed. operation at the other end of
the line.
Algorithm A procedure for processing
information.
Arc Supression An earthing reactor so Preferred to the term ‘Petersen
Coil designed that its inductive Coil’. Similar to BS 204
reactive current to earth under definition.
fault conditions balances the
capacitive current to earth
flowing from the power line so
that the earth current at the
fault is limited to practically
zero.
Arcing Time The time between the instant of
arc initiation at the circuit-
breaker contacts and the final
extinction on all poles.
Backup Protection A protection which is intended See also Local Backup

to operate when a system Protection. Similar to IEC 50 —
fault is not cleared in due time Section 448-03.
because of failure or inability of
the main protection or the
associated circuit-breaker to
operate.
(Contd...)
RECOMMENDED PROTECTION TERMINOLOGY xvii
Balanced Fault A fault giving rise to steady

state symmetrical phase
currents in a 3-phase system.
Balanced Voltage A differential protective system See also Longitudinal

Protection where equal primary current Differential Protection.
flow into and out of the
protected zone ideally
produces no net circulating
current through the CT
secondary circuit and a series
connected relay.
Biased Relay An electrical measuring relay Similar to IEC 50(16).

the operating characteristic of
which is changed by the appli-
cation of restraint derived from
input energising quantity(ies).
Blind Spot A fault location between a cir- Preferred to the term ‘dead
cuit-breaker and its associated zone’. A fault in the blind spot,
transformers where post-type although within the zone of a
current transformers are used particular protective system, is
and are provided on one side not cleared completely by the
of the breaker only. operation of that protective
system alone.
Blocking A protection system associated IEC 50 — Section 448-03.
Protection with a signalling system in
which the receipt of a signal
blocks tripping locally
initiated.
Blinder A relay whose characteristic, Usually applied to long,

when plotted on a R-X heavily-loaded lines to
diagram, is a straight line prevent tripping on power
crossing the characteristic of swings by confining the
another relay and arranged to tripping characteristic of the
prevent tripping on one side of line protection relays.
its own characteristic.
Bolted Fault A balanced fault of virtually Typically a fault arising as a

zero impedance between each result of working earths
phase and the neutral point of applied but not removed prior
a 3-phase system, reducing to energising a line. Requires
voltage measurements for special measures to ensure
protection purposes to correct clearance (see switch-
negligible values. in fault protection).
(Contd...)
xviii POWER SYSTEM PROTECTION AND COMMUNICATIONS
Breaker Failure A specific form of local backup Frequently by means of a

Protection protection which operates in sequential tripping scheme
the event of a circuit-breaker working through the busbar
failing to clear a fault and trips protection trip circuits. Also
all other circuits feeding into provides protection for blind
the same section of busbar as spots.
that breaker.
Buchholz Relay A protective relay, responsive Preferred to the term ‘gas
either to the collection of gas relay’.
produced by incipient faults or
to oil surges caused by
explosive faults within a
transformer tank, arranged to
operate an alarm or to trip the
transformer out of circuit.
Characteristic The phase angle at which the Similar to BS 142 definition.
Angle performance of a relay is
declared. It is usually the angle
between the energising
quantities at which maximum
sensitivity occurs.
Characteristic The maximum value of system Similar to BS 3950.
Impedance Ratio impedance ratio up to which
the relay performance remains
within the prescribed limits of
accuracy.
Characteristic An electrical quantity, or one of e.g. current for an over-current
Quantity its attributes, the name of relay; frequency for a frequency
which characterises the relay relay, etc.
and the value(s) of which are
the subject of accuracy
requirements.
Check Protective An auxiliary protective system Similar to BS 3950.
System intended to prevent tripping
due to inadvertent operation of
the main protective system.
Check Relay A relay forming part of a A current check relay in a
protection system which voltage balance Pilot Wire
confirms the presence of a fault Protection System which
condition, independently of allows tripping only in the
other detectors and allows a presence of fault current, e.g.
trip command to be issued. not as a result of Pilot Wire
short-circuit.
(Contd...)
RECOMMENDED PROTECTION TERMINOLOGY xix
Also, a relay which by its A current check relay in a local

contact state confirms a back-up protection system
particular primary circuit which together with main
condition and in conjunction protection status and a time
with other indicators, function, establishes the
determines trip circuit action. breaker failure criterion.
Check Zone Term applied to the non-

selective part of a multi-zone
bus protection system,
supervising current flow at the
terminals of the complete
station. Tripping is conditional
on operation of both the check
and a discriminative zone.
Circulating A differential protective system The classical Merz-Price

Current where the equal primary system. See also Differential
Protection current flow into and out of the Protective System.
protected zone ideally
produces a circulating current
through the CT secondary
circuit but no net current
through a shunt connected
relay.
Comparator A device which compares two In most distance relays, either

or more input signals and phase comparators or
produces an output when a amplitude comparators are
predetermined relationship used. Some modern solid state
between the inputs occurs. distance relays have multi-
input comparators.
Component Test A test to determine the BS 3950.

characteristics of an individual
component.
Conjunctive Test A test on a protective system, BS 3950.

including all relevant
components and auxiliary
equipment appropriately
interconnected.
Counter A logic device which cycles
through a defined set of states
in response to input transitions.
(Contd...)
xx POWER SYSTEM PROTECTION AND COMMUNICATIONS
Cross-country Simultaneous flashover to Common on unearthed

Fault earth of two different phases in systems or systems earthed
different line sections of the through arc suppression coils.
same power system generally,
although not necessarily, at the
same voltage level.
DC/DC Converter A power supply employing

solid state switching devices
and saturable electromagnetic
components with suitable
control circuitry to convert
direct current power from one
voltage level to another and
provide galvanic input/output
separation.
Dead Zone See ‘Blind Spot’. Not recommended
Dependent Time A measuring relay having an

Lag Relay operating time which is a
function of the value of the
characteristic quantity.
Differential A unit protective system in Similar to BS 3950 definition.

Protective System which an algebraic comparison
is made of currents at two or
more points in the power
system.
Digital Relay A relay which processes the

actuating quantity in a digital
format, the analogue to digital
conversion being normally
accomplished internally.
Direct Intertrip The signal initiated by local

protection which, when
received at the remote end of
the protection section, trips that
circuit-breaker without
reference to the state of the
remote protection.
(Contd...)
RECOMMENDED PROTECTION TERMINOLOGY xxi
Directional A unit protection system

Comparison associated with signalling, the
Protective System operation and selectivity of
which depends on the
comparison of the directions of
power flow at the ends of the
protected section.
Directional A non-unit protection, the

Overcurrent operation of which depends on
Protection the magnitude of the current
and its phase angle relative to
a voltage reference at the point
of measurement.
Discrimination The quality whereby a BS 3950.

protective system distinguishes
between those conditions for
which it is intended to operate
and those for which it shall not
operate.
Discriminative Term applied to the selective

Zone part of a multi-zone bus
protection system, supervising
current flow at the terminals of
a particular bus zone. Tripping
is conditional on operation of
both a Discriminative and the
Check Zone.
Distance A non-unit protection system, IEC 50 Section 448-03.

Protection the operation and selectivity of
which depend on local
measurement of parameters
from which the equivalent
distance to the fault is
evaluated and compared with
zone settings.
(Contd...)
xxii POWER SYSTEM PROTECTION AND COMMUNICATIONS
Distance A Distance Protection System

Protection with associated with signalling in
Permissive which the shortest zone setting
Overreach corresponds to a distance
longer than the length of the
protected section and in which
tripping to clear a fault in the
protected section without a
deliberate time delay is only
permitted when fault detection
at one end is coincident with
receipt of a tripping signal from
the opposite end.
Permissive which the shortest zone setting
Underreach corresponds to a distance
shorter than the length of the
protected section and in which
tripping of the end remote from
the fault without a deliberate
time delay is only permitted
when fault detection at that end
is coincident with receipt of a
tripping signal from the
opposite end.
Distance An Underreaching Distance
Protection with Protection System associated
Acceleration with signalling in which the
shortest zone setting at one end
is extended corresponding to a
distance longer than the length
of the protected section, by
transmission of a signal from
the protection at the opposite
end.
Blocking which reverse looking fault
Overreach detectors transmit a signal to
the opposite end of the
protected section to block
tripping by overreaching
forward looking elements of
the system.
(Contd...)
RECOMMENDED PROTECTION TERMINOLOGY xxiii

Blocking which the shortest zone setting
Underreach corresponds to a distance
shorter than the length of the
protected section, which
provides for clearance of faults
in the section beyond this zone
setting in less than Zone 2 time,
but which contains reverse
looking fault detectors to
initiate signalling to the
protection at the opposite end
of the section in order to extend
its tripping delay to Zone 2
time, so as to discriminate
against external faults.
Drawout Relay A relay whose construction is Usually applies to a complete
such that it can be inserted into relay chassis as distinct from
and withdrawn from a case or individual relay elements.
other such fitting and provide Provision is made in the
with sliding contact fingers for construction of the case to
establishing the electrical automatically short-circuit any
connections between the external current transformer
withdrawable element and the secondary circuit when the
fixed portion. relay is withdrawn.
Duplicate The provision of two The two protective systems
Protection completely separate protective may share common supplies.
systems at a particular relaying However, duplication in its
point, either of which is capable full sense implies that no
of fulfilling the required failure of one system to
protective function. operate correctly may prevent
the other system from
performing its intended
function.
Dynamic Test A test measuring the (relay) Typically includes the
response to both the transient transient DC component of the
and steady state components of fault current.
the primary circuit actuating
quantity.
Earth Fault Protection for faults between Similar to IEC 50(16). Use of
Protection conductors and earth. the term ‘earth leakage
protection’ is not
recommended.
(Contd...)
xxiv POWER SYSTEM PROTECTION AND COMMUNICATIONS
Electro-mechanical An electrical relay or relay

Relay element whose characteristic is
determined partly or wholly by
moving parts.
External Fault A fault outside the defined Preferred to the term ‘through
zone of protection. fault’.
Fast Transient A type test intended for ANSI 37-90-1978 and IEEE
Surge Withstand energised solid state relay Guide P472/D2-1982.
Capability Test systems, to determine whether
they will operate without
erroneous output, component
failure or calibration change
beyond normal tolerances
when specified high voltage
fast rising transients such as
occur during the interruption
of inductive devices in DC
circuits, are applied.
Fault Clearance The time interval between

Time inception of fault current and
arc extinction. The sum of the
relay operating time, circuit-
breaker opening time and arc
duration.
Fault Detector An element of a protection Alternative term for Starting

relay which responds to faults Element.
or abnormal service conditions,
not necessarily only within the
protected zone, and initiates
the operation of other elements
of the protection.
Fault Level The rms value of the fault

current in kA at rated system
voltage available at the point of
fault in a particular location of
the power system.
Fault Setting The limiting value of the May be referred either to

measured quantity at which primary or the secondary
operation of the protective circuit. BS 3950.
system occurs.
(Contd...)
RECOMMENDED PROTECTION TERMINOLOGY xxv
Fault Throwing A method in which operation IEC 50(16).

of the protection of a circuit not
provided with a circuit-breaker
at one end causes an intentional
fault on the circuit at that end,
so enabling the protection at
the other end or ends to operate
to disconnect the circuit.
Field Failure A protective system which

Protection operates when the field supply
to a synchronous machine
drops below the required level.
Final Tripping A tripping command issued by

a self-reset relay which does
not initiate reclosing.
Flag Indicator An indicating device which Preferred to the term ‘target’.

shows that the associated relay Includes indicating lights.
element has operated.
Flip-Flop A bistable logic device with

defined responses to various
input levels and/or transitions.
Frame Leakage Protection in which the Similar to IEC 50(16).

Protection actuating quantity is the
current flowing between the
metal framework enclosing the
protection zone and earth.
Frequency Relay A relay in which the May be of the under-or over-
characteristic quantity is frequency type.
frequency.
Gas Impulse Relay A relay which responds to a This device is an alternative to
sudden change of pressure of the Buchholz relay and may
the air cushion above oil- be used on transformers not
immersed equipment. fitted with a conservator.
Harmonic Bias A method of making Frequently used in differential
differential relays insensitive to protective systems.
magnetic inrush currents,
wherein the harmonics are
filtered out from the
differential current and used to
provide additional restraint.
(Contd...)
xxvi POWER SYSTEM PROTECTION AND COMMUNICATIONS
High Frequency A type test intended for AS 2481-1981.

Disturbance Test energised solid state relays to
determine whether they will
operate in a faulty manner
when specified high-frequency
transients representative of
practical system conditions are
applied.
High-Impedance A single-input relay having a Frequently used in differential

Relay high input impedance such protective systems.
that, when fed from a current
transformer, the relay current is
substantially different from
that obtained using the
transformation ratio.
High-speed A protective system whereby Preferred to the term

Protection no intentional time delay is ‘instantaneous protection’.
introduced between the onset
of a fault and the initiation of
tripping power.
Impedance Relay A specific type of distance

protection relay whose polar
characteristic, when plotted on
an R-X diagram, is a circle
having its centre at the origin.
Impulse Starter A starting element in a Often used to discriminate

protective relay system between faults and high
responsive to step changes balanced load conditions.
rather than the absolute value
of the actuating quantity.
Impulse A type test intended to AS 2481-81.

Withstand Test determine the ability of a relay
to withstand, without damage,
very high value and short
duration over-voltages.
Independent A measuring relay having an
Definite Time Lag operating time which, after a
Relay predetermined value is
reached, is independent of the
value of the characteristic
quantity.
(Contd...)
RECOMMENDED PROTECTION TERMINOLOGY xxvii
Inrush Factor The peak value of the first BS 3950.

major loop of the magnetising
inrush of the power
transformer winding being
switched, expressed as a
multiple of the rms value of the
rated current of that winding.
Internal Fault A fault within the defined zone

of protection.
Intertripping The tripping of circuit- Synonymous with Transfer

breaker(s) by signals initiated Tripping (British usage).
by protection at a remote
location.
Integrator A device whereby the output

signal is the integral of the
input signal.
Integrated Circuit A single, homogeneous device

containing a large number of
semi-conductor components,
e.g. transistors, grouped and
connected in such a way as to
form a complete circuit.
Knee Point Voltage That point on the open-circuit

excitation curve of a current
transformer where a 10%
increase in applied voltage
causes a 50% increase in
exciting current.
Level Detector A device whose output changes

state when the input signal has
attained a desired threshold
level.
Local Backup A form of backup protection in Can also involve the

Protection which determination and intertripping of remote circuit-
initiation of the required action breakers. Similar to IEC 50,
takes place at the same location Section 448-01.
as that at which the main
protection is situated.
(Contd...)
xxviii POWER SYSTEM PROTECTION AND COMMUNICATIONS
Lockout Tripping A tripping command issued by

a relay having a manual (local
or remote) reset facility. The
tripping command is
maintained in case of
attempted reclosure until the
relay is reset.
Logic Gate A device whereby the output

signal is related to the input
signals by a prescribed form of
mathematical logic.
Longitudinal A unit protection system, the IEC 50, Section 448-03.

Differential operation and selectivity of
Protection which depends on the
comparison of the phase and/
or magnitude of the currents at
the ends of the protected
section.
Main Protection A protection system expected Similar to IEC 50, Section

to have priority over backup 448-01.
systems in initiating fault
clearance.
Memory Relay A specific type of distance relay The term ‘Memory Trip’ is
whose polarising circuit is used when such a relay is
oscillatory in nature, so that a employed for switch-in fault
polarising signal persists in the protection, i.e. in distance
relay for a short time after the protection with bus VTs.
collapse of system voltage at
the instant of fault.
Mho Relays A specific type of distance relay

whose polar characteristic,
when plotted on an R-X
diagram, is a circle which either
passes through the origin
(polarised mho) or is offset
from the origin (offset mho).
(Contd...)
RECOMMENDED PROTECTION TERMINOLOGY xxix
Microprocessor An integrated circuit

incorporating the majority of
the principal components of a
computer processing unit, viz.
control unit, arithmetic and
logic unit, data registers and
possibly also clock generator,
memory and input/output
ports.
Modified A specific type of distance relay

Impedance Relay whose polar characteristic,
when plotted on an R-X
diagram, is a circle having its
centre located at a set point
along the R axis.
Negative Phase A non-unit protection system

Sequence in which the operation is
Protection dependent upon the level of
negative phase sequence
component in the characteristic
quantity.
Opening Time The time between the

application of tripping power
to the circuit-breaker when
closed and the instant of
separation of the contacts.
Operating Time The time which elapses from IEC 50(16).

the appearance of the abnormal
conditions which cause the
operation of the protection
until the protection initiates
tripping or alarm.
Operational An amplifier with very high

Amplifier gain and differential input.
Used to perform various
mathematical functions
depending on the feedback
arrangement used, e.g.
inverting, summation,
integration, etc.
(Contd...)
xxx POWER SYSTEM PROTECTION AND COMMUNICATIONS
Oscillatory Surge A type test intended for Term used by ANSI 37.90-1978
Withstand energised solid state relay and IEEE guide P472/D2-1982
Capability Test systems to determine whether to describe the High
they will operate without Frequency Disturbance Test of
erroneous output, component AS 2481-1981.
failure or calibration change
beyond normal tolerances,
when specified oscillatory
surge voltages representative
of practical system conditions,
are applied.
Out of Step Protection which separates the IEC 50(16).

Protection appropriate parts of a power
system in the event of
sustained power transfer
oscillations.
Overcurrent Protection which operates Use of ‘overload protection’ as
Protection when the current exceeds a a general term is not
predetermined value. recommended. IEC 50(16).
Overload Protection which operates Refers specifically to plant in

Protection when the protected zone is the protected zone. IEC 50(16).
overloaded.
Overreach The unintentional extension of

the protection operation zone
beyond that indicated by the
relay setting.
Overreaching A form of protection in which Usually associated with ‘end

Protection relays at one terminal are set to to end’ signalling.
operate for faults on and
beyond a specified line section.
Overvoltage Protection that operates at a

Protection predetermined voltage level
which is greater than normal.
Permissive The trip signal initiated by local

Intertrip protection which, when
received at the remote end of
the protected section, permits
the remote protection to trip the
associated circuit-breaker.
(Contd...)
RECOMMENDED PROTECTION TERMINOLOGY xxxi
Permissive A protection system associated IEC 50, Section 448-03.

Protection with a signalling system in
which the receipt of a signal
permits tripping locally
initiated.
Phase Comparator Component of a protection

relay designed to respond to
the phase angle between
energising quantities.
Phase Comparison A unit protection system, the

Protection operation and selectivity of
which depend on
measurement, at each end of
the protected zone, of the phase
angle between local and remote
end currents.
Phase Sequence Circuitry designed to separate

Network positive, negative or zero
sequence current or voltage
components or a combination
of these from a 3-phase supply
system to protective relays.
Pilot Isolating A transformer for isolating

Transformer pilot wires from the terminal
relay equipment to provide
protection against the effects of
any longitudinal voltages
induced in the pilot wires.
Pilot Supervision A method for monitoring the Frequently involves the

integrity of pilot-wire circuits. injection of a small DC
circulating current, any
significant change in which
initiates an alarm.
Pilot-wire A differential protective system

Protection in which a metallic circuit is
used to inter connect the relays
at the terminals of the protected
zone.
(Contd...)
xxxii POWER SYSTEM PROTECTION AND COMMUNICATIONS
Plug-in Relay A relay designed to the inserted IEC 50(16).

in a fitting by means of plugs
serving as terminals.
Polar The graphic representation of a See R-X diagram.

Characteristic protective relay’s operating
range in terms of a polar co-
ordinate system.
Polarisation Method of making the Similar to IEC 50(16).
operation of a relay dependent
on the phase angle between
fault current and voltage.
Polygonal A distance relay characteristic

Characteristic which, when plotted on an R-X
diagram, is a closed figure
having three or more (usually
four) straight sides.
Protected Zone That portion of a power system

(BS 3950) protected by a given protective
system or part of that
protective system.
Protective The apparatus, including Similar to BS 3950 definition.

Equipment protective relays, transformers
and auxiliary equipment, for
use in a protective system.
Protective Scheme The coordinated arrangements A protective scheme may

for the protection of one or involve several protective
more elements in a power systems. BS 2950.
system.
Protective System A combination of protective Similar to BS 3950 definition.

equipments designed to secure,
under predetermined
abnormal conditions, the
disconnection of an element of
a power system.
Reach For stepped curve distance- IEC 50(16).

time protection: the distance
corresponding to the farther
end of each step or zone.
(Contd...)
RECOMMENDED PROTECTION TERMINOLOGY xxxiii
Reactance Relay A specific type of distance

protection relay whose polar
characteristic, when plotted on
an R-X diagram, is a straight
line running parallel to the R
axis.
Reclosing Relay A relay designed to close a

circuit-breaker in accordance
with a predetermined sequence
after the circuit-breaker has
been opened by the operation
of protective equipment.
Relay Backup Protection located either locally May comprise part of a local
or remotely which will operate backup scheme.
for faults within the reach of
the relay being considered in
the event of its failure to
operate.
Relay Burden The loading imposed by the

circuits of the relay on the
energising power source
expressed as the product of
voltage and current (VA or
watts if DC) for a given
condition, which may be either
at setting or at rated current or
voltage.
Relaying Point A location on a power system

equipped with means of
deriving information, e.g.
current and/or voltage
transformers, and relay
equipment to which this
information is supplied.
Reliability The quality of a protective

system that ensures it will
operates whenever the specific
conditions required for it to
operate are present.
(Contd...)
xxxiv POWER SYSTEM PROTECTION AND COMMUNICATIONS
Remote Backup A form of backup protection in Similar to IEC 50, Section

Protection which determination and 448-01.
initiation of the required action
takes place at a location other
than that at which the main
protection is situated.
Replica Impedance Components of a relay

simulating the magnitude and
phase angle of a section of the
protected zone impedance to
ensure correct reach and/or
orientation of the comparator
polar characteristic.
Resetting Time The time which elapses
between the disappearance of
the abnormal conditions which
caused the operation of the
protection and the restoration
of the protection to its initial
condition.
Residual A method of ensuring that fault Compensation is provided by
Compensation distance relays measure adding a portion of the
accurately irrespective of the residual current from all three
distribution of the return fault phases to the current in the
current between the various faulted phase.
power system neutrals.
Residual Current The vectorial sum, in a Similar to BS 2950 definition.

multiphase system, of all the
line currents.
Residual Voltage The vectorial sum, in a Similar to BS 3950 definition.
multiphase system of all the
line-to-earth voltages.
Reverse Power Protection which detects the
Protection reversal of power flow at a
particular point (or points) in a
system.
Reverse Reach The operating range of a
distance relay in the direction
opposite to that of principal
concern.
(Contd...)
RECOMMENDED PROTECTION TERMINOLOGY xxxv
R-X Diagram The graphic representation of a See polar characteristics.

protective relay’s operating
range in terms of a rectangular
coordinate system.
Schmitt-trigger A level detector incorporating
hysteresis to separate the
trigger and reset thresholds.
Secondary Impedance of the primary
Impedance circuit referred to the
protection relay terminals
using CT and VT
transformation ratio; i.e.
CT ratio
Zs = Zp ×
VT ratio
Secondary Application of the actuating

Injection quantity to a protection relay
directly, rather than via the
primary circuits of current or
voltage transformers.
Security The quality of a protective
system that ensures it will not
operate unless the specific
conditions required for it to
operate are present.
Selective Protection which determines The selectivity is absolute if

Protection that the fault is within its own the protection responds only
zone and isolates that zone to faults within its own zone
only. (unit protection), and relative
if it is obtained by grading the
setting (e.g. time or current) of
the protections of several
zones, all of which may
respond to a given fault.
Similar to IEC 50, Section
1448-01.
Series Regulator Part of a power supply

employing a linear device to
stabilise the output voltage in
the presence of input voltage
and load current variations.
(Contd...)
xxxvi POWER SYSTEM PROTECTION AND COMMUNICATIONS
Sensitivity A measure of the minimum

operating quantity, e.g.
current/voltage necessary to
cause the correct operation of
a relay or complete protection
system.
Solid State Relay An electrical relay or relay
element whose characteristic is
determined by solid state semi-
conductor components (e.g.
diodes, transistors).
Stability The quality of a protective
system that ensures it will not
operate for faults outside the
defined zone of protection.
Starting Element An element of a protective
system which responds to
faults or abnormal service
conditions, and initiates the
operation of other more
selective elements of the
protection.
Static Relay A relay in which the designed The term includes a solid state
response is developed by relay. AS 2481-1981.
electronic, magnetic, optical or
other components without
mechanical motion.
Static Test A test measuring the (relay) Also termed Steady State Test.
response to the steady state
components only, of the
primary circuit actuating
quantity.
Summation The principle of a combination
of a number of inputs in a
polyphase system to produce
one representative output.
Switch-in-Fault An auxiliary protective system Usually associated with the
Protection which, when using distance use of line VTs, by allowing the
relays as the main protection, distance protection starting
ensures high-speed clearance elements to trip directly for a
for closing on a 3-phase fault short time after closing the
close to the relaying point. circuit-breaker. For busbar
VTs, a memory relay may be
used.
(Contd...)
RECOMMENDED PROTECTION TERMINOLOGY xxxvii
Switched Distance Distance protection which

Protection employs one comparator to
perform a number of
measuring functions, the
correct quantities being
supplied to the comparator by
a switching technique
controlled by the starting
elements.
System Impedance At a given measurement IEC 50, Section 448-03,
Ratio location, commonly at one end BS 3950.
of a line, the ratio of the power
system source impedance to
the impedance of the protected
zone.
System Instability A condition of unstable power

transfer leading to collapse of
the system in the absence of
stabilising forces.
Tapped Circuit A multi-ended circuit in which

a number of the terminations
comprise transformers with
circuit-breakers only on the
lower voltage side.
Teed Circuit A multi-ended circuit
controlled by circuit-breakers
at each end, at the same system
voltage.
Through Fault See ‘External Fault’. Not recommended.
Time Lag A delay intentionally BS 3950.

introduced into the operation
of a protective system.
Transfer Tripping The tripping of circuit- Synonymous with Inter-

breaker(s) by signals initiated tripping (American usage).
by protection at a remote Not recommended.
location.
Transient Factor The factor by which the steady IEC 50, Section 448-03.
state value of the core flux of a
protection CT is multiplied
during the period of transient
asymmetric primary current.
(Contd...)
xxxviii POWER SYSTEM PROTECTION AND COMMUNICATIONS
Transverse A unit protection system

Differential applied to parallel connected
Protection circuits in which operation
depends on unbalanced
distribution of currents
between them.
Trip Circuit A method for monitoring the May be done either by a

Supervision continuity of a circuit-breaker supervision relay or by a ‘trip
trip circuit. healthy’ lamp.
Underreach The unintentional reduction of

the protection operating zone
below that indicated by the
relay setting.
Underreaching A form of protection in which

Protection the relays at a given terminal
are intended to operate only for
faults at or nearer than a
specified remote location on
the protected line or
equipment.
Undervoltage Protection that operates at a

Protection predetermined voltage level
which is less than normal.
Unit Protection A protection system, the IEC 50, Section 448-01.

operation and selectivity of
which are solely dependent on
the comparison of electrical
quantities at the boundaries of
the protected section.
Zener Diode A semiconductor diode with a

sharply defined reverse
breakdown voltage, most often
used as a constant voltage
source.
(Contd...)
RECOMMENDED PROTECTION TERMINOLOGY xxxix
DEVICE NUMBERS
The following is a selected list of device numbers commonly used on
protection drawings:
2 Time delay starting or closing coil
3 Checking or interlocking relay
12 Overspeed device
14 Underspeed device
21 Distance relay
25 Synchronising or synchronism check relay
27 Undervoltage relay
30 Annunciator relay
32 Directional power relay/reverse power
37 Undercurrent or underpower relay
40 Field failure relay
46 Reverse phase or phase balance current relay
47 Reverse phase sequence voltage relay
49 Machine or transformer thermal relay
50 Instantaneous earth fault relay
50E Instantaneous earth fault relay
51 AC time overcurrent relay
52 AC circuit breaker
52(a) Circuit breaker auxiliary switch-normally open
52(b) Circuit breaker auxiliary switch-normally closed
55 Power factor relay
56 Field application relay
59 Overvoltage relay
60 Voltage and current balance relay
63 Buchholz gas device
64 Earth fault protection relay
67 AC directional overcurrent relay
68 Blocking relay
74 Alarm relay
76 DC overcurrent relay
78 Phase angle or out-of-step protective relay
79 AC reclosing relay
81 Frequency relay
83 Automatic selective control or transfer relay
85 Carrier of pilot receive relay
86 Locking-out relay
87 Differential protective relay
95 Neutral displacement relay
96 Undervoltage control for tap changers
This page
intentionally left
blank
PART A
POWER SYTSEM PROTECTION
This page
intentionally left
blank
CHAPTER
BASIC PRINCIPLES
1.1 INTRODUCTION TO PROTECTIVE RELAYING

‘Protective Relaying’ or ‘Protection’ is the term that defines the branch of
electric power engineering that is concerned with the detection and
disconnection of short-circuits (faults) and other abnormal conditions on
the power system.
There are three aspects of the design and operation of a power system
that are important in considering the role of protective relaying:
• Normal operation
• Prevention of electrical failure
• Mitigation of the effects of electrical failure.
The term ‘normal operation’ assumes no failures of equipment, no
mistakes of personnel, nor ‘acts of God’. It involves the minimum
requirements for supplying the existing customer load and a certain amount
of anticipated future load. Design of the power system for normal operation
involves major expense for equipment and includes consideration of:
• Choice between hydro, steam, or other sources of power
• Location of generating stations
• Transmission of power to the load
• Study of the load characteristics and planning for its future
growth
• Metering
• Voltage and frequency regulation
• System operation
4 POWER SYSTEM PROTECTION AND COMMUNICATIONS
• Maintenance requirements
• The consequences of equipment or plant failure.
Protection systems must not interfere with or limit the normal
operation of the system but must continuously monitor the system to detect
electrical failure or abnormal electrical conditions.
Further important aspects in the design of the power system are:
• Incorporation of features aimed at preventing failures, and
• Provisions for mitigating the effects of failure when it occurs.
Modern power system design employs both recourse as dictated by
the economics of any particular situation. Notable advances continue to
be made toward greater reliability. However also, increasingly greater
reliance is being placed on electric power. Consequently, even though the
probability of failure is decreased, the tolerance of the possible harm to the
service is also decreased.
The type of electrical failure that causes greatest concern is the
short-circuit, or ‘fault’ as it is usually called, but there are other abnormal
operating conditions peculiar to certain elements of the system that also
require attention. Some of the features of design and operation aimed at
preventing electrical failure are:
• Provision of adequate insulation
• Coordination of insulation strength with the capabilities of
lightning surge arresters
• Use of overhead ground wires and low tower-footing resistance
• Design for mechanical strength to reduce exposure, and to
minimise the likelihood of failure caused by animals, birds,
insects, dirt, sleet, bushfires, etc.
• Proper operation and maintenance practice.
Some of the features of design and operation for mitigating the effects
of failure are:
• Features that mitigate the immediate effects of an electrical failure
1. Design to limit the magnitude of short-circuit current
(a) By avoiding too large concentrations of generating
capacity
(b) By using current-limiting impedance
2. Design to withstand mechanical stresses and heating owing
to short-circuit currents
3. Time-delay undervoltage devices on circuit breakers to
prevent dropping loads during momentary voltage dips
4. Ground-fault neutralisers.
BASIC PRINCIPLES 5
• Features for promptly disconnecting the faulty element

1. Protective relaying
2. Circuit breakers with sufficient interrupting capacity
3. Fuses.
• Features that mitigate the loss of the faulty element
1. Alternate circuits
2. Reserve generator and transformer capacity
3. Automatic reclosing.
• Features that operate throughout the period from the inception of the
fault until after its removal, to maintain voltage and stability
1. Automatic voltage regulation
2. Stability characteristics of generators.
• Means for observing the effectiveness of the foregoing features
1. Automatic oscillographs
2. Efficient human observation and record keeping.
• Frequent surveys as system changes or additions are made, to be sure
that the foregoing features are still adequate.
Thus, protective relaying is one of several features of system design
concerned with minimising damage to equipment and interruptions to
service when electrical failures occur. WHEN WE SAY THAT RELAYS
‘PROTECT’, WE MEAN THAT, TOGETHER WITH OTHER EQUIPMENT,
THE RELAYS HELP TO MINIMISE DAMAGE AND IMPROVE SERVICE
AT A MINIMUM COST. It will be evident that all the mitigation features
are dependent on one another for successfully minimising the effects of
failure.
1.2 POWER SYSTEM PLANT AND LAYOUT

The following section gives an overview of plant and the electrical
characteristic that are relevant to design of the protection system. They
also outline some of the various switching arrangements that are likely to
be encountered on the power system.
The aim is to identify, for the reader, some of the underlying issues
that need to be considered in the design of protection.
1.2.1 Power System Plant
The system for generation, transmission and distribution of electricity is
made up of generators, lines, transformers, reactive plant (capacitors and
static compensators) etc. connected in a network to provide reliable
transport of electrical energy from the generation source to the customer.
The parameters of the plant (size and electrical specifications) and its
associated auxiliary equipment, together with the arrangement of the

network, have a significant influence on the design of the protection system.
Plant impedances and earthing arrangements will determine the
magnitude and path of fault currents. Number and location of current and
voltage transformers will determine the arrangement of protection zones
which, in turn affects the reliability of the whole power system.
Consequently, the protection engineer must have a sound knowledge of
the design of both the electrical plant and the power system in order to
influence the design.
The following paragraphs briefly outline some of the plant and
system design considerations.
1.2.2 Generators
Generators appear in a number of sizes ranging from less than 1 MW
(typically in a cogeneration plant) to 600 MW or more in a large fossil
fuelled station. Generated voltages are generally constrained in the range
of 6.6 kV to 33 kV due to design limitations in the generator insulation
systems. This means that step-up transformers are generally needed to
connect the generator to the transmission system. Important parameters
in the design of protection for the system and the generator are the generator
impedances.
A distinction is made for two conditions, namely the direct and
quadrature axis which cover the positions when the axis of the rotor poles
are in phase with the machine poles, or 90 electrical degrees out of phase.
Fault currents (resulting from a short-circuit on the power system) are
mainly reactive and as they cause drops in the direct axis voltage, we use
the direct axis impedances for fault calculations.
The impedance of the generator varies with time following inception
of a fault, due to the inductive nature of the generator electrical circuit.
The value depends on the time that has elapsed from the inception of a
short-circuit. Impedances in three time zones are specified for calculation
of currents and voltages:
• Subtransient impedance (X d ”)—determines the level for
short-circuit current during the first 1 to 3 cycles after short-circuit
inception.
• Transient impedance (Xd’)—determines the level of current that
a particular generator will contribute to a short-circuit during
the transient period between 3 to 20 cycles.
• Synchronous impedance (Xd)—determines the steady state value
of short-circuit current after the transient period.
BASIC PRINCIPLES 7
The time constant that determines the duration of the subtransient

and transient periods and related ‘offset’ of the short-circuit current is
determined by the inductance and resistance of the generator. It is often
referred to as the X/R ratio of the generator. In a multi-generator system
the X/R ratio is highest near the generation source and reduces as lines
and transformers are interposed between the generation and the load. The
X/R ratio is important in the determination of required current transformer
performance, as you will see in Chapter 4.
For protection calculations, we assume that the nominal terminal
voltage of all machines is acting behind the machine impedance i.e., all
machines are unloaded, and their voltages are all in phase. Some
organisations use the subtransient impedance Xd” for fault calculations
and apply a decrement to reduce current with time, depending on the
measuring and operating time of the protection relays. This can be
appropriate if accurate high speed measurement is required, however, the
majority use the transient impedance Xd’ and assume that the current does
not change during the protection relay operating period. This is adequate
for most applications and these sections are based on the use of the transient
impedance Xd’.
Also of importance in the design of the protection system is the
method of earthing, which determines the paths for earth fault currents in
the system. Generator neutrals are generally earthed through a high
impedance to limit the flow of earth fault currents in the generator windings
and eliminate the damage that this would cause. The path for earth fault
currents on the external power system is established through earthing of
transformer neutral connections.
1.2.3 Transformers
Power transformers of various sizes are located throughout the power
system. ‘Step-up’ transformers convert the generator voltage to levels
suitable for the transmission system which transmits bulk power to the
load centres. Depending on the size of the system, transmission voltages
will range from 132 kV to 500 kV. ‘Step-down’ transformers reduce the
voltages at the bulk load centres to typically 66 kV or 33 kV for distribution
through a ‘subtransmission network’ which supplies the high voltage
distribution system. The distributions system is typically 33, 22 or 11 kV
and supplies distribution substations that transform the voltage to the
customer level.
Large transformers in the generating stations or transmission stations

may be made up of three single phase units or a single three phase unit.
Physical size and transport limitations can frequently determine the choice
that has to be made. Single phase units, as well as taking up more space,
have more complex connection arrangements, particularly with the
connection of the delta or tertiary winding. The external delta connections
are more exposed to faults and failure can result in high short-circuit
currents, which can be disastrous mechanically for the delta windings.
Inter-winding impedances, winding connections (i.e., delta, star,
interstar) and earthing arrangements are important for the protection
engineer. These factors determine the magnitude and path of fault currents
and consequently the ability for protection systems to selectively detect
and clear faults from the system.
It is usually sufficient to use the inductive component of the
transformer impedance in protection calculations and this will usually be
expressed as a per cent or per unit at rating i.e., per cent impedance is the
percentage voltage drop across the transformer at rated voltage and current.
With a short-circuit on the terminals the current will be:
I Rated × 100
%Z
1.2.4 Lines
Impedances, for calculation of fault currents, are the most important line
parameter for protection purposes. These are usually calculated in resistive
and reactive ohms at system frequency and are expressed in the form R + jx
or Z∠Q. The R term is the resistance per phase and the jx term is obtained
from the basic equation of the type
jx = k log10
F Separation of conductors I
GH k × Radius of conductor JK
1
Factors that influence the impedance include the presence of

overhead earth wire and mutual coupling with parallel lines.
1.3 SWITCHING ARRANGEMENTS

Switching arrangements used in a particular power system or individual
stations within the system are influenced by a number of factors and there
is no clear right or wrong arrangement. Factors that need to be considered
are:
BASIC PRINCIPLES 9
• Economic and business investment criteria,

• History of development of the individual power system i.e.,
decisions made in the past can be uneconomic to change because
of widespread changes that may be required,
• Ease and safety of operation and maintenance,
• Security, reliability and quality of supply to the customer,
• Flexibility for future development.
There are many switching arrangements used on the power system,
all of which influence the design of the protection system.
A major consideration for the Protection Engineer is the ability to
establish appropriate protection zones that will selectively isolate faulty
items of plant. In this respect the number and location of current and voltage
transformers is a major consideration. The preference would be to locate
current transformers on each side of the circuit breaker, transformer and
generator so that independent overlapping zones of protection can be
established for each plant item. This practice can result in significant costs,
either in the cost of the plant item itself (e.g., if the CT’s are mounted within
the CB structure) or in the cost of additional space and structures to mount
free standing CT’s in the switchyard. A frequent compromise is to provide
CT’s on one side of the plant. With this arrangement it is possible to achieve
overlapping zones of protection but it can result in “blind spots” or “dead
zones” which requires special measures. For example, with CT’s located
on the line side of a circuit breaker, a fault between the CB and the CT post
will be detected by the busbar protection zone but is outside the line
protection zone. The bus protection will operate to trip the local circuit
breakers but the protection at the other end of the line must detect and
clear the fault from that end.
1.3.1 Single Switching

Each item of plant has its own CB. This arrangement: (see Fig. 1.1)
• is economic in terms of plant requirements,
• is straight forward and safe to operate and maintain,
• has few complications from a protection viewpoint, apart from
selecting the location for CT’s and VT’s.
The major disadvantage is the inflexibility in programming
maintenance. For example, an outage of a CB will result in the loss of the
associated plant item to the system (possible a major generator or
transformer).
FIGURE 1.1 Single switching
1.3.2 Double Switching

Each plant item has two circuit breakers to provide the ability to switch to
either of two bus-bars.
This is a very flexible arrangement (Fig. 1.2) and has the major
advantage that any item of plant can be transferred from bus to bus without
interrupting the circuit that it feeds. Again there are no particular design
problems from a protection viewpoint. It is relatively easy to establish
selective zones for protection of each plant item, the bus-bars and the
incoming and outgoing circuits.
FIGURE 1.2 Double switching

BASIC PRINCIPLES 11
The major disadvantage is the high cost of providing circuit breaker

and their associated auxiliary equipment and space requirements. This
additional expenditure has to be weighed against the gain in revenue or
convenience of keeping generation and other plant in service during
outages of circuit breakers or bus-bars for maintenance or as a result of
plant failure.
Some savings can be achieved by using a combination of single and
double switching. For example, it can be argued that the generator could
be single switched and any maintenance requirements on the CB would
be to coordinate with generator maintenance. In this case the generators
shown in the double switched arrangement, two CB’s could be eliminated
by single switching the generators to alternate bus-bars.
1.3.3 Mesh Layout
This scheme (Fig. 1.3) has most of the advantages of a double bus layout,
in that all plant can be kept in service for the outage of any one CB. But, it
requires only one CB for each item of plant in its simplest form compared
with the two CB’s for each item of plant in the double switched
arrangement.
The limit on the number of items of plant in a mesh layout is usually
around six in order not to prejudice the system in the event of outages e.g.,
if CB A is open for maintenance and a fault occurs on Feeder 1, the system
G1
Feeder 1
G2
FIGURE 1.3 Six circuit breaker mesh
is left with generator 2 disconnected. The mesh layout is flexible and uses
less CB’s than the double switched arrangement. There are no particular
design problems from a protection viewpoint. It is relatively easy to

establish selective zones for protection of each plant item and the incoming
and outgoing circuits provided current transformers are provided with
each circuit breaker and plant item and, depending on the protection
selected, voltage transformers are provided in the outgoing lines.
1.3.4 1½ CB Switching
A more elaborate system than the mesh system is the 1½ CB arrangement
(Fig. 1.4).
FIGURE 1.4 1½ Circuit breaker
This arrangement uses more CBs than the mesh arrangement but
gives better reliability for faults in the transmission lines or generation
plant. Again, provided current transformers and voltage transformers are
carefully located the protection arrangements is straight forward.
1.3.5 Transfer Bus Arrangement
This arrangement (Fig. 1.5) is applicable to stations where there are a large
number of feeders. It permits more flexibility than the single switched
arrangement as any feeder may be kept in service while its CB is out of
service, by using the transfer bus and connecting the feeder either in parallel
with another feeder or to a spare CB.
The system is more complex to operate and can require switching of
current transformers and protection circuits through auxiliary switches
on the transfer isolators, to maintain adequate protection on the feeders.
Problems can also arise with the operation of earth fault protection
when feeders are operated in parallel due to the unbalance in load currents
BASIC PRINCIPLES 13
giving rise to artificial ‘earth fault current’ in the relay circuits. Special
operating procedures may be required to overcome this problem.
FIGURE 1.5 Transfer bus
1.4 THE FUNCTION OF PROTECTIVE RELAYING

The function of protective relaying is to cause the prompt removal from
service of any element of a power system when it suffers a short-circuit, or
when it starts to operate in any abnormal manner that might cause damage
or otherwise interfere with the effective operation of the rest of the system.
It achieves this through relays and protection schemes that measure power
system quantities, detect a fault or abnormal condition and open (trip)
appropriate circuit breakers.
Circuit breakers are generally located so that each generator,
transformer, bus, transmission line, etc., can be completely disconnected
from the rest of the system. These circuit breakers must have sufficient
capacity so that they can carry momentarily the maximum short-circuit
current that can flow through them, and then interrupt this current; they
must also withstand closing in on such a short-circuit and then interrupting
it according to certain prescribed standards.
Fusing is employed where protective relays and circuit breakers are
not economically justifiable.
A secondary function of protective relaying is to provide indication
of the location and type of failure. Such data not only assists in expediting
repair but also, by comparison with human observation and automatic
oscillograph records, they provide means for analysing the effectiveness
of the fault-prevention and mitigation features including the protective
relaying itself.
1.5 PRINCIPLES OF PROTECTIVE RELAYING

The protection system can be divided into two main groups:
• ‘primary’ relaying
• ‘back up’ relaying.
Primary relaying is the first line of defence, whereas back up relaying
provides for failure of the primary protection to clear the fault or
abnormality, either through failure of protection equipment or primary
plant.
1.5.1 Primary Relaying
Fig. 1.6 illustrates primary relaying.
Circuit Low voltage High voltage High voltage
breaker switchgear switchgear switchgear
Generator Power transformer Transmission line
FIGURE 1.6 Single line diagram of a portion of an electric

power system showing primary relaying
Observation:
• Circuit breakers are located in close proximity to each power
system element. This provision makes it possible to disconnect
only a faulty element. Occasionally, a breaker between two
adjacent elements may be omitted, in which event both elements
must be disconnected for a failure in either one.
• A separate zone of protection is established around each system
element. The significance of this is that any failure occurring
within a given zone will cause the ‘tripping’ (i.e., opening) of all
circuit breakers within that zone, and only those breakers.
It will become evident that, for failures within the region where
two adjacent protective zones overlap, more breakers will be
tripped than the minimum necessary to disconnect the faulty
element. However, if there were no overlap, a failure in a region
between zones would not lie in either zone, and therefore no
BASIC PRINCIPLES 15
breakers would be tripped. The overlap is the lesser of the two

evils. The extent of the overlap is relatively small, and the
probability of failure in this region is low; consequently, the
tripping of too many breakers will be quite infrequent.
• Adjacent protective zones of Fig. 1.6 overlap around a circuit
breaker. This is the preferred practice because, for failures
anywhere except in the overlap region, the minimum numbers
of circuit breakers need to be tripped. When it becomes desirable
for economic or space-saving reasons to overlap on one side of a
breaker, as is frequently true in metal-clad switchgear, the relaying
equipment of the zone that overlaps the breaker must be arranged
to trip not only the breakers within its zone but also one or more
breakers of the adjacent zone, in order to completely disconnect
certain faults.
1.5.2 Back up Relaying

Back up relaying is intended to operate when a system fault is not cleared
in due time because of failure or inability of the main protection or the
associated protection to operate.
A clear understanding of the possible causes of primary-relaying
failure is necessary for a better appreciation of the practices involved in
back up relaying. When primary relaying fail several things may happen
to prevent primary relaying from causing the disconnection of a power
system fault. Primary relaying may fail because of failure in any of the
following:
• Current or voltage supply to the relays
• DC tripping-voltage supply
• Protective relays
• Tripping circuit or breaker mechanism
• Circuit breaker.
It is highly desirable that back up relaying be arranged so that
anything that might cause primary relaying to fail will not also cause failure
of back up relaying. Two principles are applied:
• Remote back up
• Local back up.
With remote back up the back up relays are located so that they do
not employ or control anything in common with the primary relays that
are to be backed up. So far as possible, the practice is to locate the back up
relays at a different station. Consider, for example, the back up relaying
for the transmission line section EF of Fig. 1.7. The back up relays for this
line section is normally arranged to trip breakers A, B, I, and J. Should

breaker E fail to trip for a fault on the line section EF, breakers A and B are
tripped; breakers A and B and their associated back up relaying equipment,
being physically apart from the equipment that has failed, are not likely to
be simultaneously affected as might be the case if breakers C and D were
chosen instead.
Station K
A C G I
E F
B D H J
FIGURE 1.7 Illustration for back up protection
The back up relays at locations A, B, and F provide back up protection

if bus faults occur at station K. Also, the back up relays at A and F provides
back up protection for faults in the line DB. In other words, the zone of
protection of back up relaying extends in one direction from the location
of any back up relay and at least overlaps each adjacent system element.
Where adjacent line sections are of different length, the back up relays
must overreach some line sections more than others in order to provide
back up protection for the longest line.
A given set of back up relays will provide incidental back up
protection for faults in the circuit whose breaker the back up relays control.
For example, the back up relays that trip breaker A of Fig. 1.7 may also act
as back up for faults in the line section AB. However, this duplication of
protection is only an incidental benefit and is not to be relied on to the
exclusion of a conventional back up arrangement when such arrangement
is possible; to differentiate between the two, this type might be called
‘duplicate primary relaying’.
A second function of back up relaying is often to provide primary
protection when the primary-relaying equipment is out of service for
maintenance or repair.
It is perhaps evident that, when back up relaying functions, a larger
part of the system is disconnected than when primary relaying operates
correctly. This is inevitable if back up relaying is to be made independent
of those factors that might cause primary relaying to fail. However, it
emphasises the importance of the second requirement of back up relaying,
that it must operate with sufficient time delay so that primary relaying
will be given enough time to function if it is able to. In other words, when
a short-circuit occurs, both primary relaying and back up relaying will
BASIC PRINCIPLES 17
normally start to operate, but primary relaying is expected to trip the

necessary breakers to remove the short-circuited element from the system,
and back up relaying will then reset without having had time to complete
its function. When a given set of relays provides back up protection for
several adjacent system elements, the slowest primary relaying of any of
those adjacent elements will determine the necessary time delay of the
given back up relays.
Local back up provides for the initiation of the required action at
the same location as that at which the main protection is situated. Local
back up usually involves the provision of two completely independent
(duplicate) protection systems including relays, current transformers,
circuit breaker trip coils, etc.
For many applications, it is impossible to abide by the principle of
complete segregation of the back up relays. Then one tries to supply the
back up relays from sources other than those that supply the primary relays
of the system element in question, and to trip other breakers. This can
usually be accomplished; however, the same tripping battery may be
employed in common, to save money and because it is considered only a
minor risk.
1.6 UNIT AND NON-UNIT SCHEMES

The purpose of an electrical power generation system is to distribute energy
to a multiplicity of points for diverse applications. The system should be
designed and managed to deliver this energy to the utilisation points with
both reliability and economy. As there is a natural conflict between these
two requirements, some compromise is necessary. Reliability in system
design is very important and although it is possible to achieve very high
reliability, the economics of doing so due to the excess plant required are
prohibitive. Several ways of improving security of supply without adding
too much to the costs are by:
• improving plant design
• increasing the spare capacity
• arranging alternative circuits to supply loads.
Also such division of the system into zones, each controlled by its
own switchgear in association with protective gear, provides flexibility
during normal operation and ensures a minimum of dislocation following
a breakdown.
In practical power systems any fault condition, especially a
short-circuit, is a potential threat to a secure supply as such a condition
cannot only disrupt supply to consumers but can also cause irreparable
damage to very expensive equipment. The importance of removing such

abnormal conditions as rapidly as possible, is therefore, quite obvious. This
is where the protective gear plays its part.
It is the function of protective gear to detect and initiate action to
remove disturbances, as soon as it is practicable. Protection is therefore
applied in overlapping zones to cover the system completely, leaving no
part unprotected. Another important requirement of the protective
equipment is that only the faulted section should be disconnected and
protective devices must therefore be selective i.e., when a fault occurs the
protection is required to select and trip only the nearest circuit breakers.
This property of selective tripping is also called discrimination and is
achieved by two general methods.
1. Non Unit Schemes
These are invariably time-graded systems that utilise information
(voltages and currents) derived from a particular point on the system.
Protection systems in successive zones as shown in Fig. 1.8 are arranged to
operate in times that are graded through the sequence of equipments to
that upon occurrence of a fault, although a number of protective equipments
respond, only those relevant to the faulted zone complete the tripping
function. The others make incomplete operations and reset. Distance
protection and time graded overcurrent devices are prime examples of
non-unit protection.
End Zone
Z3A
Z2A
Z1A
Y
A X B
Time
Z1B
Z2B
Z3B
End Zone
Zone 1 = 80-90% of protected line
Zone 2 = Protected line + 50% of shortest line
Zone 1 = Protected line + longest second line + 25% of third line
X = Circuit breaker operating time
Y = Discriminating time
FIGURE 1.8 Protective systems arranged in successive zones

BASIC PRINCIPLES 19
2. Unit Protection
These are schemes that respond to fault conditions lying within a
clearly defined zone. They utilise information from two or occasionally
more points in a system. In most cases a unit protection system involves
the measurement of quantities at each end of the zone, and the transmission
of information between the equipment at zone boundaries. Examples of
unit protection are differential current relays where the current entering a
zone is compared with that which leaves it. Also phase comparison carrier
protection is another example.
1.7 ZONES OF PROTECTION

The protected zone is that part of a power system guarded by a certain
protection and usually contains one or at the most two elements of the
power system. For a non-unit scheme, the zone lies between the current
transformers and the point or points on the protected circuit beyond which
the system is unable to detect the presence of a fault (Figs. 1.9 & 1.10). For
a unit scheme, the zone lies between the two or several sets of current
transformers and the point or points which together with the relays
constitute the protective system (Fig. 1.11).
A B
Protected zone
FIGURE 1.9 Protected and back up zones of a non-unit system of protection
A C
Y
R R
Protected Back up
zone zone
FIGURE 1.10 Application of a non-unit scheme of protection (i.e. distance

protection with its associated VTs on the line side of the
isolator) and the standby protection zone of the normally
shorted standby protection
Standby protection
zone
Line
Voltage
transformer
X Y
X = Main protection relay

Y = Standby protection relay
FIGURE 1.11 Protected zone of a unit system protection
1.8 COMMON TERMINOLOGIES

A list of Recommended Terminology is included at the beginning of the
book. Some of the terms that are important for understanding the basic
principles of the protection system are:
Stability This term refers to the ability of the system to remain
inoperative to all load conditions and faults external to
the relevant zone. This quality is present in unit system,
as they remain inoperative under all conditions, with
faults outside their own zone. However, non-unit
systems can respond to faults anywhere on the power
system.
Selectivity Protection is arranged in zones so as to assure no part is
left unprotected. When a fault occurs the protection is
required to select and trip the nearest circuit breakers
only. Also known widely as ‘Discrimination’. In the non-
unit systems the discrimination is not absolute, but it is
dependant on responses of a number of similar systems,
all of which respond to a given abnormal condition.
However, for the unit systems, the discrimination is
absolute and it is able to detect and respond to abnormal
condition occurring within the zone of protection.
Sensitivity This term is frequently used when referring to the
minimum operating current of a complete protective
system. Hence protective system is sensitive, if the
primary current is low. The requirements of all relays
should be quite sensitive for reliable operation. This term
is usually expressed in amperes referred to the primary
BASIC PRINCIPLES 21
circuit or as a percentage of the rated current of the

current transformers.
Reliability Power system represents a large capital investment and
in order to get maximum return it must be loaded to its
maximum. The purpose of power system is not only to
supply energy but also to keep the system in full
operation, in order to give the best service to the
consumers and earn revenue for the supply authority.
Failure is not confined to the protective gear but may
also be due to the failure of the circuit breaker. Hence
every component involved in fault clearance can be
regarded as a source of failure.
Failures can be reduced by:
• reliable designs
• regular maintenance
• site testing.
Speed The objective of speed is to safeguard continuity of
supply. Hence if fault can be isolated in the shortest time,
the greater the system can be loaded. Fig. 1.12 shows
typical values of power that can be transmitted as a
function of fault clearing times for various types of faults.
It can be seen that the fault involving phases has marked
effect on stability compared with the line-to-earth faults.
The other advantage of having fast clearance times is
that unnecessary changes can occur in the system due
to:
• high fault arc
• burn copper conductors
• machine or transformer lamination weld.
Load power
Phase-earth
Phase-phase
Two phase-earth
Three phase
Time
FIGURE 1.12 Typical values of power that can be transmitted as
a function of fault clearance time
Fault currents can cause irreparable damage if allowed
to continue for more than a few seconds. Hence fast
fault clearance is imperative.
CHAPTER
2
Network Analysis and Fault
Calculations
2.1 FAULTS ON POWER SYSTEM

A fault is the intentional or unintentional connecting together of two or
more conductors that ordinarily operate with a difference of potential
between them. The connection between the conductors may be by physical
metallic contact or it may be through an arc. At the fault, the voltage
between the two parts is reduced to zero in the case of metal-to-metal
contacts or to a very low value in case the connection is through an arc.
Currents of abnormally high magnitude flow through the network to the
point of fault. These short-circuit currents will usually be much greater
than the designed thermal ability of the conductors in the lines or machines
feeding the fault. The resultant rise in temperature may cause damage by
the annealing of conductors and by the charring of insulation.
In a power system consisting of generators, circuit breakers,
transformers, transmission and distribution circuits, it is inevitable that
sooner or later in such a large network some failure will occur somewhere
in the system. The probability of such failures is more on the power
transmission lines, because of their greater length and bare exposure to
atmosphere.
A fault or short-circuit may occur due to:
• deterioration of insulation
• damage due to unpredictable causes such as perching of birds,
accidental short-circuiting by snakes, tree branches, bush fires
etc.
• abnormal voltage viz., lightning or switching surges.
NETWORK ANALYSIS AND FAULT CALCULATIONS 23
However, faults must not be confused with overcurrent. The latter

implies that loads greater than the designed values have been imposed on
the system. Under such conditions voltage at the load point or in it vicinity
may be low, but not zero. The currents in the overloaded equipment are
high and may exceed the thermal design limit. Nevertheless, such currents
are substantially lower than in the case of a fault. Service frequency may
be maintained, but at below-standard voltage.
2.2 FAULTS TYPE

Faults may be classified as permanent or transient. Permanent faults are
those in which insulation failure or structure failure produces damage that
makes operations of the equipment impossible and requires repairs to be
made. Transient faults are momentarily faults that may be removed by
de-energising the equipment for a short period of time; short-circuits on
overhead lines frequently are of this nature.
In general, faults on transmission systems may be categorised under
two headings: Series and shunt type. Series faults may involve single-pole
switching and one or more conductor opening. These conductors may be
at one busbar or at different busbars. They may occur either due to breaking
of the conductors or through the action of the circuit breakers and other
devices that may not result in the opening of all the three phases
simultaneously. Series faults form some sort of unbalance in the system
impedances and does not involve either the earth or any interconnection
between phases. Shunt faults such as single-phase-to-ground, two-phase-
to-ground, phase-to-phase, three-phase faults with or without ground and
their combination, again form some sort of unbalance between phases or
between phases and ground. These faults may occur either through
impedances or direct short-circuits.
The most serious result of a major uncleared fault is fire that may
not only destroy equipment but may spread into the system and cause
complete outage. Short-circuit may have the following consequences:
• Reduction of line voltage, which will lead to the breakdown of
supply to the consumer.
• Damage may be caused to the elements of the system.
• Damage to other apparatus due to overheating and abnormal
mechanical forces.
• Make the system unstable.
• Reduction in voltage, causing relays with pressure coil to
maloperate.
• Reduction in voltage on healthy feeders connected to the system

having fault, which may cause consumer’s motor to draw
excessively large currents.
In order to achieve designs that result in a reliable system it is of
importance to have some idea of the frequency of the incidence of faults
on different equipment in a power system. A typical analysis of fault types
is:
Overhead lines 50%
Cables 10%
Circuit breakers 15%
Transformers 12%
Current and voltage transformers 2%
Control equipment 3%
Others 8%
It has already been mentioned that the two types of shunt and series
faults may produce balanced or unbalanced currents and hence they can
be classified as balanced or unbalanced faults. However, in practice,
majority of the faults that occur on power transmission systems are
unsymmetrical shunt faults. The frequency of occurrence of these faults is
as follows:
Single line-to-ground 90%
Line-to-line 5%
Line-to-line-to-ground 3%
Line-to-line-to-line 2%
2.3 FAULT LEVEL CALCULATIONS

It is usual to express the short-circuit capacity in kVA or MVA. The short-
circuit level is obtained from the product of the greatest r.m.s. current that
can be interrupted and the r.m.s. voltage across the contacts immediately
after final arc extinction.
Short-circuit MVA = √3 × (Nominal kV) × ISC × 10–3
The Thevenin’s equivalent circuit that represents the system is an
e.m.f. equal to the nominal line voltage divided by 3 in series with an
inductive reactance of
XTH =
eNominal kV/ 3 j × 1000 Ω
I SC
(Nominal kV) 2
= Ω
Short-circuit MVA
For ease of calculation, it is usual to express all impedances on a

common base. If base kilovolts is equal to nominal kilovolts, then
Base MVA
XTh = per unit
Short-circuit MVA
I base
= per unit
I SC
The elements of a power system are specified as follows:
(a) Generators and transformers are shown in percentage impedance
on rating.
(b) Feeders and interconnectors are based on actual impedance per
phase.
(c) Reactors are based on voltage drop at the rated current.
To convert these to p.u. values on a common base.
Z% MVA base
(i) Z p.u. = ×
100 MVA rating
MVA base
(ii) Z p.u. = Z Ω ×
kV 2
V MVA base
(iii) Z p.u. = R ×
IR kV 2
Let us consider a typical impedance value of the components of a
power system as indicated in Fig. 2.1.
A B Transformer C D E
Feeder Reactor
0.009 W
42.5 V
4 MVA 750 A
7% 3.3 kV
FIGURE. 2.1 Typical impedance value of the componet of a power system
Using a 10 MVA base

Generator: 25 MVA, 30%
30 10
XG = × = 0.12 p.u.
100 25
Interconnector: 0.04 Ω, 11 kV
10
XI = × 0.04 = 0.0033 p.u.
112
Transformer: 4 MVA, 7%
7 10
XT = × = 0.175 p.u.
100 4
Feeder: 0.009 Ω, 3.3 kV

10
XF = × 0.009 = 0.0083 p.u.
3.3 2
Reactor: 42.5 V, 750 A.
10 42.5
XR = 2
× = 0.052 p.u.
3.3 750
10
Fault at A= = 83.33 MVA
0.12
10
Fault at B= = 81.10 MVA
0.12 + 0.0033
10
Fault at C= = 33.52 MVA
0.12 + 0.0033 + 0.175
10
Fault at D= = 32.62 MVA
0.12 + 0.0033 + 0.175 + 0.0083
10
Fault at E= = 27.89 MVA
0.12 + 0.0033 + 0.175 + 0.0083 + 0.052
Tutorial problem Example 1, gives a further example of the
calculations.
For more complex networks with a larger number of components,
the process of ‘network reduction’ to obtain the Thevenin equivalent is the
same as in the above examples and is just a matter of applying normal
circuit analysis theorems. Tutorial Example 2 gives a worked example to
demonstrate the techniques commonly used.
2.4 LIMITING SHORT-CIRCUIT LEVELS

When a short-circuit to earth or between phases occurs the current is limited
by the system impedance (which is fundamentally the impedance of the
alternators, bus-bar interconnectors, transformers and feeders). The
impedance of a small system with limited generator capacity may be
sufficient to limit the short-circuit kVA at any point to a value that the
circuit breakers are capable of interrupting. In large systems, however,
additional impedance may be required. This is provided by reactors that
limit the short-circuit current to a value that can be interrupted by the
breakers before damage to plant occurs.
There are three possible locations for reactors:

(a) in series with the alternators
(b) in series with the feeders
(c) between the bus-bar sections.
(a) This scheme is not commonly used as:
(i) modern power station alternators have sufficient leakage
reactance to withstand short-circuit
(ii) under normal healthy conditions large voltage drop and
power loss in each reactor occurs due to load current
(iii) in case of short-circuit on or near the bus-bar end of a feeder,
large fault current causes large voltage drops in the reactors
which reduce the bus-bar voltage to such a low value that
the alternators can fall out of synchronism.
(b) These are not commonly used for the following reasons:
(i) normally there are hundreds of feeder circuit breakers but
only a few alternator circuit breakers. It is essential to limit
the short-circuit which the feeder breakers will have to
interrupt in order to reduce cost
(ii) in case of fault there is large voltage drop in its reactor with
small reduction in bus-bar voltage and synchronism is not lost
(iii) feeder reactor do not protect the alternators against bus-bar
faults, however leakage reactance should afford enough
protection.
The disadvantage of using feeder reactors are:
(i) large voltage drop and power loss
(ii) in case of number of alternators, the value of reactance has to
be increased to keep the short-circuit levels within the ratings.
(c) These are the most commonly used and gives all the advantages without
the disadvantages of (b) above.
2.5 TRANSIENTS DURING A BALANCED FAULT

In order to consider the basic points a simple circuit model of an a.c.
generator (or alternator) is shown in Fig. 2.2. The voltage e(t) is a fictitious
generated voltage that is assumed R X
to be dependent only on the speed
of the machine and the value of the +
field current. The resistance R is the e(t)
–
a.c. resistance of the machine
winding. The reactance X = jxL is
also fictitious, and its value is FIGURE 2.2 Approximate equivalent circuit
chosen to fit the problem, as follows: of an a.c. generator (or synchronous motor)
Xd is synchronous reactance (steady state value)

Xd’ is transient reactance (3–10 cycles after sudden changes on the
machine)
Xd” subtransient reactance (1–3 cycles immediately after sudden
changes on the machine)
The various values of X are represented in Table 2.1. This value
depends upon whether condition of steady state or rapid change is being
considered.
Suppose a short-circuit is suddenly applied at the terminals of an
a.c. generator. The power frequency that flows in anyone of the windings
will vary with time as shown in Fig. 2.3. The current initially has a high
magnitude, falls in value rapidly in the first 5–15 cycles, and eventually
comes to steady state.
Subtransient Transient Steady state
1
X²d
1
X¢d
Current
1
Xd
Time
FIGURE 2.3 The circuit in a short-circuited a.c. generator
TABLE 2.1: Typical sequence reactance values for

synchronous machines
Two Pole Turbine Four Pole Salient Pole Salient Pole Synchronous
Generator Turbine Generator Generator Compensators
Generator with Dampers Without
Dampers
Low Av. High Low Av. High Low Av. High Low Av. High Low Av. High
Xd 0.95 1.2 1.45 1.0 1.2 1.45 0.6 1.25 1.5 0.6 1.25 1.5 1.25 2.65 2.65
Xd’ 0.12 0.15 0.21 0.2 0.23 0.28 0.2 0.3 0.5 0.2 0.3 0.5 0.3 0.48 0.6
Xd” 0.07 0.09 0.14 0.12 0.14 0.17 0.13 0.2 0.32 0.2 0.3 0.5 0.19 0.32 0.36
X_ 0.07 0.09 0.14 0.12 0.14 0.17 0.13 0.2 0.32 0.35 0.48 0.65 0.18 0.31 0.48
X0 0.01 0.03 0.18 0.015 0.08 0.14 0.03 0.18 0.23 0.03 0.19 0.24 0.025 0.14 .018
The phenomena that occur with a synchronous machine that make

necessary use of the three aforementioned values of reactance involve
transients in the field and in the rotor of the machine. An analysis of this
behaviour is mathematically considered as follows:
The voltage e(t) is assumed to be
e(t) = Em sin (ωt + α)
It can be seen that at time t = 0, a d.c. term exist whose initial
magnitude may be equal to the magnitude of the steady state current term.
The transient current i(t) is given by:
i(t) = (Em/Z)[sin (ωt + α – θ) – sin (α – θ) exp (–Rt/L)]
where Z = √(R2 + ω2L2)
and θ = tan–1 (ωL/R)
The worst possible case occurs when θ – α = (i) 0° or (ii) 90°.
(i) In this case the transient current is given by
i(t) = (Em/Z) sin ωt
The waveform for this case is shown in Fig. 2.4. (i).
(ii) For this case, the current wave form [Fig. 2.4 (ii)] approaches twice
the steady state maximum value just after the initiation of
short-circuit. The transient current is given by
i(t) = (Em/Z)(– cos ωt + exp (– Rt/L)]
For small values of t, exp(– Rt/L) = 1
Thus, i(t) = (Em/Z)(1 – cos ωt)
∴ i(t) = 2Em/Z
i i
t
t
(i) (ii)
FIGURE 2.4 Short-circuit wave form for case (i) and (ii)
Looking at the response of the Figs. 2.3 and 2.4, we can note that the
reactance of the machine appears to be time varying. The symmetrical r.m.s.
short-circuit currents are usually desired. These are calculated using the
synchronous, transient and sub-transient reactance (Fig. 2.5).
Steady state current = I = E/Xd = 1.414

Transient current = I′ = E/Xd’ = 2.83
Sub-transient current = I″ = E/Xd″ = 3.54
i(t)
Xd
2l² x(t)
lmax(t)
2l¢
Xd¢
i(t)
2l
X²d t
Subtransient Transient Steady state
FIGURE 2.5 Symmetrical short-circuit and reactances for a synchronous machine
In fault calculations the initial symmetrical r.m.s. current, which is

the sub-transient current is usually desired. On occasion, the symmetrical
r.m.s. current must be approximated at a time of (say) five cycles after
fault occurs. Then the transient current is used as approximation.
For calculations of fault current in a power system, each generator
is represented as an e.m.f. E in series with an approximated reactance
(usually Xd”). The reactance of all transformers and lines are inserted into
the single line diagram and the short-circuit calculation is performed using
any valid circuit solution method.
2.6 SEQUENCE NETWORKS FOR CALCULATION OF

UNBALANCED FAULTS
Up to date, our analysis of fault currents and voltages has been based on a
three phase balanced short-circuit and the equivalent impedance network
has represented the reactances of one phase. However, the majority of faults
on the power system involve one or two phases and earth. This means that
the currents and voltages are no longer a balanced three phase system. The
maximum unbalance occurs at the fault point where one or more phases will
have their voltage reduced to near zero and fault current will flow.
A technique called ‘Symmetrical Components’ is used to represent
the unbalanced three phase currents and voltages as three symmetrical
systems, each of which can be solved using three phase calculation
techniques.
Symmetrical components were first proposed for the calculation of
fault currents and voltages in 1918, in a paper to the AIEE by C.L. Fortescue.
They have been a major tool for the calculation of fault currents since that
time.
The concepts of symmetrical components are relatively easy to
understand but the application, in dealing with different combinations of
phases and earth for different fault conditions, can be complex and very
time consuming.
As mentioned previously, there are many computer packages
available today to perform fault calculations and most protection
practitioners would have limited need to perform complex manual
calculations. Nevertheless, it is important to understand the concepts and
be able to employ the techniques to small systems.
An overview and some practice with the use of the techniques is
given in the following paragraphs. If the reader wishes to investigate the
technique further there are a number of texts that give a comprehensive
treatment, including the Westinghouse, Electrical Transmission and
Distribution Reference Book.
Symmetrical component analysis is based on the principle that three
sets of balanced (symmetrical) vectors can be used to represent an
unbalanced three phase voltage or current. One set (a, b, c) using the
impedances associated with the normal operation of the power system,
called the positive sequence impedances, a second set using impedances
that would be presented by the network to a set of vectors having a phase
sequence of a, c, b, called the negative sequence impedances and a third
with three vectors of the same angle called the zero sequence impedances.
The sequence impedance of plant in a network can be imagined as
the impedance derived from the equation of the voltage drop resulting
from the circulation of the sequence current through the network
impedance i.e., the voltage could actually be measured by measuring the
voltage drop when the current of a particular sequence is circulated through
the network element.
The following tabulation demonstrates the concepts:

a a a b c
c b b c
Positive phase sequence Negative phase sequence Zero phase sequence

(PPS) (NPS) (ZPS)
Denoted: Ea1 Eb1 Ec1 Denoted: Ea2 Eb2 Ec2 Denoted: Ea0 Eb0 Ec0
Ia1 Ib1 Ic1 Ia2 Ic2 Ib2 Ia0 Ib0 Ic0
For a phase to phase fault between phases b and c of a three phase.

With the positive and negative sequence vectors having the relationship
shown it can be seen that adding the two sequence vectors gives the
resultant fault currents in phases b and c. There is no earth or neutral
connection so the zero phase sequence vectors are zero.
a
b
c
a c b
Ic Ib
c b a
Consider an unloaded generator connected through a reactor as

shown in Fig. 2.6. The equations for the components are as follows:
V1 = Ea – I1Z1
V2 = – I2Z2
V0 = – I0Z0
la
a
ln +
Ea
Zn Ec – – E
+ – b
+ lb
b
c
lc
FIGURE 2.6 Circuit diagram of an unloaded generator

grounded through a reactance
where Ea is the positive sequence no-load voltage to neutral and Z0 is

defined as
Z0 = 3Zn + Zg
0
Zg is the sequence impedance per phase of the generator.
0
The sequence components of current are shown in Fig. 2.7. The
generated e.m.f. in the positive sequence network is the no load terminal
voltage to neutral, which is also equal to the transient and subtransient
internal voltage since the generator is not loaded.
la1
a
Z1 Reference bus
+
E Ea –
Ec – – a +
Z1 + – Eb Ve
+ Z1 lb1 Z1 1
b
c
a
le
lc 2
1
(a) Positive sequence current paths (a) Positive sequence network
la
1
a
Z1 Reference bus
lb1 Z2 Ve
2
Z1 Z1 b
c
a
le
2
lc1
(b) Negative sequence current paths (b) Negative sequence network
la la0
0
lb0 – la a
0
lc0 – la0 Reference bus
Zg
0
3Zn
Zn Zg Z0
0 lb Ve0
0
Zg b Ze
c 0 0
a
le 0
lc 0
(c) Zero sequence current paths (c) Zero sequence network
FIGURE 2.7 Paths for current of each sequence in a generator and

the corresponding sequence networks
The reference bus for the positive and negative sequence networks
is the neutral of the generator. So far as the positive and negative sequence
components are concerned, the neutral of the generator is at ground
potential if there is a connection between neutral and ground having a
finite or zero impedance since the connection will carry no positive and
negative sequence current. Remember the negative sequence paths are the
same as those of positive sequence currents. However, the impedance to
the flow of negative sequence current in a rotating machine is not the same
as that of the positive sequence. This difference is not significant for most
fault calculations, and is usually neglected. Transformers and lines have
equal positive and negative sequence impedances.
The zero sequence currents are a single phase set, and must flow up
through the neutral from ground; consequently the zero sequence network
is different from the other two, both in the values of the impedances and
the structure of the network. In apparatus that has no neutral (i.e., delta
connected) or that has no connection between ground and neutral, there is
no path for the flow of zero sequence current. The zero sequence currents
from each of the three phases of the wye connected generator must flow
through the neutral impedance Zn. The current flowing in the impedance
Zn between neutral and ground is 3I0 and the voltage drop across the neutral
impedance is 3ZnI0. Since the current in the zero sequence network is I0,
the impedance must be 3Zn.
In case of an ungrounded wye connected load, the positive and
negative sequence networks are identical, whereas the zero sequence
network is not connected from the neutral, since the wye is ungrounded.
If the wye is grounded, then there is a path for zero sequence currents
to flow, and the neutral will be connected to the reference bus.
The positive and negative sequence networks for a delta connected
load have the same form as the wye connected load. The zero sequence
current is, however, different. No zero sequence current paths exist for
line currents, however, zero sequence currents may circulate around the
delta. This circulating, path is not normally energised, except in the case of
third harmonic exciting currents of transformer windings in a delta
winding. Third and its odd multiple harmonics are necessarily zero
sequence components.
2.6.1 Transformer
The positive and negative sequence networks of transformers are identical
to the per phase equivalent circuit used in normal balanced analysis. The
form of the zero sequence network is very much dependent on the actual
connection of the primary and secondary. Windings in delta connections

prevent zero sequence line currents, but allow zero sequence phase currents
to circulate around the delta. Windings in ungrounded wye prevent all
zero sequence current flow. In all cases, current flow on one side of the
transformer must produce a current flow on the other side (to balance the
mmf’s produced by each winding). Magnetising paths are not included,
since they are very high impedance paths. Table 2.2 gives zero sequence
equivalent circuit for three phase transformers with different possible type
of connections.
TABLE 2.2: Zero sequence equivalent circuits of three
phase transformer banks
Zero-sequence
Symbols Connection Diagrams
Equivalent Circuits
P Q Z0
P Q
P Q
Reference bus
P Q P Z0 Q
P Q
Reference bus
P Q Z0
P Q
P Q
Reference bus
P Q Z0
P Q
P Q
Reference bus
P Q P Z0 Q
P Q
Reference bus
The phase shift introduced by delta-wye transformers has no effect

on the magnitude of the fault currents at the point of fault, but it should be
taken into account in determining voltages on the far side of the
transformer. Include the normal 30° phase shift for the required
transformation.
Three winding transformers and auto-transformers pose interesting

problems. On considering grounded wye-wye three winding transformers
with a delta tertiary, which has the same equivalent circuit as the grounded
wye auto-transformer with a delta tertiary. The tertiary is assumed to be
brought out to terminals, although this is not always the case (particularly
in the case of auto-transformer). The positive and negative sequence are
shown in Fig. 2.8. The zero sequence network must show that no zero
sequence line currents can flow from the delta tertiary, but that the zero
sequence phase currents in the delta allow the bank to serve as a source of
ground currents. Fig. 2.9 shows the zero sequence network for three
transformer banks, together with diagrams of connections and the symbols
for one line diagram.
ZHV ZLV
ZT
ZT
ZHV ZLV
FIGURE 2.8 Positive and negative FIGURE 2.9 Zero sequence

sequence network for auto-transformer network for auto-transformer and
and three winding transformer three winding transformer
On most fault calculations, the main interest in using this circuit is

to calculate through faults involving ground. Thus the tertiary connection
is often not shown on the positive and negative sequence diagrams. The
tertiary must be shown connected to ground in the zero sequence diagram
to correctly represent the fact that the bank can supply ground current to
the fault. If the delta tertiary were omitted from the bank, then the wye
auto-transformer would have sequence networks of the same form as those
of a wye-wye two winding transformer, which can pass ground fault
currents from one side to the other, but cannot supply zero sequence current
to a fault.
2.6.2 Synchronous Machines

In case of synchronous machines, the only source of voltage on the system
is the positive sequence. Machines will contribute to fault currents whether
operating to produce voltage or operating as spinning reserve. The positive
sequence impedance Z1 is the normal transient/subtransient value.
Negative sequence current set up a rotating magnetic field in the opposite
direction to that of positive sequence currents and which rotates around
the rotor surface at twice the synchronous speed; hence the effective
impedance, being the negative sequence impedance Z2 is different from
Z1. In the absence of information Z2 is approximately 70% of Z1, as the
system impedance will swamp the small difference for system fault
calculations.
The zero sequence impedance Z0 will depend on the nature of the
connection between the star point of the windings and earth and the single
phase impedance of the stator windings in series. Because zero sequence
currents are all in phase, there will be considerable demagnetisation of the
generator iron circuit which means that Z0 is very low, say half of Z1.
Resistors or reactors are frequently connected between the star point of
the windings and earth for reasons associated with protective gear and
limitation of overvoltages and this completely swamps the machine
impedance. Further as the machines are usually connected to the system
via delta-wye transformers, they cannot supply zero sequence currents to
system faults because they cannot flow outside the delta winding.
2.6.3 Induction Motors

Even though there is no external excitation, the flux present within an
induction motor will contribute to the fault current, but the contribution
will decay to zero. For system fault studies, only the very large motors
need be considered and they may be lumped as an equivalent motor with
power rating equal to the sum of the motors considered. Four times full
load current of this motor may be used to give a reasonable figure for the
equivalent source impedance.
2.6.4 Lines and Cables

The positive and negative sequence impedances are the normal balanced
values and are identical because the impedances are independent of phase
order for balanced applied voltages.
The zero sequence impedance depends on the nature of the return
path through the earth if no fourth wire is provided. It is also modified by
the presence of earth wires on the towers.
The zero sequence mutual impedance between parallel circuits can
be appreciable because they share an earth return path, in addition to
mutual coupling effects.
Table 2.3 summarises the expressions to be used to determine fault
current and the forms of interconnection of the sequence networks for
various unsymmetrical fault conditions.
TABLE 2.3: Summary of fault current of sequence

networks for various fault conditions
a a
G b G b
c c
F1 F2 F0 F1 F2 F0
N1 N2 N0 N1 N2 N0
V V
Three phase fault, a–b–c One phase to ground fault
a a
G b G b
c c
F1 F2 F0 F1 F2 F0
N1 N2 N0 N1 N2 N0
Double phase fault, b–c Double phase to earth fault
2.7 CALCULATION OF VOLTAGES IN THE NETWORK

While studying a power system, it is essential to know the short-circuit
currents and kVA the system carries under various fault conditions and at
different points in the system. Since the system remains balanced during
symmetrical (three phase) faults, analysis can proceed on a single-phase
basis. The solution can be obtained directly by normal method of reducing
the network up to point of fault, expressing the system parts in terms of
impedances in per unit values, and then solving the related equations by
network laws. The faults can also be represented by symmetrical
components and the solution obtained by method of symmetrical
components. However, as earlier mentioned various types of
unsymmetrical faults (both shunt and series) do occur. These unbalanced
type of faults can be solved by the use of symmetrical component method.
The faults may occur at the terminals of an unloaded generator or at

any point in the system under different operating conditions; the fault may
be complete dead short or it may be leakage through some impedance. In
the latter case, the fault impedance comes into the circuit at the proper
place depending on the type of fault when solving symmetrical component
sequence networks. When considering the fault at any point in the system,
the fault point should be located in various sequence impedance networks
and voltages and impedances up to the point of fault in each sequence
network should be considered for solution of fault current.
When the network currents have been determined, the sequence
voltage at any point in the network can be found by subtracting the
impedance drops of that sequence from the generated voltage (if any),
taking the neutral point of the network as the point of zero voltage.
The voltages at any point in the network (including the fault point)
are given by:
V1 = Ea – I1Z1
V2 = Ea – I2Z2
V0 = Ea – I0Z0
The actual voltages on each phase can be calculated using the
transformation as covered in standard Power System textbooks.
[Vabc] = [A] [V012]
A similar procedure may be used to calculate the potential rise on
neutral points, given that there is some impedance between the neutral
point and earth. Such an impedance appears only in the zero sequence
network, and then as three times the actual impedance, as described earlier.
2.8 SHORT-CIRCUIT FAULT CALCULATIONS
2.8.1 Tutorial Problems and Solutions

Example 1. A single line diagram of a generating station is shown in Fig. 2.10.
The ratings and percentage reactances of different elements are as indicated. If a
3-phase short-circuit occurs on any feeder near transformer secondary (indicated
by point F), find the short-circuit MVA fed to the fault.
A B C
10 MVA 10 MVA 10 MVA
30% 30% 30%
10 MVA 10 MVA 10 MVA

10% 10% 10%
5 MVA 5 MVA 5 MVA

5% 5% 5%
FIGURE 2.10
Solution. Choose 5 MVA as the base value.
The percentage reactance of each generator on the base MVA
5
= 30 × = 15%
10
The percentage reactance of each reactor on the base MVA
5
= 10 × = 5%
10
The percentage reactance of each transformer on the base MVA
5
= 5 × = 5%
5
When a 3-phase short-circuit occurs at point F on the feeder near
the secondary of the transformer, the reactance diagram will be as shown
in Fig. 2.11(i). This circuit can be further reduced to Fig. 2.11(ii).
A B C
(i) (ii)
FIGURE 2.11 Reactance diagram for example 1
The total percentage reactance from generator neutral up to fault

point F
= (10 + 5)% in parallel with (15 + 5)%
15 × 15
= +5
15 + 15
= 7.5 + 5 = 12.5%
100
Short-circuit (MVA) = MVAbase ×
% Fault reactance
100
=5× = 200.
2.5
Example 2. Consider the network shown in Fig. 2.12(i). With the plant details
listed below, it is required to calculate the three phase current for a fault F on line M.
The example calculations show the application of a number of techniques used in
network reduction. Fig. 2.12(ii) shows the impedance network, with all impedances
converted to a 25 MVA base and Fig. 2.12(iii) the steps in reducing the network to a
single impedance to the fault. Note the use of the star-delta conversion.
E
R
GA
Line H
11/66 kV
D GC
GB Line K 66/3.3/6.6 kV
11/66 kV
Line L
Line
G M
Fault
GD
11/6.6/66 kV
(i) Power system
R
9.38 12.5
A
10 25
20 8 Line H
B 4
6 8
8 3 18.18
Line K Line L C
8 2.29 6.04 5
Line M
3.96
(ii) Equivalent impedance network

12.28 10 20.24
6 5
8
16.33
12.28 20.24
5
12.2 2.5
3.33
2
16.33
(iii) Steps in reducing newtork
FIGURE 2.12 Network reduction
Note: All impedances shown as % on 25 MVA Base

Transformers E 11/66 kV ∆/Y, rating 20 MVA
Test—with short circuit on 66 kV winding
1.1 kV applied to 11 kV winding produced
full load current.
D 11/66 kV, ∆/Y, rating 25 MVA
Impedance 8% on rating (25 MVA)
G 66/11/6.6 kV, Y/Y/∆, rating 30/20 MVA
Impedances—66/11 kV, 10% on 30 MVA
66/6.6 kV, 8% on 20 MVA
11/6.6 kV, 5% on 20 MVA
N 66/6.6/3.3 kV, Y/Y/∆
Impedances—HV-LV 12 %, at 25 MVA
HV-MV 11% at 25 MVA
MV-LV 7% at 25 MVA
Generators A Z1 = j15% at rating, 40 MVA
B Z1 = j20% at rating, 25 MVA
C Z1 = j20% at rating, 27.5 MVA
D Z1 = j8% at rating, 25 MVA
Compensator R Z1 = j25% on 25 MVA
Lines (all % on 25 MVA Base).
H = 10, K = 6, L = 8, M = 5.
For this example we will use a common 25 MVA base for analysis of
the circuit, so all impedances must be converted to this base.
For a three winding transformer, an equivalent diagram of
impedances is used to represent the three windings. For example,
transformer N is represented in the network by three impedances:
8% 3%
66 kV 6.6 kV
4%
3.3 kV
The impedances are obtained from the data provided, by solving
the three simultaneous equations
HV + LV = 11
HV + MV = 12
MV + LV = 7
The equivalent impedance obtained from reduction of the network,
Fig. 2.12 is:
12.2% on 25 MVA base.
25
Fault MVA = × 100 = 204.9 MVA
12.2
204.9 × 10 6
Fault current = = 1782 A.
3 × 66 × 10 3
Example 3. A wye connected balanced 3-phase load is shown in Fig. 2.13. This
load draws 90 A from a balanced 3-phase supply. There are fuses in phase B and Y.
Find the symmetrical components of the line currents.
R
Fuse
Y
Fuse
B
FIGURE 2.13
(i) Before the fuses are removed

(ii) After the fuses are removed.
Solution. (i) Before fuse removal
The system under this condition is balanced and current in each
line is 90 A.
Hence, IR = 90 ∠0°; IY = 90 ∠240°; IB = 90 ∠120°
Since the system is balanced, there will be no negative and zero

sequence components.
IR = IY = IB
0 0 0
1
= [IR + IY + IB]
3
1
= [90 ∠0° + 90 ∠240° + 90 ∠120°]
3
1
= [90 ∠0° + 90 ∠ –120° + 90 ∠120°]
3
=0
1
IR = [I + a2 IY + a IB]
2 3 R
1
= [90 ∠0° + 1 ∠–120° × 90 ∠240° + 1 ∠120° × 90 ∠120°]
3
1
= [90 ∠0° + 90 ∠120° + 90 ∠240°]
3
1
= [90 ∠0° + 90 ∠120° + 90 ∠− 120°]
3
=0
Also, IY = a IR = 1 ∠120° × 0 = 0
2 2
IB = a2 IR = 1 ∠240° × 0 = 0
2 2
The positive sequence components will have finite values.
IR = IR = 90 ∠0°
1
IY = IY = 90 ∠240°
1
IB = IB = 90 ∠120°
1
(ii) After fuse removal
When the fuses are removed the system becomes unbalanced
IR = 90 ∠0°; IY = IB = 0
The sequence currents in the three lines are as follows:
IR = IY = IB
0 0 0
1
= [IR + IY + IB]
3
1
= [90 ∠0° + 0 + 0]
3
= 30 ∠0°
1
IR = [I + a IY + a2 IB]
1 3 R
1
= [90 ∠0° + 0 + 0]
3
= 30∠0°
IY = a2 IR = 1 ∠240° × 30 ∠0° = 30 ∠240°
1 1
IB = a IR = 1 ∠120° × 30 ∠0° = 30 ∠120°
1 1
1
IR = [I + a2 IY + a IB]
2 3 R
1
= [90 ∠0° + 0 + 0]
3
= 30 ∠0°
IY = a IR = 1 ∠120° × 30 ∠0° = 30 ∠120°
2 2
IB = a2 IR = 1 ∠240° × 30 ∠0° = 30 ∠240°
2 2
In spite of the fuse removal there are sequence currents in both the
Y and B phases. They are just mathematical components of the currents.
However the current in the phases are actually zero.
IY = IY + IY + IY
0 1 2
= 30 ∠0° + 30 ∠240° + 30 ∠120°
= 30 ∠0° + 30 ∠–120° + 30 ∠120°
=0
and similarly IB can be shown = 0.
Example 4. A 25 MVA, 11 kV, 3-phase generator has a subtransient reactance of
20%. The generator supplies two motors over a transmission line with transformers
at both ends as shown in the one-line diagram of Fig. 2.14. The motors have rated
inputs of 15 and 7.5 MVA, both 10 kV with 25% subtransient reactance. The
3-phase transformers are both rated 30 MVA, 10.8/121 kV, connection delta/wye
with leakage reactance of 10% each. The series reactance of the line is 100 W.
Note the following:
• Negative sequence reactance of each machine is equal to its subtransient
reactance.
• Omit resistances.
• Select generator rating as base in the generator circuit.
• Zero sequence reactance for the generator and motors are 6% each.
• Current limiting reactors of 2.5 W each are connected in the neutral of
the generator and motor number 2.
Mot
1
Gen
Mot
2
FIGURE 2.14 One-line diagram
• Zero sequence reactance of the transmission line is 300 W.

• Zero sequence reactance of the transformer is equal to its positive
sequence reactance.
Draw the positive, negative and zero sequence networks of the system.
Solution. A base of 25 MVA, 11 kV in the generator circuit requires a 25 MVA
base in all other circuits and the following voltage bases.
121
Transmission line voltage base = 11 ×
10.8
= 123.2 kV
10.8
Motor voltage base = 123.2 ×
121
= 11 kV
The reactances of transformers, line and motors are converted to
per unit values on appropriate bases as follows:
2
Transformer reactance = 0.1 × FG IJ
25 10.8
= 0.0805 p.u.
30 11 H K
100 × 25
Line reactance =
(123.2) 2
= 0.164 p.u.
2
Reactance of motor 1 = 0.25 ×
25
×
10FG IJ
15 11 H K
= 0.345 p.u.
2
Reactance of motor 2 = 0.25 ×
25
×
FG IJ
10
7.5 11H K
= 0.69 p.u.
– – –
Eg Em1 Em2
+ + +
j0.345 j0.69
j0.0805 j0.164 j0.0805
FIGURE 2.15 Positive sequence network
The positive sequence diagram is shown in Fig. 2.15. The negative

sequence reactances are identical to the positive sequence reactances, hence
the network is identical. However, note that in the negative sequence
network there is omission of voltage sources. The negative sequence
network is illustrated in Fig. 2.16.
j0.2 j0.345 j0.69
j0.0805 j0.164 j0.0805
FIGURE 2.16 Negative sequence network
Transformer zero sequence reactance = 0.0805 p.u.

Generator zero sequence reactance = 0.06 p.u.
2
Zero sequence reactance of motor 1 = 0.06 ×
25 FG IJ
×
10
15 H K
11
= 0.082 p.u.
2
25 F 10 I
Zero sequence reactance of motor 2 = 0.06 × ×G J
7.5 H 11 K
= 0.164 p.u.
2.5 × 25
Reactance of current limiting reactors =
(11) 2
= 0.516 p.u.
Reactance of current limiting reactor included in zero sequence
network
= 3 × 0.516
= 1.548 p.u.
Zero sequence reactance of transmission line
300 × 25
=
(123.2) 2
= 0.494 p.u.
The zero sequence network is shown in Fig 2.17.
j1.548 j1.548
j0.082
j0.06 j0.164
j0.0805 j0.494 j0.0805
FIGURE 2.17 Zero sequence network
2.8.2 Further Problems

1. A simple system representing a distribution line is shown in Fig. 2.18.
Assume that the sending end voltage is held constant at 6.6 kV and
the line impedances are as indicated. Load points are A, B and C.
Find the currents and voltages at A, B and C.
2. Faults on network.
Fig 2.19 shows a simple network consisting of a transformer bank
supplying a load over two parallel lines on one of which a fault is
assumed to exist. Neglecting the load find the voltage at each bus
and the current in each line.
6 ohm 6 ohm 6 ohm
A B C
L L L
FIGURE 2.18 Normal distribution circuit with three load points
X fr
I II
Line A
L
Line B
Source X Fault
FIGURE 2.19 One-line diagram of a typical power system with a fault
3. General approach
Using Thevenin’s theorem a general procedure for obtaining fault
currents in any network can be evolved. In Fig. 2.20 a single phase
impedance diagram is shown where F is the fault point and N the
neutral. Find the Thevenin’s equivalent at point F in the absence of
a fault.
0.5
0.1
0.25 0.25
V F
N
FIGURE 2.20 Single phase equivalent circuit in the absence of fault
4. Calculate the symmetrical short circuit current at point F in Fig. 2.21

given the following data:
F
G
Heat
+
Light
FIGURE 2.21 Power system
Generator: 6.6 kV, 20 MVA, Z = (2 + j 25)%.

Motor: Induction, 3.0 kV, 6000 h.p., Z = (2 + j 20)%.
Transformer 1: 6.6/13.6 kV, 20 MVA, Z = (0.4 + j 6)%.
Transformer 2: 13.6/3.0 kV, 7 MVA, Z = (1 + j 7)%.
Transformer 3: 13.6/0.415 kV, 12 MVA, Z = (1.5 + j 10)%.
Line: Z = 0.05 + j 0.35 Ω
Choose 20 MVA as the base value and neglect the heating and
lighting load.
5. A 20 kVA, 440/110 V single phase transformer has a per unit
reactance of 0.08. Calculate the primary current and the short-circuit
kVA, when a zero impedance short-circuit occurs on the secondary
whilst rated voltage is applied to primary. Change the base to 40 kVA
and repeat the calculations.
6. A three-phase, 22 kV alternator having p.u. X of 0.25 at 25 MVA is
connected to the primary of a three-phase transformer having per
unit X equal to 0.15 at 40 MVA. Calculate the primary current and
short-circuit MVA when a three-phase symmetrical short-circuit
occurs on the secondary of the transformer (a) using a 25 MVA base,
(b) using a 40 MVA base.
7. Calculate the short-circuit fault current and short-circuit MVA when
a three phase symmetrical short-circuit occurs at F in the three phase
system represented by the one-line diagram in Fig. 2.22 (a) using a
50 MVA base, (b) using a 100 MVA base.
X = 0.3 pu X = 0.3 pu
Alt Alt
at 50 MVA at 50 MVA
X = 0.3 pu
11/66 kV
at 100 MVA
Fault
8. Using a 50 MVA base, calculate the p.u. X and inductance of reactor

‘x’ required to limit the short-circuit MVA to 100 when a three phase
symmetrical short circuit occurs at F in the three-phase, 50 Hz system
represented by the one-line diagram in Fig. 2.23.
X = 0.2 pu Alt Alt X = 0.25 pu
at 25 MVA A B at 50 MVA
X = 0.00 pu
33/132 kV
at 10 MVA
Fault

CHAPTER
Earth Fault and Interferences
3.1 INTRODUCTION
Safety reliability (continuity) of supply and welfare of plant are the
fundamental reasons or justifications for the protection and control
arrangements provided in electric power systems.
Protection and control viewed in this context obviously covers a
broader field than that concerned with the workings of circuit breakers
relays and other devices used to sense and control faults and abnormalities:
It includes consideration of aspects of system design especially system
earthing; circuit breaker provision; safety earthing; prevention and/or
mitigation of interference with public or private facilities or systems e.g.,
telecommunication systems, pipelines, railway systems, swimming pools
etc.
The protection engineer by virtue of their need to study prospective
and actual faults and abnormalities, which effect a power system, is well
placed, indeed obliged, to be active and accept considerable responsibility
in ensuring that the power system with which they are concerned has been
designed and is operated in a manner which maximises safety, continuity
of supply and welfare of plant. This arises largely because considerations
in these areas are more intimately connected with their work of providing
protection systems than they are with other aspects of system design
construction and operation.
3.2 POWER SYSTEM ARRANGEMENTS AND

CONSTRUCTION FEATURES
3.2.1 General Comments

Power system arrangements evolve overtime from considerations of factors
such as: load size, type and location; energy source type and location;
economics of voltage levels; economic, reliability and other considerations
of interconnections between sources and loads for the provision of the
desired degree of continuity of supply. Adequate planning will ensure
that at any time there is a high probability that the system load can be
carried despite the non-availability of one or more system elements due to
breakdown or maintenance.
However, a deeper look into how the system is interconnected and
operated is required to ensure overall integrity of the system and safety,
particularly during abnormal conditions ranging from overloading to faults.
For example, system design considerations, particularly those of system
stability and plant welfare, are inseparable from protection considerations:
A too slow protection operation may precipitate instability. Furthermore
when applying protection schemes at any point in a system it is necessary
to consider its application in the broader context referred to above and at
the same time, bear in mind that a protection scheme applied at any point
in a system has a duty to the system, as well as the plant it nominally
protects. The relative importance of these duties, to the system and the
plant protected, is related to the voltage level and/or its position in the
system e.g., at the lowest level, say for a 240 V appliance the protection is
almost exclusively concerned with the appliance and its environment,
whereas, at the EHV end of the scale the concern is largely with the stability
of the system and consequently, in this case, the amount of money spent
on protection usually greatly exceeds the particular requirements of the
line or plant concerned.
3.2.2 System Types
Public utility HV electric power systems are almost always three phase
systems, but may be either three or four wire systems and have their
neutrals either solidly earthed or earthed through an impedance.
Australian HV systems are predominantly 3-phase three wire solidly
earthed systems. However, some small sections of these systems usually
have impedance earthing e.g., generators and some HV distribution. Other
small sections need to be treated as unearthed systems during fault
clearance.
EARTH FAULT AND INTERFERENCES 53
Fig. 3.1A shows some typical arrangements used to connect power

sources to loads. The single line system diagram is amplified in the insets
to show the three phase arrangement and the system earthing. Inset 1 shows
a solidly earthed system being fed from the output side of the transformer
by virtue of its neutral being connected directly to earth. This arrangement
could be converted to impedance earthing by inserting an impedance
(inductance or resistance) in the connection to earth.
Other power
& terminal STNs
Main transmission CCT
BKR
Power station
Subtransmission
1 CCT
Zone 1 If
If B BKR B
substation
W W
R R
1
If If If If
N
Resistance
to
Inset1 Earth
If
Distribution Fault current path

substation
HV distribution
11
LV IR B IR¢ IR B
customer 11 W IW¢ IW W
IW
11 R IB¢ IB R
IB N
Metal
LV
structures Earthing system Fuses
Station earthing Customer

HV
GRD load
Earthing
Inset 2 system
Load current flow 3-phase system
FIGURE 3.1A Power system diagram

3.2.3 Protection/Safety Implications of System Arrangements

It can be seen from Fig. 3.1 how the high voltage winding of a zone sub-
station transformer can be back fed from the lower voltage side and how
under these conditions the part of the system it feeds would be unearthed
e.g., with the high side B/T open and only a radial line fed from the bus we
would have an unearthed system feeding out into the public domain. Whilst
this type of condition is not usually created deliberately, because
operational authorities are aware of it, care has to be taken that if it does
occur during a fault clearance sequence, protection exists to detect it and
control it. The hazard to be avoided either by protection operation or by
suitable operating procedures is a conductor on the ground fed from an
unearthed system i.e., a condition where no fault current flows.
The type of system earthing used, greatly influences the currents
and voltages associated with earth faults and consequently the safety
considerations and the characteristics required of the protection schemes.
Recently impedance earthing of more HV distribution systems has
become necessary, because of safety/protection problems associated with
conducting overhead line support structures, e.g., concrete poles. In this
context it should be noted that replacing wood poles (virtually a double
insulated arrangement) with conducting poles markedly increases the
incidence of high current earth faults and with it voltage disturbances to
other parts of the system, thus imposing a requirement of faster fault
clearance and/or reduction of fault magnitude.
3.2.4 System Earthing and Equipment Earthing
Earthing systems serve two distinct but related purposes: System and safety
earthing.
System Earthing
System earthing involves connecting the neutral points of
transformers, generators, reactors, capacitor banks, etc. either directly to
earth or through an impedance, in order to hold the voltages of the system
substantially symmetrical with respect to the earth and to provide a path
for certain earth fault current.
It is possible to operate a high voltage system without having the
system earthed, however faults must then be detected by the displacement
of the voltages of the phases of the system with respect to earth. With this
arrangement the location of the fault usually involves a complex set of
operating procedures.
Where the system neutrals are earthed (directly or through an
impedance) current operated earth fault protection can provide a fast
and/or sensitive means of locating and clearing an earth faults.
The introduction of an impedance into the earthing of neutrals will

limit the magnitude of the earth fault current and consequently reduce
earth potential rises. This is the subject of the next section.
Another effect of the introduction of earthing impedance is increased
difficulty in detecting earth faults within the transformers or on the output
side buses of zone substations and in providing back up from the station,
or stations supplying the zone substation.
Practical solutions of these problems may involve one or more of
the following: Limited choice of system arrangement, more circuit breaker
on the input side of the zone substation and signalling to the source stations.
Zone substation protection is covered in another section.
Safety Earthing
This involves connecting to earth all the conducting metal cases,
frames, supporting structures etc. associated with the generation,
transmission, distribution, utilisation of electric power, to ensure that in
the event insulation failure (fault) these cases, frames, etc. will be held at,
or as close as practicable to the potential of the general mass of the earth.
Otherwise these cases, frames, etc. would assume the voltage of the phase
conductor where the insulation failed and present a hazard to life.
In practice many cases arise where difficulty is experienced in
providing a sufficiently good connection to earth to ensure the elimination
of the possibility of hazardous voltages appearing on equipment frames,
etc. during faults.
The foregoing remarks are made in the context of protection of high
voltage systems, however, the principles involved have application in low
voltage systems and indeed where common or bonded HV and LV earthing
systems are used safety requirements of the LV system will be a major
consideration in the HV earthing system design.
The philosophy in designing LV system earthing has emphasised
keeping the impedance specified through resistance limitations of
connections within them low enough to operate the protective devices
provided (CB’s or fuses) and keep voltage rise on any part of the earthing
system low enough to avoid hazard to life.
The safety aspect has been enhanced by the availability of more
reliable and low cost RCD's which because of their very high sensitivity
and rapid operation can prevent electrocution.
There is no equivalent detection and fault clearance equipment
available for HV systems consequently emphasis has to be placed on
avoiding hazardous voltages and preventing direct contact.
Various forms of earthing systems are depicted in Fig. 3.1B.
Bon
d
e utral
LV n
A
N
Conducting
pole (concrete)
or metal
Bond
May or may not

have a supplementary
earthing system
LV earthing system
HV earthing system
(case of separate HV & LV
earthing systems bonded
for conservation to CMEN)
(a) Some typical earthing systems—CMEN
(Common multiple earthed neutral) system
HV cable
Earthing
Station boundary
electrodes
Buried grid
LV earthing
system
(b) Some typical earthing systems— (c) Some typical earthing systems—
earthing for large station one type of kiosk substation
(Note: Hazard zone may earthing systems
extend beyond station boundary)
FIGURE 3.1B
3.3 EARTH POTENTIAL RISE

3.3.1 Occurrence
Current flowing through the resistance to earth of an earthing system, or
the resistance to earth of the medium of earth contact, causes the earthing
system, or point of contact with the earth, as the case may be, to rise in
potential above that of the general mass of the earth, i.e., above the potential
of remote points.
The gradients of the potential drop away from the fault point and
the source station earthing system will be steep near the earthing system
and the fault point and slight well away from them. The step and touch
potentials adjacent to the earthing system will be the critical ones with
respect to the safety of people who are in the vicinity of earthing system
when an earth fault occurs, nevertheless, the total rise of potential above
remote earth is of concern for people working on metallic circuits, e.g.,
telephone circuits, passing through that part of the gradient area where
the potential rise is significant in relation to personnel safety. The total rise
is the main concern in respect to the rating and treatment of the circuits
entering the hazard zone and/or station. The total rise of potential is termed
the transfer potential. Fig. 3.2 depicts gradients and transfer potentials.
Hazard zone Hazard zone
surrounding surrounding
substation fault point
If
Station VR
earth grid E
(Resistance to E
earse RSE) (Remote point)
Fault point
Telephone
(Resistance to earth RFE)
cable
FIGURE 3.2 (a) Geographic arrangement
V
EPRFP Fault
point
A
Station
E
VAE Remote
EPRS earth
V
FIGURE 3.2 (b) Voltage relative to remote earth

Power
system
impedance Fault point
IF
Power
source VS EPR of
voltage fault point
EPR of station relative to
Station remote earth
RSE grid relative to RFE
earth
grid REM earth
Remote earth
FIGURE 3.2 (c) Simplified electrical circuit transfer to potential
3.3.2 Magnitude
Various factors which influence the magnitude of earth potential rises are
discussed in the following paragraphs.
Earth System Resistances
The resistance to earth of the earthing system depends upon its size
and arrangement and the soil resistivity.
Resistances to earth of earthing systems vary from small fractions
of an ohm (0.001 Ω) for a large power or terminal station to perhaps several
hundred ohms for an isolated distribution structure without supplementary
earthing, e.g., a concrete pole in an area of high soil resistivity.
Station Earthing
Even low earthing system resistances may be associated with
significant voltage rises if the fault current is high enough, e.g., the earthing
system of a particular zone substation with a 40 m × 40 m earthing grid in
an area with 10 ohm metre soil resistivity (resistance to earth 0.2 ohms)
would rise to 2400 volts above remote earth for a maximum fault current
of 12,000 amps. [(refer Fig. 3.3a)]
Constant
voltage
contours
Earthing
grid 40 × 40 m
Hazard zone may extend

outside station boundary
FIGURE 3.3 (a) Zone substation hazard zone

The hazard zone in this case would extend roughly 30 to 40 metres

beyond the edge of the grid.
In cases where a very low resistance is obtained there may be no
hazard zone, despite large fault currents, e.g., a case in point with a
resistance of 0.006 ohms gives 96 volts rise above remote earth for a current
of 16,000 amps.
In areas of high soil resistivity, extensive earthing systems similar
to those referred to above may have quite high resistance to earth and,
consequently, be subject to quite high voltage rises, necessitating measures
to control both the rise of the grid above remote earth and the steepness of
the voltage gradients (particularly those at the edge of the grid).
Any non-homogeneity in the soil will result in distortion of the
uniformity of voltage gradients which would otherwise be expected, e.g.,
for a simple symmetrical grid.
Also more significant distortion of the constant voltage contours
and gradients may occur due to buried conductors, e.g., high voltage
underground cable metal sheaths and/or armouring [refer Fig. (3.3b)].
Earthing
grid 40 × 40 m
FIGURE 3.3 (b) Distortion of hazard zone by UG cable sheath in

contact with earth and grid
Furthermore, other metal circuits such as pipelines, running in close

proximity to earthing systems, will distort the constant voltage contours
and hence gradients.
Where services, e.g., water supply or communication circuits enter
stations, insulated sections, interruption of sheath continuity and
transformer isolation of cable cores, as appropriate, are employed to avoid
transferring the EPR voltages to points outside the hazard zone to minimise
distortion of voltage gradients and to avoid the possibility of substantial
earth fault currents being carried by these circuits.
Where high voltage underground cables are required, distortion of
the voltage gradients can be avoided by using substantial plastic sheathing
and earthing metal sheaths and armouring etc. only at the station end.
Isolated Structure Earths
Commonly encountered resistance values lie in the range 1 to 50
ohms.
The extent of the hazard zone will depend on the fault current
available, the soil resistivity, the earthing system geometry and the safety
criteria used.
For the purpose of illustration we will use the criteria adopted by
ESAA and Telecom Australia in the ‘Earth Potential Rise Code’(1). The
relevant criterion for a distribution system earthing system is an EPR of
430 volts. For a high reliability system an EPR of 1000 volts is acceptable,
without special precautions being taken.
Fig. 3.4 illustrates the case of a conducting distribution or sub-
transmission pole, e.g., a concrete pole, in homogeneous soil.
lf Voltage above remote
earth (EPR)
1m VT
d rl
Hazard zone ERP = for d ® r
2pd
VT (See pS of code)
Voltage gradient
E1000
Surface 1 mm
of VS
I VS earth E430
I
0 r d1000 d430 d
Distance
from structure
(a) Constant voltage contours (b)
22 kV r = 100 Wm
20
d430 volt
15
10
5 22 kV r = 10 Wm
11 kV r = 10 Wm
0 5 10 15 20 kM
Distance from structure
(Fault levels at source
22 kV 500 MVA 11 kV 350 MVA)
(c)
FIGURE 3.4 Variation of extent of hazard zone with soil resistivity and fault
level for metal pole without supplementary earthing system
Note the considerable influence that fault current and resistivity have
on the extent of the hazard zone.
The physical arrangement of the earthing system and the differing
resistivities of the layers of soil, considerably modify the voltage gradients,
the step and tough potentials at the surface and the location and shape of
the 430, 1000 and 1500 volt hazard zone contours (refer Fig. 3.5).
Wood
Transformer
pole
Insulated
earth
lead
Buried
earthing
Earthing system
electrodes
FIGURE 3.5 Pole mounted substation earthing
Control and Mitigation of Earth Potential Rises and Step and Touch
Potentials
Power authorities use safety criteria similar to those used by Telecom
to determine maximum step and touch potentials and earth potential rises.
These are discussed in more detail in the next section of these notes.
It is only within ‘hazard zones’ that step and touch potentials are
significant. The control of these potentials to within acceptable limits is
usually the major objective in the design of earthing systems for HV
installations, since it is frequently economically impossible to keep the
potential rise with respect to remote earth to a safe value. As a consequence,
methods of working on power circuits which run to remote points, are
designed to ensure that personnel are not subjected to transfer potentials
caused by faults on adjacent circuits.
Fig. 3.6(a) illustrates a method of isolating workmen from transfer
and step and touch potentials.
Fig. 3.6(b) illustrates an arrangement used to facilitate work on
communication circuits entering a substation surrounded by a hazard zone.
In Fig. 3.5 the advantage of wood poles in respect to the avoidance
of touch potentials is clearly evident. The insulation of earth lead can also
help in reducing potential across the surface by deeper burial of the
electrodes.
Live
circuit
Circuit
under access
Safety earthing
Insulated arranged to keep
ladder out of service &
pole at same
potential if live
circuit faults to pole
(a) Precaution for power

authority work
Isolating links
Insulated
box Insulating
met station
earth grid
Cable to telecom Cable to substation
(b) Work on telecom cable

entering substation
FIGURE 3.6 Safe working arrangements for work inside a hazard zone
Earthing systems for large substations can and are designed

specifically for each application taking into account site conditions. The
design is often verified by current injection tests after installation of the
earthing system. In the case of distribution earthing systems this practice
is impracticable and standard designs have to be used. The results achieved
can vary widely from expectations. Furthermore it is usually not possible
to get a satisfactory standard for conducting poles.
Methods of reducing EPR’s and step and touch potentials for
conducting structure cases as shown in Fig. 3.2 will involve supplementary
earthing systems and/or overhead earth wires to common the earthing
systems of a number of adjacent structures, with a view to reducing the
resistance to earth and distributing the fault current through a number of
adjacent earthing systems.
Reduction of earth fault current and hence, reduction of potential
rises can be achieved by inserting impedance between the transformer
neutrals and the station earth grid.
Justification for adopting these measures and solutions to some of

the protection problems they introduce, will be apparent after examining
the safety requirements and difficulties in realising them which are
discussed in the next section of these notes.
3.4 SAFETY CONSIDERATIONS

3.4.1 Nature of Hazards Associated with High Voltage
Power System Earth Faults
The Step, Touch, Transfer and Induced Voltages associated with HV earth
faults very seldom last longer than a few seconds. Earth faults involving
high current, i.e., those involving significant step, touch, transfer and/or
induced voltages, are usually cleared in less than one second.
Since these voltages are by nature transient, the prevention of
electrocution is concerned with avoiding ventricular fibrillation rather than
let-go currents, as is the case with low voltage circuits, (where the voltage
frequently remains applied to the victim and asphyxiation will most likely
be the cause of death).
Consequently the currents associated with HV faults which can flow
through the human body, without threatening life, are considerably higher
than the let-go current. The current allowable in any case is a function of:
the duration of the shock; condition of application (arm to arm, arm to leg
etc.); body weight; health factors.
3.4.2 Assessment of Risks
The relationship between the available voltage and the resultant current
will be a function of: body resistance; contact resistance and voltage [body
resistance is a function of voltage and is approximately 1000 ohms at 250
volts (2)]. Contact resistance will vary widely with the type of contact and
factors such as perspiration.
When prospective EPR’s and induced voltages are determined for
typical existing earthing systems associated with HV distribution (e.g.,
substation earthing systems), it is frequently found that they exceed those
required to cause fibrillation, yet there is virtually no historical evidence
of death or injury from this reason. The major cause of this discrepancy is
the low probability of somebody being present when a fault occurs.
If conducting poles eventually replace all wood poles then the
number of earthing systems will have increased about ten-fold, changing
the probability dramatically.
Power system engineers have opted to make this type of installation
inherently safer. Various national and international standards, codes and
technical papers are available to enable policies and practices to be
determined. Some of these are listed in the references. The graphs as shown
depict allowable touch voltage versus time as specified by some of these

standards and codes (refer figure given below).
10
4
(1) rs = 10000[W Metre]
(2) rs = 10000
(3) rs = 5000
(4) rs = 5000
(5) rs = 1000
(6) rs = 1000
(7) rs = 100
(8) rs = 10
(9) rs = 100
(10) rs = 10
103
1
Touch voltage (Volt)
2
3
4
IEEE STD 80–1961

5
IEEE STD 80–1976

102
6
7
8
9
10
101 –2 –1 0 1 2
10 10 10 10 10
Time (Seconds)
FIGURE 3.7
3.5 APPLICATION OF SAFETY CRITERIA

3.5.1 Control of Touch Potentials
By way of example let us use AS2007 and see what the implications are in
relation to commonly encountered fault levels and protection operating
times. A clearance time of 0.1 sec. (the fastest practical time for
subtransmission and distribution) would allow 300 volts max. for touch
potentials; a clearance time of 1 sec. would allow 120 volts max.
Consider an earth fault on a conducting pole in a 22 kV feeder 1 km
from a zone substation with a max. fault level of 500 MVA. Assume the
pole has a resistance to earth of 10 ohms. The fault current will be
approximately 1200 amps. and the touch potential will be about 7 kV.
If the pole was 10 km from the zone substation the fault current
would be about 800 amps. and the touch voltage about 5 kV.
Reduction of earthing resistance to say 1 ohm would result in
voltages of about 3 kV and 800 volts respectively.
A resistance of 1 ohm is impractically low in most cases for a pole
earthing system.
Commoning all pole and LV earths will reduce the touch voltages
to about 800 and 200 volts respectively.
Reducing the fault current by installing a resistor between the
transformer neutrals and earth, to limit the maximum fault current to say
2000 amps, together with CMEN (common multiple earthed neutral), will
reduce the prospective voltages to about 150 and 100 volts respectively,
i.e., within the limits previously postulated, when typical clearance times
apply.
The comparable voltages for 11 kV, and 6.6 kV systems for faults
distant 1km stations with maximum fault levels of 350 and 250 MVA
respectively and no CMEN or EFCL (earth fault current limitation) are of
the order of 2 kV and 1.5 kV respectively. Still a problem.
Subtransmission may also use similar poles and the maximum fault
currents are often high so solutions have to be found. In some respects the
difficulties are greater and CMEN and EFCL may not provide a solution.
In the case of main transmission other factors enable economically
practical solutions to be obtained e.g., specifically designed tower earthing
systems and ground wires are required for other reasons; the lines are run
in easements; faults are cleared rapidly; faults are rare.
3.5.2 Control of Step Potentials
These will usually be the major concern only where touch potentials have
been eliminated by appropriate insulation.
3.5.3 Control of Transfer Potentials
Communication circuits and other metallic services entering or passing
through hazard zones are prevented from transporting potentials out of
the hazard zone by suitable isolation and insulation, as mentioned earlier.
Where the means of transport could be the neutral of a CMEN system
the transfer potential may need to be kept within the voltage limits for
touch potentials.
3.6 DEMANDS ON PROTECTION ARISING FROM SAFETY,

RELIABILITY AND INTERFERENCE CONSIDERATIONS
3.6.1 General Comments
Sensitivity, reliability and security are the characteristics required of the
protection systems installed on a power system, if the inherent reliability
of the elements of the power system, is to be realised, as a power system
operating safely, to provide a high degree of continuity of supply to its
customer.
Continuity implies more than simply remaining connected, it also

implies maintenance of voltage and frequency and indirectly selectability
and discrimination by the protection systems, together with security against
mal-operation, speed of operation and reliability.
Sensitivity and reliability are the main attributes required for safety.
Protection equipment and systems generally exhibit a high degree
of reliability: Failures occur on only a few per cent of occasions when
protection is required to operate. However, this is not good enough since
the consequences of failure to operate for many fault conditions would be
catastrophic. It is, therefore, necessary to provide back up.
Generally speaking it is more difficult to provide back up than
primary protection.
Measures required for safety such as EFCL add greatly to the
difficulties and cost of providing protection. If EFCL had to be added to an
existing large system the total cost would most likely run to hundreds of
millions of dollars. Some illustration of this is contained in the following
paragraphs.
3.6.2 High Voltage Distribution Systems
There are always conflict of interest between the requirements phase fault
sensitivity and loading and between earth fault sensitivity and out of
balances in the system. For example, on a long radial feeder there may be
heavy load near the start which will effectively limit the sensitivity of the
phase fault protection and make it difficult to see faults at the remote end
of the feeder. Also, if there were fuses in the backbone of the feeder with
long lines beyond them, earth fault sensitivity will be restricted (when 1
fuse is blown there will be an unbalance of capacitance in the circuit due to
the missing phase conductor and consequently zero sequence current in
the earth fault relay).
In both cases safety and welfare of the feeder can only be assured, if
consideration of protection requirements are an integral part of design
process for establishing the line and operating it as part of a system where
load transfers and alternative feeds apply.
E/F sensitivities, in the range 5–10 amps primary, can be achieved
easily, for both primary and back up protection. Greater sensitivity can be
achieved with short feeders and for longer or more complex feeders by
careful attention to distribution design.
The addition of resistors to limit the maximum earth fault current has
an insignificant effect on feeder protection design, however, it considerably
increases the difficulty of protecting the plant within the zone substation.
3.6.3 Zone Substation and Sub-transmission Line Protection
Differential, distance and other protection schemes for zone substations
and sub-transmission are dealt with comprehensively in other lectures and
all that is intended here is to discuss aspects of the philosophy of

application.
Figure 3.8 depicts a zone substation switching arrangement intended
to ensure reliability (or continuity) of supply to the feeders despite the loss
of either an incoming line or a transformer. Obviously separate protection
is required on each element (separately switched portion) of the system to
realise the inherent reliability of the system.
Shorting switch
(Fault make capability)
Isolation of cap
bank neut
FDR FDR
(a) Arrangement of NER’s in a fully switched sub
with line CBs and no pilot cable system
Master
& BU E/F
X & Y Diff.
Res
E/F
(b) Application of restricted earth fault protection (a differential protection

covering earth faults on the star connected winding and its connection)
FIGURE 3.8
The reliability of supply to the feeders is jeopardised by protection

equipment or circuit breaker failure or inadequacy to cope with all the
prospective fault conditions.
The design process must begin with a thorough examination of all
the prospective fault conditions for normal and abnormal system
conditions, taking into account conditions applying during the sequences
of fault clearance, failures of relaying equipment and circuit breakers,
system stability requirements (if applicable), short and long term ratings
of lines and plant (protection to be stable for all loads up to line or plant
ratings or maximum expected load and in some situations responsive to
overloads).
Figure 3.9 depicts a situation where the primary protection
(differential and gas) has some local back up relaying (overcurrent) but
would use distance protection at the source station to cover failure of the
line CB, since it would usually be possible to detect most faults on the low
side of the transformer under radial feed conditions from the remote station.
Diff
O/C
20 MVA
Gas
MEF
1f – G
Fault
FIGURE 3.9 Typical pilot for fully SW subs
The addition of a resistor to the transformer neutral circuits, as

depicted in Fig. 3.10 eliminates overcurrent and the remote distance
protection as satisfactory back up necessitating duplication of the
differential protection and supplementation with restricted earth fault and
back up earth fault. (Note: for the case of a 66/22 kV 20 MVA transformer
with an NER limiting earth fault current to 2000 amps the resulting current
on the 66 kV side is approximately 400 amps, but the maximum load current
could be as high as 350 amps with forced cooling i.e., no margin to set an
overcurrent relay). Circuit breaker failure protection would have to be
achieved by signalling to the remote end. An alternative to signalling may
be to use a shorting switch on the resistor.
Dup diff.
20 MVA
66/22
Gas
MEF BU RCS
1f – G EF EF
Dup diff.
REF
FIGURE 3.10 Likely arrangement after inclusion of NERs
It is possible that the arrangement depicted in Fig. 3.9 may be

protected by a protection scheme similar to that of Fig. 3.10 if high speed
back up was required, say, because of system stability requirements. This
would probably be the case with most main transmission stations.
Note: Local back up coupled with back up for CB failure will achieve
the minimum disconnection of healthy plant or lines when failures occur
and will do so at a faster speed than remote back up. Thus it will contribute
to greater reliability of supply.
3.6.4 Zone Substation Feeder Protection
Fault Current Paths
The fault current paths for earth faults fed from zone substations
are depicted in Figs. 3.11, 3.12 and 3.13 for various transformer winding
connections and neutral earthing arrangements.
I66
I22
max 5 KA
FIGURE 3.11 Fault currents obtained with NER for a 66/22 kV 20 MVA transformer
I22
I66 =
3n
max abt 400°C

n = Tr ratio
I22
Max
NB. 20 MVA TRS may have

short time load ratings up to
40 MVA i.e. 350
FIGURE 3.12 Fault currents obtained without NER for

a 66/22 kV 20 MVA transformer
Note: For 66/11 kV 20 MVA transformers currents on 66 kV side for NER
case are half those for 22 kV case)
1
I ²
2 F
1
I ²
2 F
IF²
IF IF
IF¢
IF¢
IF
(a) Single phase to earth fault fed through a star/delta/star

transformer with unearthed hv star winding
(Note: No zero sequence currents on supply side of transformer
No scope for back up for light earth by monitoring currents on high
side of transformers No LFI on high side of transformers as a
consequence of low side faults EPR of station earth grid).
Si0IF¢²
IA
IB
IC
IF IF
IF²
IF¢² IF
(b) Single phase to earth fault fed through a star/delta/star

transformer with hv star winding earthed
(Note: No zero sequence currents on high side for low side fault
May be a need for co-ordination of E/F prots on high and low sides
LFI and EPRs may need to be considered).
Ic¢
R
Ib¢ Ib
Y
Ic¢ Ia
B
IE
IE
(c) Two phase to earth fault fed through a delta/star transformer
(Note: No zero sequence currents on high side for low side fault.
Max current not as high as for single phase to earth fault.
No LF1 on high side.)
FIGURE 3.13
Except for the case of an earthed star connection on the input winding
of the transformers, currents flowing in the earth are confined to the low
side of the transformer.
The most common arrangements for zone substations and HV
customer substations fed from sub-transmission involve unearthed HV
windings.
Note: 1. Main transmission systems usually have earthed input and
output star connected windings consequently ‘reflected’ or
compensating currents flow in the earth for other parts of the
system in response to a fault on element of system, e.g., in the
system depicted in Fig. 3.13(c) compensating currents will flow
in both the sub-transmission and main transmission systems
(albeit of low magnitude compared with fault current).
2. Distribution substations have unearthed HV windings and
customer generators and SWER systems are isolated by
transformers with unearthed supply side HV windings.
Fig. 3.14 depicts a typical HV distribution system arrangement.
Protection Implications of System Loading and Earthing

Arrangements
The essential features of these system arrangements which impinge
on safety, reliability and interference considerations are:
(a) Loads on HV distribution systems are effectively connected
between the phases and consequently current flows in the earth
only when there is an earth fault, hence earth fault protection
sensitivity is limited only by: (i) small earth current caused largely
by small out of balance of line to ground capacitance and (ii) CT

errors.
(b) Under normal system conditionals there is only one point at
which the system is earthed, hence there can be no currents in
the earth arising as a consequence of source voltage imbalances
and thus sensitive earth fault protection can be applied.
This would not be the case if there were two or more points at which
transformer neutrals were earthed.
Note: Where two sources are paralleled to transfer a feeder from one zone
substation to another it will often be necessary to suppress the
sensitive earth fault protection whilst the sources are in parallel.
In cases where two feeders from the same source are paralleled to
effect transfers of load master control of the feeder earth fault
protections can be used to avoid the need to suppress protection
during the switching operations. Suppressing protection during
switching operations is an undesirable procedure since faults can
be precipitated by operating equipment e.g., an insulator may break
when an isolating switch is operated.
(c) Since earth currents will in most cases be seen as such only on
the output side of zone substation transformers (except where
there are earthed star windings on the input windings of the
transformers) back up for failure of the protection must be done
locally and the broader implications in respect to the
independence of the primary and back up protections must be
considered if the success rate in the detection and clearance of
faults is to be maximised.
These latter remarks also apply to faults not involving earth and to
high current earth faults, however in these cases remote back up can often
be achieved and complete independence of the detection and clearance
equipment is assured (relays, CTs, auxiliary supplies and circuit breakers).
In the arrangement of sensitive primary the (master earth fault
scheme) and back up earth fault protection depicted in Fig. 3.13 the
independence of the fault detection functions of the primary and back up
protections are obvious. Independence of the fault clearance arrangements
is achieved by using the bus tie and transformer CBs; for back up for CB
failure and using independent DC supplies for powering primary and back
up relaying and CB tripping.
DC supply for
parallel prot
Master DC supply for BU

E/F
FDR BU
E/F E/F
Master BU TP FDR
E/F E/F TP TR + B/T
CBs
FDR o/c
E/F
Zone
substation
HV customer (may include generators)

substation
LV Distr.
LV customer
IL¢
SWER LV customer
ISOL,
substation IL SWER
LV Distr.
substation
FIGURE 3.14 Typical arrangements for 3-phase hv feeders showing

(i) hv feeder system earthed at one point only and all loading
connected phase to phase (ii) primary and backup E/F prot.
Calculation and Analysis of Earth Currents

Calculation of prospective fault currents using symmetrical
component analysis is dealt with in another section however to reinforce
the description of current flows given in, Section on ‘Fault Current Paths’
described earlier, Fig. 3.15 depicts graphically the symmetrical components
and the real currents for a single phase to ground fault fed through a
delta/star transformer.
Z1s Equiv Z1T Z1F

Z2s imp of Z2T Z2F
system Z0T Z0F
c¢ c
b¢ If b
lb¢ IF
a¢ a
la¢
n
iaa iba ica IF
ia¢ ia¢
1 2
RF
ic¢ ib ¢ ib¢ i¢
1
1 2 c2
ic ia ia 2 ia ib ic
1¢
ia¢ ib¢ 1 0 00
2 1
ia¢ ib¢ ic
2¢ ib1 ib2
ia¢ ib¢ ic¢ = 0 ic1 ic2
2 2
ib 1 ic 1
Ean ib
ic 1
ia = ia2 = ia0 = ic
0
1 Z1 + Z2 + Z0 + 3RF ib 2 ib 1 Ia
ic 2
IF = Ia = IN = ia + ia + ia0 ia 1
1 2
IB = 0 IC = 0
Z1 = Z1s + Z1T + Z1F
Z2 = Z2s + Z2T + Z2F
Z0 = Z0T + Z0F
FIGURE 3.15 Single phase to earth fault for fed through a delta/star transformer
Problems Associated with High Impedance in the Path of the Fault

Current
Not all faults can be detected because in some instances the
magnitude of the fault current will be less than spurious currents arising
out of system out of balances.
One type of example is a failed insulator on a wooden pole which
may result in a pole fire before there is sufficient current to cause protection
operation.
Another more serious problem arises when a fallen conductor is
back fed from a lightly loaded distribution substation. A case involving a
single phase transformer is illustrated in Fig. 3.16 where it can be seen that
for even a fully loaded 10 kVA transformer the fault current is limited to
about 60% of the full load current of the transformer i.e., about 0.3 amps
for a 22 kV case. Thus to get current above a minimum operating current
of 5 amps for the earth fault protection requires a fairly sizeable load.
IF
IF = 0.6IL IF
EF – E Load
IL = Normal load
current
IF
IF
FIGURE 3.16 Conductor on earth back fed from single phase substation
Fortunately this case has a low probability but it does illustrate the
desirability of arranging the system to minimise spurious currents and
hence be able to use low protection settings. One of the measures which
could be taken is to avoid long sections of fuse protected line: the blowing
of one fuse will unbalance the phase to ground capacitances and increase
the spurious earth currents.
3.6.5 Main Transmission Systems
There is obviously more need and justification for maximising the reliability,
security (stability) and selectability of protection systems in this type of
application.
Duplicated protection with back up for CB failure together with
various enhancements to achieve faster clearance are essential.
Automatic monitoring and self checking arrangements are quite
common.
3.7 INTERFERENCE ON SUPPLIED AND OTHER SYSTEMS

3.7.1 Preamble
There are numerous ways in which electrical interference can occur between
electric power systems and other systems and between parts of the power
system. Examples of interference with other systems are: Transferred and/
or induced 50 Hz voltages during earth faults; injections of high or low
voltage due to direct contact between the systems; TV and radio interference
due to faulty line hardware; induction into communication circuits of
50 Hz fundamental and harmonic voltages from load currents in the power
system. Examples of interferences within the power system are: Voltage
distortion due to fluctuating loads (e.g., arc furnaces, traction); induction
into in service and out of service circuits during faults on adjacent circuits;
EPR's in the vicinity of out of service lines, etc. which are under access for
work; interference by low and high frequency noise signals generated by
faults and their associated arcs, in the power system's protection and control
circuits.
As mentioned earlier the protection engineer is usually in a good
position to deal with these problems because of the type of studies he must
do, to set and examine the performance of his equipment. Indeed it is
absolutely vital, in the interests of reliability of supply, that the protection
and control systems are immunised against interferences.
Some of these interferences are examined in the following
paragraphs.
3.7.2 Induced Voltages and EPR’s Imposed on Other Systems

Induction in Telecommunication Circuits
Fig. 3.17 depicts the electromagnetic coupling which exists during
earth faults on a power line. The induced voltage is given by the formula
Power line
Alternating magnetic
field
Induced current
Telecommunication
line Earth
Induced voltage fault
Low frequency electromagnetic induction

(a)
Induced
longitudinal
voltage
E
0
Length of exposure l
Induced longitudinal
voltage in an open-circuit
telecommunication line
(b)
Induced
longitudinal + 1
voltage E
2
relative 1 0
E Length of
to earth 2 – exposure l
(c)
FIGURE 3.17 Induced longitudinal voltage between a
telecommuinication conductor and earth
E = C × I × L × K where, C = 50 Hz mutual impedance per unit length of

exposure; I = fault current; L = length of exposure; K = shielding factor. C
is a function of the separation and the soil resistivity and a 10 metre
separation, C = 0.28 ohms/km; for a 50 metre separation, C = 0.05 ohms.
Telecommunication lines often run parallel to power lines for
considerable distances and fault currents can be very large, hence induced
voltages can be high even with large separation, e.g., for L = 10 km, I = 10,000
amps, K = 1, separation = 0.5 km, E = 2.5 kV. If the separation was only
two or three metres, as may apply where an aerial supervisory cable is
strung under the line, the induced voltage would be 17.5 kV.
EPR’s in Telecommunication Circuits

Telecommunication cables entering or passing through a gradient
area will encounter EPR’s. These EPR's will be significant in respect to
equipment and personnel safety, only if a hazard zone is entered. Refer to
Figs. 3.17 and 3.18.
Hazard zone
Earthing
system
PIT
or p Cable
illar
Exchange
During fault voltage
at this point exceeds
alloable voltage
FIGURE 3.18 Hazard zone
Pipelines
Oil and gas pipelines are often insulated from earth for electrolysis
reasons and thus they need to be treated similarly to telecommunication
cables in respect to EPR’s and induced voltages.
3.7.3 Induction into Adjacent Lines
Voltages induced into adjacent lines are usually of little consequence, except
in a few circumstances, notably where an out of service circuit is being
worked on and where sensitive earth fault protection is applied to a line in
a network which shares poles with other lines, (which perhaps may be
operating at a different voltage level e.g., distribution HV often shares poles
with sub-transmission).
The diagram of Fig. 3.19 helps to explain how sensitive earth fault
protection, on lines in a network, can mal-operate during a fault on a feeder
running parallel to one of the lines in the network, viz. in each phase of the
network continuous paths can be traced, one leg of which is in close
proximity to the faulted feeder. Currents resulting from the voltage induced
in each of the phases will be summed in the earth fault protection circuits
associated with the network. If the exposure is long enough and the sub-
transmission protection is sensitive enough the relays will respond and
may operate before the feeder protection has been able to clear the fault.
Obtaining discrimination is usually fairly easy in this case, however in
cases where different networks are involved it is difficult to achieve
discrimination and retain sensitivity.
B
A Ei E/F C
OK
Zm
Earth
fault
Distribution feeder sharing poles with subtransmission
(a)
A
e
h as e B
P a s C
e
Ph as
P h Ii
C
Ii
A
I i I i Ii
Ii
3Ii
Induced current are summed into E/F pilot at both ends of each line
(b)
FIGURE 3.19 Induced currents interfering with protection on another pilot of system
(Note: Problem only arises where exposure is long-earth fault too low for quick
operation for feeder protection-sensitive earth fault protection on loop).
3.7.4 Impulses and High Frequency Disturbances

Protection and control equipment is often situated in very hostile
electromagnetic and electrostatic environments especially when faults occur
in close proximity to relay houses. It must not be damaged nor mal-function.
Electromechanical equipments used in the past and still in common
use exhibit considerable resistibility to damage by quite high transient
voltages and are insensitive to high frequency disturbances. Semi-
conductor equipments now used need to be designed and housed to resist
and avoid severe impulses and a wide band of electrical noise, if their
performance is to allow the intrinsic reliability of the primary system
elements to be fully utilised. Specifications for impulse and high frequency
tests are contained in standard specifications such as IEC Publication 255-4,
Appendix E. I.8.
CHAPTER
Relaying Transducers
4.1 INTRODUCTION
In power systems, the levels of voltage or current are usually too high to
permit direct connection to measuring instruments or relays, coupling is
therefore made through voltage and current transformers that are designed
to reproduce an accurate scaled down replica of the input quantity.
Protective relays are required to measure during the transient period (which
invariably follows a fault) so that both accuracy and transient response of
the transducers are important. Errors in transformer output may
abnormally delay the operation of the protection, or cause unnecessary
operations. The functioning of such transformers must, therefore, be
examined analytically.
4.2 VOLTAGE TRANSFORMERS

The equivalent circuit of a wound electromagnetic transformer is of the
well-known form as shown in Fig. 4.1. The system voltage Vp is connected
across the primary winding. The voltage Vs developed across the burden
(relays of meters connected to the secondary winding) is required to be
independent of Zb over the specified range. One way of achieving this is to
make the winding impedances Zp and Zs as low as possible. Secondly, the
nominal core flux density is arranged to be well below the saturation level.
RELAYING TRANSDUCERS 81
n:1
ZP ZS
Vp Ze Zb VS
(Perfect coupling assumed)
C1 = High voltage capacitance

C2 = Low voltage capacitance
T = Intermediate transformer
FIGURE 4.1 Equivalent circuit of a transformer
Errors: Two errors are important. Firstly, the ratio error which is
defined as (n Vs – Vp)/Vp × 100%, where n is the turns ratio primary/
secondary. If the error is positive the secondary voltage exceeds the nominal
value. The turns ratio need not be equal to a whole number, some turns
compensation often being applied to ensure that error is positive for low
burdens and negative for high burdens. The phase error is the phase
difference between the primary and secondary voltages. Permissible errors
are +/– 3% (ratio) and +/– 2% (phase).
All voltage transformers are required by standards to have ratio and
phase angle errors within prescribed limits over a 80% to 120% range of
voltage and a range of burden from 25% to 100%. For protection purposes,
accuracy of measurement may be important during fault conditions when
the voltage is greatly suppressed. Therefore, a voltage transformer for
protection must meet the extended range of requirements over a range of
5% to 80% rated voltage and for certain applications, between 120% and
190% rated voltage.
4.2.1 Transient Performance of a Voltage Transformer
Transient errors cause few difficulties in the use of electromagnetic VTs
although some do occur. If a voltage is suddenly applied, an inrush transient
will occur, as with ordinary power transformers. The effect is however
less severe than the power transformers because of the lower flux density
for which the VT is designed. When the supply to a VT is interrupted (e.g.,
primary open circuited) the core flux will not immediately collapse and a
more or less exponential current flows in the burden. The condition must
not be confused with collapse of primary voltage due for example to a
system fault when the secondary voltage collapses also.
4.2.2 The Capacitor Voltage Transformer (CVT)
At voltages above 132 kV, the cost of conventional VTs is often prohibitive
because for higher voltages, the size becomes largely proportional to the
rated voltage. The CVT is a useful and more economical alternative, the
arrangement being as shown in Fig. 4.2.
C1 L T
C2 Zb
FIGURE 4.2 Capacitor voltage transformer arrangement
C = C1 + C2 L
C1
Vi = Vp Zb
C1 + C2
FIGURE 4.3 Thevenin’s equivalent circuit of a capacitor voltage transformer
This device is basically a capacitor potential divider, the low-voltage

end of which energies a wound voltage transformer. There are several
variations of this basic circuit. The inductance L may be a separate unit or
it may be incorporated in the form of leakage reactance in the intermediate
transformer. A Thevenin equivalent circuit is shown in Fig. 4.3 and it can
be seen that at nominal system frequency when C and L are in resonance
and therefore cancel, the circuit behaves in a manner similar to a
conventional electromagnetic VT. At other frequencies, however, a reactive
component exists which modifies the errors. Provided that the reactive
voltages across C and L are not too large in relation to Vi, the change of
error with frequency is not excessive. For a typical design, in which C has
the value 2000 pF and Vi = 12 kV, the change in phase error, with 150 VA
UPF output per Hz of frequency change (in the region of rated frequency)
is about 15/20/40 minutes for a 400/275/132 kV capacitor voltage
transformer.
Transient Behaviour of CVTs
The transient response of CVTs is much inferior to that of a electromagnetic
VT and there are errors influenced by so many factors. A CVT as can be
seen is a series resonant circuit. For example, under conditions of sudden
voltage change (such as a collapse due to a fault) on the primary, oscillations
caused by the interaction of the series capacitance, and the magnetising
inductance of the intermediate transformer (of typically 10 kHz) will often

occur. Oscillations at around 300 Hz will also occur under similar conditions
due to the interaction of the series tuning reactance and the winding
capacitance of the intermediate transformer. The exact equivalent circuit
of a CVT is shown in Fig. 4.4.
C L RP Rs Ls
C1 RM Lm
Vi = V p Cm Vx
C1 + C2
FIGURE 4.4 Simplified equivalent circuit of a capacitor voltage transformer
It should be noted that the power factor of the burden greatly

influences transient responses. It can dominate the CVT behaviour, allowing
the same CVT with two different burdens to exhibit different transients. In
general, an increase in resistive burden results in larger-amplitude
transients but faster decay. On the other hand, a reactive burden constitutes
a further tuned circuit and introduces new modes of oscillations that can
persist for several cycles of the system frequency. It is generally found that
higher power factor burdens are better.
Ferro-resonance
It is possible when energising CVTs to induce sustained sub-
harmonic ferro-resonance that takes place between the inductance of the
exciting impedance and the capacitance of the potential divider. One quite
common condition occurs in which a stable one-third sub-harmonic
persists. A resistive burden tends to reduce the likelihood of such
oscillations and it is common to build in special anti-ferro resonance devices
that use parallel tuned circuits. Some compromise is however necessary
because the latter tends to impair transient response.
The High Fidelity CVT
It is now widely appreciated the CVTs
cannot respond quickly to changes in primary C1
voltage. A capacitor divider on the other hand Buffer
(as shown in Fig. 4.5), overcomes the transient- amplifier
response difficulties of CVTs by interposing a C2
buffer amplifier between the divider and its

burden. This preserves the divider’s
homogeneous nature and allows it to respond FIGURE 4.5 Capacitor
accurately over a wide range of frequencies. divider buffer amplifier
Ideally one would want new purpose-built units of such CVTs but this is
not economically viable at present.
A closer approach to the response of a capacitor divider can be
achieved by modifying a conventional CVT to produce a capacitor divider
output in addition to the traditional output from the intermediate voltage
transformer. Such a scheme is shown in Fig. 4.6.
Ic
VC1 C1
L
VC2 C2 B Vs
Vp
Power
Pre-amplifier
VC3 C3 B
Matching impedance Power amplifier
FIGURE 4.6 A hi fi capacitor voltage transformer
The capacitor C3 is introduced into the CVT circuit so that it conducts

the whole of the capacitor current ic and therefore reflects the voltage VC .
1
The value of C3 is large compared with C1 and C2 and does not affect the
transient or steady-state characteristics of the traditional output voltage
Vs in any way. With this arrangement, the less accurate output Vs of the
conventional CVT can be used for purposes of metering, telemetry, back
up protection, etc., and the very accurate output (VA) of the capacitor
divider used for the high speed distance relays.
4.2.3 Other Voltage Transducers

A high voltage resistor divider is an attractive alternative from a transient
point of view, but is prohibitively expensive and produces a ratio error
which is burden dependent. Another more practical alternative is a cascade
voltage transformer which has a better frequency and transient response
than the conventional CVT. The principle is illustrated in Fig. 4.7. The
conventional type of VT has a single primary winding, the insulation which
presents a great problem for voltages above 132 kV. The cascade VT avoids
these difficulties by breaking down the primary voltage into several distinct
and separate stages. The complete VT is made up of several individual
transformers, the primary windings of which are connected in series. Each
magnetic core has primary windings P on two opposite sides and the
secondary winding S consists of a single winding on the last stage only.
Coupling windings C connected in pairs between stages, provide low
impedance circuits for the transfer load ampere-turns between stages and
ensure equal voltage distribution over the several primary windings. The
primary and coupling windings are connected to the cores of selected points
and all stages are housed in a vertical stack which is filled with oil and
sealed together with a cushion of nitrogen to permit expansion during
temperature changes. These cascaded devices are presently very expensive,
but in view of their superior transient response, they are likely to find
more widespread use in the future.
A
S n
N a
FIGURE 4.7 Schematic diagram of a typical cascade voltage transformer
4.3 CURRENT TRANSFORMERS (CTs)

Current transformers convert the primary current to a level suitable for
measurement by the protective relays, usually with a rated secondary
current of 5 or 1 amp.
The primary winding of a CT is connected in series with the power
circuit. Its equivalent circuit is the same as for any other transformer (as
shown in Fig. 4.8). The object is:
(a) minimum interference with power circuit i.e., Zp and Zs are made
very small by design.
(b) that Is is a faithful replica of Ip.
Ip Zp Ip¢ I:n Zs
Is = Ip¢/n
Rm Xm Es Vs Zb
FIGURE 4.8 Equivalent circuit of a current transformer
4.3.1 Construction of CT
Current transformers are usually designed so that the primary winding is
the line conductor that is passed through an iron ring which carries the
secondary winding. They are mostly of this type and are known as bar-
primary or ring-wound current transformers.
The construction of a typical ring-wound current transformer is
shown in Fig. 4.9. Grain-oriented sheet-steel strip is wound to form a core
and is covered with a layer of insulation. The secondary winding is wound
over this and consists of the number of turns needed to produce the required
ratio of wire of sufficient cross-sectional area to carry rated current, followed
by a further layer of insulation which covers the secondary winding. When
installed, the primary conductor which acts as a single turn, is passed
through the centre of the ring. The making of the core by stacked annular
laminations has now been superseded by the wound-type method.
Core construction
P1
Insulation
Secondary
winding
Tape
S2 wound Insulator
S1
core
FIGURE 4.9 Construction of a current transformer

4.3.2 Design of CT
For protection purposes, current transformer specifications are normally
laid out in terms of the knee-point e.m.f. The knee point of the excitation
curve is defined as that point of which a further increase of 10% of secondary
emf would require an increment of exciting current of 50% (Fig. 4.10). We
can say that at voltages above knee point, saturation occurs quickly.
Vk 10% Vk
Exciting voltage (Vk)
50% Irk
Irk
Exciting current (Ir)
FIGURE 4.10 Definition of knee point of excitation curve

Magnetising current, and consequently flux, changes from zero to
maximum in 1/4 cycle and therefore the rate of change of flux is
φ−0
= 4φ webers/cycle
1/4
or at a frequency of ‘f ’ cycles/s = 4φ f webers giving an average induced
voltage Vav of 4φ fN, where N is the number of turns or in r.m.s. values, the
knee-point voltage is
V = 4.44φf N as V = 1.11Vav
Also as flux φ = flux density which is B(tesla) × core area × s (m2),
the knee-point voltage is
V = 4.44Bsf N
The flux density of electrical sheet steel is about 1.5 tesla at knee-point
which for a ring-type current transformer of known ratio makes the knee-
point voltage fairly easy to estimate if the approximate dimension of the
core is known. For example, a CT ratio of 200/1 with a core area of
30 × 20 mm2 would have a knee point flux of
1.5 × 30 × 20 × 10–6 = 0.0009 weber
which on a 50 Hz system would produce a knee point voltage of

V = 4.44 × 0.0009 × 200 × 50 = 40 V.
4.3.3 Burden of CT
The load connected to a current transformer is called the burden and can
be expressed either as a VA load or as an impedance. In the former case the
VA is taken to be at the CT nominal secondary current. For example, a
5 VA burden on a 1 A transformer would have an impedance of 5 ohms:
5 VA
=5V
1A
5V
impedance = =5Ω
1A
or on a 5 A current transformer:
5 VA
=1V
5A
1V
impedance = = 0.2 Ω
5A
All burdens are connected in series and the increase in impedance
increases the burden on the current transformer. A current transformer is
unloaded if the secondary winding is short-circuited as under this condition
the VA burden is zero because the voltage is zero. The errors of
transformation depend on the angle of the burden as well as its impedance.
4.3.4 Errors of CT
Besides the ratio and phase errors, composite errors are to be considered.
The composite error is defined as the r.m.s. value of the difference between
the ideal secondary current and the actual secondary current (included
are the effects of phase displacement and harmonics of the exciting current).
In CTs with negligible leakage flux and no turn correction i.e., with turns
ratio equal to the nominal current ratio, the composite error corresponds
to the r.m.s. value of the exciting current (usually expressed as a percentage
of the primary current).
CTs are therefore designed to give as high a value as possible of
magnetising impedance. This involves high quality core material. The more
closely the value Ie tend to zero the greater is the accuracy, but it is not
possible to eliminate both ratio and phase errors. The phasor diagram
(as shown in Fig. 4.11) illustrates this point. It can be clearly seen that the
exciting current Ie is responsible for all the errors. The component Ir of Ie,
which is cophasal with Is causes the ratio error and the component Iq which
is in quadrature with Is causes the phase error θ. The problem is that Ie is

also dependent on the magnitude of the p.f. of the burden. It will be seen
that with a moderately inductive burden, with Is and Ie approximately in
phase, there will be little phase error and the exciting component will result
almost entirely in ratio error. It is possible to partially compensate for ratio
error by reducing the secondary winding by one or two turns.
IsRs
Es
IsXs Iq
Ir
Ip
Vs
q
Is
Ie
F
Es = Secondary induced e.m.f. IsRs = Secondary resistance voltage drop

Vs = Secondary output voltage IsXs = Secondary reactance voltage drop
Ip = Primary current Ie = Exciting current
Is = Secondary current Ir = Component of Ic in phase with Is
q = Phase angle error Iq = Component of Ic in quadrature with Is
FIGURE 4.11 Vector diagram for a current transformer
On assuming 1/1 turns ratio Is = Ip – Ie. The exciting current depends

on induced emf Es which is approximately equal to Is(Zs + Zb) and also in
exciting impedance Ze. If Ze is linear, then Ie would be the composite error.
In practice, Ze is non-linear and if Ie contain harmonics of the fundamental
frequency which increases its r.m.s. value and hence the composite error
increases too. This is quite noticeable in the saturation requirement.
4.3.5 Specifications of Current Transformers
Australian Standard AS 1675–1986 provides standards for the specification
of CT’s.
For protection schemes where the operating times are greater than
around 0.2 second i.e., there are no significant transients or dc component
present in the fault current, a Class P CT is used, specified in the format:
10 P 150 F20
10 — represents the composite error, % at the accuracy limit current
(preferred values 2.5, 5, 10),
P — protection class,
50 — rated secondary reference voltage; secondary voltage required

to drive secondary current through the connected burden at
the accuracy limit current,
F20 — accuracy limit factor, either 5, 10, 15, 20, or 30 times rated
secondary current (20 preferred).
For protection schemes which operate in less than 0.2 second and
where it is necessary for the CT to perform in the face of the DC transient
present in the fault current, a PL Class CT is used. This Class is specified in
terms of the knee point voltage, in the format:
0.05 PL 950 R3
0.05 — represents the magnetising current (amps) at the knee point
voltage of the CT excitation curve,
PL — class,
950 — chosen knee point voltage,
R3 — maximum secondary winding resistance (ohms).
The knee point of the magnetisation curve (Vk) is defined as that
point at which an increase of 10% of secondary voltage would increase the
magnetising current by 50%.
For the PL Class CT Vk is chosen to provide a margin so that the CT
will not saturate during the primary system DC transient, by allowing for
the primary transient factor (TF).
Vk = TF × I(Rs + Rb)
Rs—CT secondary resistance (ohms)
Rb—Resistance of connected relays plus leads (ohms).
British Standard BS 3938 1973 is a little different in that burden, accuracy
class and accuracy limit are used to define the Class P type CT.
Standard values of rated burden are: 2.5, 5, 7.5, 10, 15 and 30 VA.
Two accuracy classes are quoted 5P and 10P which gives a composite
error at rated accuracy limit of 5% and 10% respectively.
Standard accuracy limit factors are:
5, 10, 15, 20 and 30.
The method of describing a current transformer is as follows:
15 VA Class 5P20
which means that it is rated for a burden 15 VA and will not have more
than 5% error at 20 times rated current.
4.3.6 Application
In specifying current transformers the connected burden and mode of
operation must be taken into account paying attention not only to the wide
range of devices which may be connected, but also to the variation of
impedance over the range of setting any relay. For example, the normal
burden of an overcurrent relay is 3 VA at setting. The normal setting range
of the relay is 50% to 200% of nominal current. Therefore, a 1 A relay set to
50% would have a setting current of 0.5 A and the voltage across the coil at
this current would be
3 VA
V= =6V
0.5 A
and the impedance would be
6V
Z= = 12 Ω
0.5 A
At a setting of 200% the setting current would be 2 A, the voltage
3 VA
V= = 1.5 V
2A
and the impedance
1.5 V
Z= = 0.75 Ω
2A
If the characteristic of the relay is to be maintained up to 20 times
the relay setting, then a knee point voltage of not less than
20 × 6 V = 120 V for a 50% setting
or 20 × 1.5 V = 30 V for a 200% setting
would be required. The former is more onerous and therefore the lowest
setting must be taken into account when specifying the knee point voltage.
There is, however, an alleviating factor in that a relay operating at 20 times
its setting will saturated magnetically and therefore the impedance will be
reduced. The reduction for an overcurrent relay is about half the impedance
at setting, which means that in the above case a knee point voltage of 60 V
would be satisfactory.
In many cases the current transformers associated with the over
current protection must also cater for earth-fault relays. An earth-fault relay
having a minimum setting of 20% would have voltage at setting of
3 VA 15 V
= 15 V and impedance of = 75 Ω
0.2 A 0.2 A
The maximum earth-fault level may be restricted to, say, twice the
CT primary rating and therefore 10 times the relay setting. The knee-point
voltage should therefore be greater than 10 × 15 V = 150 V, or allowing for
saturation, 75 V.
In this case, the size is determined by the earth-fault relay. A suitable

current transformer would be a 7.5 VA Class 5P10. This would produce a
voltage of 7.5 V at rated current when connected to a 7.5 Ω burden and
would have only 5% error at 10 times rated current, i.e., at a voltage of
10 × 7.5 V = 75 V.
From the specification in the form 7.5 VA class 5P10, the knee point
voltage can be estimated. If it has a 5 A secondary winding then at rated
current it would produce 1.5 V across the rated burden and at 15 times
rated current 22.5 V. As a rough guide the knee point voltage is the product
of the VA rating and the accuracy limit factor divided by the rated secondary
current.
Class 5P is specified when phase-fault stability and accurate time
grading is required. When these are unimportant, Class 10P is suitable.
It may be that more than one relay is to be connected to one set of
current transformers in which case the total burden must be calculated. It
is generally sufficient to add the burdens arithmetically but it should be
borne in mind some alleviation may be available by adding the burden
vectorially in case of difficulties in design.
It is not good engineering practice to specify a current transformer
which is substantially larger than necessary as there is no advantage in
performance and its cost would be higher and its dimensions greater.
4.4 GUIDANCE IN APPLICATION OF CTs

4.4.1 Correlation of Transformer Class, Maximum Operating
Level and Categories of Protection (AS1675–1986)
The class, output voltage and accuracy required for a CT is determined by
the type of relay system to be driven, together with the fault level and
possibly the X/R ratio of the supply system.
Typical protection schemes are categorised in two ways:
(a) low speed schemes
(b) high accuracy high speed schemes.
Low Speed Protection Schemes
These are protection schemes having intentional time delays (fixed
or time dependent) of 0.2 sec. or greater. For such schemes good transient
performance is not necessary, since the initial saturation at maximum fault
current does not seriously affect the timing of the relay. For example,
induction disc IDMT overcurrent and earth fault; definite time overcurrent
and earth fault.
Typically a class 10P CT is satisfactory and terminal voltage should
be selected to drive the steady state maximum fault current through the
combined relay and secondary lead burden.
Care should be exercised when using solid state relays which

measure peak current and can be affected by offsets. (c.f. induction disc
relays measuring r.m.s. of fundamental components)
Short cuts in the CT design have been undertaken. For example,
under dimensioning CTs which saturate at much less than maximum fault
current where time delays and discrimination are not critical.
High Accuracy High Speed Protection Schemes
These are schemes (Fig. 4.12) for applications in which either
accuracy of setting and operating times, or through fault stability, or both
are important.
Class P
Spaces (optional)
e.g. AS 1675–1986 : 10P 150 F15
Rated composite error

at accuracy limit
Class
Rated secondary reference
voltage (volts)
Rated accuracy limit factor
Class PL
Spaces (optional)
e.g. AS 1675–1986 : 0.05PL 950 R3
Maximum secondary exciting current

(amps) at rated knee point voltage
Class
Rated knee point voltage
Maximum secondary winding resistance

(ohms) at 75°C or maximum
service temperature
FIGURE 4.12 Examples of class P and class PL
At distribution levels high impedance differential protection schemes

are typical.
While a Class P CT may be suitable a Class PL will always be better.
For this application, the PL CT is not essential for accuracy or
transient performance but it is convenient because the magnetising current
and CT secondary resistance are both required to be known for design of
the relay scheme.
It is intentionally allowed to saturate heavily at maximum fault
current and the output voltage should be designed to be in the range 2 to
5 times necessary for relay pick up.
4.4.2 Core Saturation and Current Limiting Action

In CT practice core saturation is frequently used to limit secondary current
which could otherwise reach a level high enough to cause damage to
equipment connected into the secondary circuit.
A measurement CT with a comparatively low accuracy class can be
designed as a much more effective saturation CT than a measurement CT
with a high accuracy class.
For example, Consider a CT of say class 5 (5%) (the permissible
current error is relatively high and no limit on the phase error).
At 125% of rated primary current the core can be operated at a fairly
high flux density.
A small increase in primary current above 125% raises the operating
point of the magnetisation curve into saturation and current limiting will
occur.
For a CT of higher accuracy class say 0.5 (0.5%) made of the same
core material the operating flux density at 125% of rated current has to be
established further down the magnetisation curve to maintain the ratio
and phase errors.
The core will be operating low down on the linear portion of the
magnetisation curve and any increase in current will require significant
change in magnetisation before saturation will occur.
The principle applies equally to protection CTs so that a class 10P
CT may be preferable for a saturation CT rather than a class 2.5P.
4.4.3 Through Primary and Wound Primary CTs
Current transformers appear in countless shapes, sizes, forms and types.
The through primary CT may be complete with a bar primary
provided by the manufacturer or it may be a window type or bushing
type to be fitted at installation stage on an existing primary bar or cable
(Fig. 4.13).
THROUGH-PRIMARY CT WOUND-PRIMARY CT
S2
P2 P1
S2
P1
S1
S1 P2
FIGURE 4.13 Primary current transformer

The primary is essentially a straight conductor which does not

mechanically encircle the core but effectively constitutes a single turn.
The through type CT is inherently extremely robust, both
mechanically and electrically.
When correctly designed, it is able to withstand the heavy
electromagnetic forces and thermal effects of major short circuits.
The inherent robustness and reliability have made this type of CT
the first choice except in those cases where the available primary current is
too small to enable the desired output, accuracy or performance to be
obtained economically.
A wound primary CT provides additional primary amp turns by
passing the primary current through a coil having any specified number
of turns.
This easy way of increasing output and accuracy has a penalty. In
the event of a short-circuit on the system, the mechanical forces on the
primary conductors are directly proportional to the number of turns.
This type of CT is not capable of withstanding heavy fault currents.
A further penalty is the increase in the internal burden of the CT
due to increased resistance of windings.
Often for testing and perhaps other reasons, it is convenient to
convert through primary CTs to a wound primary CT.
The primary cable can be looped through the window several times
creating an inserted primary CT.
As the primary amp-turns on the CT are unchanged the output and
accuracy will be the same.
However this approach should not be adopted in situation where
the prospective fault level is high enough to damage the CT.
Limits for a typical CT, it is worth considering some current limits
and their relative size.
For example, 500/5 CT rated primary current 500 A
(thermal limit) 1000 A
Primary Currents (accuracy limit) 10000 A
(short-time rated current) 50000 A
4.4.4 Derived Burden
The derived burden of a protection CT is roughly equivalent to the rated
burden of a measurement CT.
As the design of a CT is based on its performance at accuracy limit
primary current, it is important that the burden specified is the value at
accuracy limit primary current.
The burden represented by an electromagnetic relay when expressed

in ohms may be considerably less at twenty times setting current than at
setting current (as a consequence of relay core magnetic path saturation).
In the selection of a protection CT what matters most is the maximum
impedance of connected burden at rated accuracy limit current.
Secondary reference voltage accuracy limit
Derived burden =
Rated secondary current × Accuracy limit factor
4.4.5 Special Dimension C Ts

In the field due to close spacing of busbars or other dimensional limitations,
it is sometimes found that a standard CT has too large an OD (outside
diameter).
This problem can be solved by staggering the CTs. For example,
with the two outer phase CTs in line and the centre phase offset the CT OD
can be almost doubled.
A rectangular shape CT can overcome problems if staggering is not
appropriate or fails to meet the physical requirements.
4.4.6 CT Selection
Relay burdens are usually obtained in manufacturers specifications in units
(VA) at various current settings.
The VA obtained should be adjusted to a VA value at rated secondary
current. The rated CT burden can then be selected.
Where a problem has occurred in matching a CT to a relay or
specifying a CT, more often the selection has been made without the above
mentioned appropriate step.
For example, if a nominal 5 amp induction relay is to be used at the
50% tap (2.5 amps) and the burden of the relay is “2 VA at current setting”
then the actual burden at the selected tap will be
52
× 2 VA = 8 VA at 2.5 amp setting
2.5 2
As the current setting is reduced more relay turns are connected
across the CT, hence it stands to reason that the impedance will increase.
Using the relay manufacturers information and the ratio of squares
of current as a multiplier of nominal rated burden may lead to a very
conservatively designed CT.
A grossly over specified CT may be required to meet the calculated
demand. A larger CT would be necessary to provide a higher output
voltage.
Measurements on relays (electromechanical) indicate that the

impedance is not a fixed value and that the impedance could be a fraction
of its initial nominal value at 20 times setting current.
Unfortunately, few manufacturers provide relay impedance values
in ohms on the various tappings for a range of overcurrent conditions.
Some persistence from the user in demanding this information can
provide the desired outcome. Somewhere in the relay manufacturers
organisation, e.g., in a laboratory, full scale tests including ohmic values
for various taps on ranges of overcurrent conditions will be available.
Alternatively the relay user can perform its own tests (Fig.4.14).
CDG11 CDG14
2.5-10 A 0.5-2 A
0.4 1.4
3 VA 0.5 A
Impedance (W)
Impedance (W)
1A
2.5 A
5A 0.2 2A
10 A
1 Log-current 20 1 Log-current 20
times plug setting times plug setting
FIGURE 4.14 Realy impedance curves
If multi-tap relays are to be used over a full range of current tappings,

the burden on the CT should be calculated on the lowest relay current
setting.
This setting incorporates the maximum number of turns and the
highest coil impedance.
In practice, at distribution voltage levels, the range of settings is
limited and it is rare to vary the tapping point of the relay from one end to
the other. The min-op conditions typically being set by the CB rating.
If CTs are required for a triple pole relay it is important to determine
whether the relay incorporates three overcurrent elements or alternatively
two overcurrent and one earth leakage unit.
In the latter case a higher secondary reference voltage will be required
because of the typically low current setting adopted for the residual earth
leakage relay.
The relay manufacturers characteristic curve is a good preliminary
indicator of the accuracy limit factor of the selected CT.
For example, if relay operating times are still significant up to 20
times current setting then an accuracy limit factor of 20 will be required
for the CT.
If a 1 amp relay is used rather than a 5 amp relay then the secondary
reference voltage will be 5 times that of the equivalent 5 A secondary CT.
Saturation CTs with low accuracy limit factors should be used for
thermal relays (or devices) as heating is proportional to the square of the
CT output current.
4.4.7 CT Secondary Currents other than 5 A
Throughout most of the world CT ratings of 5 A or 1 A prevail. North
American utilities use the 5 A rating almost exclusively, however, elsewhere
lower ratings predominate.
It is worth examining the factors and relationships which influence
the decision to select a particular rating.
4.4.8 Insulation Requirements

Two aspects are considered, insulation between primary and secondary
for the system voltage and the insulation of the secondary winding to
withstand the secondary voltage produced during fault and test conditions.
The primary insulation is independent of the CT secondary current
rating, but the secondary insulation is not. For a given core and five times
the number of turns, a knee point voltage of the 1 A CT would be five
times that of a 5 A CT.
Volts per turn would be unchanged.
Where a lower knee point voltage is acceptable, the core section can
be reduced, resulting in a lower volts per turn.
It is apparent that a prime consideration is designing the secondary
system with coordinated insulation capability of the relays, interconnecting
cables, terminal blocks etc.
It may be considered that even when discussing 1 A and 5 A
secondaries that an arbitrary decision has been made. What is a ‘convenient
value’ of secondary current with respect to technical requirements and
economics.
Except in special cases nominal secondary currents above 5 A are
neither convenient nor economical.
Relay systems are designed to withstand fault currents in the range
20 times rating. Higher secondary currents result in excessive secondary
currents.
Lower secondary ratings offer advantages where lead burden
represents a significant portion of the total burden.
For a given wire size the lead burden at 1 A is only 4% of that
for 5 A.
Cost savings can be achieved using the 1 A secondary by reducing
the secondary wire size or using a smaller core etc.
4.4.9 Working Voltage

For a given amount of power out of a CT the required voltage must increase
in direct proportion to the current reduction.
Existing voltages in many 5 A systems are probably at a maximum
desirable level.
The requirement for CT output power will tend to be less with lower
secondary currents due to reduced copper losses in the leads. Modern static
relays require less operating power 0.5–1.0 VA versus nominal 3 VA for
electromechanical.
Although the CT knee point of a 1.0 voltage secondary is higher, it
should not be considered as a problem.
This voltage is normally present under excitation tests and the
voltage is distributed across the CT resistance.
Under operational conditions the terminal voltage is the more
important factor.
The connected burden under the most arduous fault conditions will
determine this. It is probably prudent to ensure that the terminal voltage
does not exceed 1 kV to allow adequate safety margin for insulation rating.
Where tapped CTs are used, a check should be performed to ensure
that autotransformer action does not overstress the open end of the CT.
Generally the terminal voltage of a 1 A CT will be less for cases
where the lead resistance RL is more than 20% of the relay burden at rated
current.
4.4.10 Transient Performance

Where consistent relay operating times are required, the CT may need to
have a large saturation factor, Ks.
The saturation factor relates the CT knee point voltage to the
maximum required steady state secondary voltage. The higher this factor
the better the CT performance under transient conditions.
The CT should not saturate when fully offset faults occur.
This requirement usually necessitates that the physical size of the
CT could be a problem, situating CTs within bushings or in metal clad
switchgear. The smaller 1 A secondary CT offers an advantage.
For equivalent transient performance the CT lead resistance of a 1 A
system can be 12.5 times that of a 5 A system (assuming that the relay
resistance is a small fraction of the lead resistance).
4.4.11 Practical Considerations

Interchange ability and spares may be considered as detrimental factors
where an organisation is already committed to a 5 A system.
Note that perceived problems are just that. Many modern relays are
suitable for use on both systems.
Interposing CTs can be used to marry and mix 5 A and 1 A systems.
4.4.12 Safety Considerations
The open circuit voltage of a 1 A CT could be higher than an equivalent
5 A CT. Safe operating procedures are essential for all CTs and hence the
hazard is not a relative measure worth considering.
The 1 A CT can offer the following advantages:
• improved transient performance
• reduced cable or CT size
• reduced voltage stress in certain applications
In all cases, advantages are strongest when the relay resistance to
cable resistance ratio decreases.
Although 5 A secondaries are still desirable for the lower end of
high voltage at high current ratio applications all the pertinent parameters
should be considered for an engineering evaluation of specific installations.
CHAPTER
Overcurrent Protection
5.1 INTRODUCTION
This type of protection which was developed some 70–80 years ago, is the
earliest, cheapest and simplest form of protection still widely used in the
power industry today. It operates on the principle that once a
predetermined level of current is reached the relay will operate in a
predetermined time. These relays are generally classified broadly in terms
of their time characteristic e.g., instantaneous, inverse time, extremely
inverse time, fixed or definite time.
Overcurrent protections should not be confused with ‘overload’ type
protections which consider the thermal capability of the plant to be
protected. The overcurrent protections are directed entirely to the clearance
of faults, although with the settings usually adopted some sort of overload
protection is normally provided.
These overcurrent protection schemes are commonly used to provide
short-circuit protection for many different types of plant. As well as being
applied to detect phase-phase short-circuits, such relays can be used to
provide earth leakage (phase-ground fault) protection. The overcurrent
relay in this case is connected in the neutral of the current transformer
secondary circuit.
Although the overcurrent protections are inherently non-directional
(i.e., they will operate for primary current flow in either direction), it is
possible for them to be controlled so that they effectively respond to primary
current in one direction only and this is done using a separate ‘directional’
relay to control the ‘secondary’ circuit of the overcurrent protection. A
‘directional’ over current protection facilitates the setting of overcurrent

protections on tie lines or ring systems.
Depending on the particular application, overcurrent relays would
be used:
• in unit or non-unit type schemes (the more common is in non-
unit type schemes) or,
• with directional or non-directional features (the more common
is non-directional).
In the case of the non-unit type schemes, it is necessary to achieve
discrimination between the local and the remote over current protections.
This discrimination is usually achieved by the application of appropriately
chosen settings.
With the advent of microprocessor based overcurrent protections,
additional and more enhanced features have been provided to the
traditional electromechanical overcurrent protections, these enhanced
features have increased the potential range of applications.
Overcurrent protections find greatest application in the distribution
area and to a somewhat lesser extent in the sub-transmission area, both for
the purpose of phase fault and earth fault protection. These protections
may be found on generators, lines/feeders, busbars, transformers,
transformer/feeder combination, reactors, capacitor banks, etc.
However, non-unit type overcurrent protections do not have
significant application on EHV transmission plant because of their limited
sensitivity and non-discriminatory nature requiring operating times that
are generally unacceptable at that system level.
5.2 TYPES AND CONSTRUCTION

The design/construction technology of the overcurrent relays range from
the electromechanical single characteristic inverse curve relay (induction
disk type) to the microprocessor based relay with the choice of multi-
characteristic curves.
The operating time characteristics and limits of accuracy of the
overcurrent relays are in accordance with the limits set by IEC255-4 and
BS142 for an inverse definite minimum time (IDMT) overcurrent relay.
Typical operating time characteristic and tolerance limits are as shown in
Fig. 5.1.
The operating time of the instantaneous and definite time relays is,
practically, independent of the magnitude of the fault current once the
magnitude exceeds the operating threshold of the relay.
OVERCURRENT PROTECTION 103
50
Pick-up 1.05-1.3 Times setting
40
30
20
Time (Seconds)
10
8
7
6
5
4
3
1
2 3 4 5 6 7 8 10 20 30
Current (multiples of plug setting)
Time/Current characteristic allowable limit:

At 2 times setting 2.5 × Declared error
FIGURE 5.1 Typical operating characteristic and operating tolerance limits
The range of typical operating time characteristics (including

mathematical formula applicable for the static/digital) of the various types
of inverse overcurrent relays are shown in Fig. 5.2, and are as follows:
(a) Definite time
(b) Standard inverse (IDMT) Tm = 0.14/((PSM)0.02 – 1.0)
(c) Very inverse Tm = 13.5/(PSM – 1.0)
(d) Extremely inverse Tm = 80/((PSM)2 – 1.0)
(e) Long time standby earth fault Tm = 120/(PSM – 1.0)
where Tm = Relay operating time (seconds) at TMS = 1.0
PSM = Plug setting multiplier (multiple of tap/plug setting)
It should be noted that all the operating time characteristics,
including mathematical formula are shown at 1.0 time multiplier setting
(TMS).
The microprocessor based versions which are available from most
relay manufacturers have many advantages over the electromechanical
relay, these include the following:
• lower ac burdens,
• wider application ranges,
• more accurate, not affected by mechanical problems such as pivot
friction,
• can be utilised to provide disturbance records, event records,
system load data and fault quantities,
• has communication capability to enable remote interrogation,
• self monitoring to ensure improved reliability and availability
capabilities.
2
10
1
10 Definite
8 seconds
Operating time, t(seconds)
Definite
4 seconds
Long time standby
Earth fault
Definite 120
t=
2 seconds I–1
Standard inverse
0.14
t = 0.02
0
10 I –1
Very inverse
13.5
t=
I–1
Extremely inverse
80
t= 2
10
–1 I –1
0 1 2
10 10 10
Current (multiple of setting)
FIGURE 5.2 Various types of inverse type characteristics

including mathematical formulae
Although the microprocessor version will find increasing

applications, the electromechanical induction type relays will continue to
remain in service for many years.
All relays are provided with a number of tap/plug settings, each of

which represents the minimum current at which the relay will start to
operate (this is referred to as the minimum pickup value). A relay that has
been set on a particular tap/plug will begin to operate at that tap/plug
setting plus/minus the manufacturer’s tolerance. However at this current
the time can be extremely long and unpredictable, any small deviation at
this level will result in significant time changes. For this reason
manufacturer’s generally do not show their time curves below 1.5 to 2.0
times minimum pickup. In practice this part of the curve should not be
used for protection.
100
80
60
50
40
30
20
10
8
6
5
4
3
2 1.0
Time (Seconds)
0.9
0.8
0.7
0.6
1 0.5
0.8 0.4
0.6 0.3
0.5
0.4 0.2
0.3
inst = 1 × Is
0.2 0.1
inst = 2 × Is
0.1 0.05
inst = 3 × Is
0.08
0.06 inst = 4 × Is
0.05 inst = 5 × Is to 31 × Is
0.04
0.03 Instantaneous
0.02 Unit
0.01
1 2 3 4 5 6 8 10 20 30 40 5060 80100
Current (multiple of setting)
FIGURE 5.2 (a) Typical family of inverse type operating time

characteristics for the various time multiplier settings
10 10
9 9
Time current characteristic Time current characteristic
8 8
inverse time relay CDGIl inverse time relay CDGII
7 7
1-3 sec 3 sec to B.S.142
6 6
5 5
4 4
3 3
1.0
2 2 0.9
0.8
Operating time in secs
Operating time in secs

0.7
0.6
0.5
1 1.0 1
0.9 0.9 0.9 0.4
0.8 0.8 0.8
0.7 0.7 0.7 0.3
0.6 0.6 0.6
0.5 0.5 0.5
0.2
0.4 0.4 0.4
0.3 0.3 0.3

0.1
0.2 0.2 0.2
0.1 0.1 0.1

2 3 4 5 6 7 8 910 20 2 3 4 5 6 7 8 910 20
Multiples of plug setting current Multiples of plug setting current

Curve reference 398. s23. 15 Curve reference S77398z06.017
FIGURE 5.2 (b)
The current axis of the characteristic curves is shown in multiples of

tap/plug or pickup current, this is for the convenience of providing one
scale for all tap/plug settings.
In addition to the tap/plug settings, there are further adjustable
settings which are identified by the scale known as ‘time multiplier’, ‘time
lever’ or ‘time dial’ settings. These settings provide different operating
times for the same operating current level, thus a family of operating time
curves are available for use. Typical family of inverse type operating time
characteristics for the various time dial settings are shown in Fig. 5.2(a).
5.3 SYSTEM ANALYSIS

To achieve correct coordination of overcurrent protection, it is necessary
to have a detailed knowledge of:
• the protection schemes themselves,
• the possible operating conditions of the system to be protected,

and
• the minimum and maximum fault currents that can flow in each
part of the network to be protected.
Since large scale tests of the system are not practical, theoretical
system analysis must be carried out. As a consequence to enable the
protection settings to be determined, the data that may be needed includes
the following:
(a) A single-line diagram of the power system, indicating the type
and rating of the protective devices and their associated current
transformers.
(b) The impedance in ohms, per cent or per-unit of all power
transformers, rotating machines and feeder circuits.
(c) The starting current requirement of large motors, as well as
starting and stalling times of the induction motors.
(d) The maximum and minimum values of short-circuit currents that
are expected to flow through each of the protective devices.
(e) The maximum peak load current (which includes all short-time
overloads due to motor starting or otherwise and does not refer
to the peak of the current waveform) through protective devices.
(f) The decrement curves showing the rate of decay of the fault
current supplied by the generators, if applicable.
(g) The excitation curves of the current transformers and details of
secondary winding resistance, load burden and other connected
burdens.
5.4 SETTINGS OF IDMT RELAYS

Given that the type of characteristic to be applied has been selected (i.e.,
standard inverse, very inverse, etc.), then there are two factors to be
considered whilst grading IDMT relays, these are the basic adjustable
settings which apply to all inverse time relays:
(i) TMS —time multiplier setting, and
(ii) PSM—plug setting multiplier.
The operating time characteristics covering the range of relay
characteristics and TMS settings, for static/digital relays can be expressed
by the following general formula:
T = K × [TMS]/[(I/IS)α – 1.0]
where T = actual relay operate time (seconds)
K = constant for the particular characteristic:
0.14 for standard inverse
13.5 for very inverse

80 for extremely inverse
120 for long time standby earth current
TMS = time multiplier setting (seconds)
I = fault current (ampere)
IS = current threshold setting of the relay – tap/plug
setting (ampere)
α = constant for the particular characteristic
0.02 for standard inverse
1.0 for very inverse
2.0 for extremely inverse
1.0 for long time standby earth current
The value of TMS for an inverse time relay is generally adjustable
from 0.05–1.0, and its setting is defined as:
TMS = T/Tm (for electromechanical relays)
TMS = T/Tm = (T × [(PSM)α – 1.0])/K (for static/digital relays)
where T = The required time of operation (seconds)
Tm = The time obtained from the relay characteristic curve
(or from the formula) at TMS = 1.0, and using the PSM
equivalent to the maximum fault current.
The value of PSM is defined as:
Fault current (I)
Current threshold setting of the realy – Tap/plug setting (I S )
Note that the current values for both I and IS must have a common
reference, it can either be in terms of primary currents or in terms of
secondary currents.
5.4.1 TMS Examples

Consider the standard inverse characteristic as shown in Fig. 5.2(a), if TMS
is 0.1 and the time Tm obtained from the curve, for a particular PSM current,
is 4.0 seconds then the actual operating time will be:
T = T m × TMS
= 4.0 × 0.1
= 0.4 seconds.
Conversely, if the required operating time is 0.4 seconds for a
particular PSM current, and the time Tm obtained from the curve is 4.0
seconds then the required TMS should be
T 0.4
TMS = = = 0.1
Tm 4.0
Increasing the TMS has the affect moving the curve higher on time
scale.
5.4.2 PSM Examples
Consider the situation where the maximum fault current flowing through
the relay location is 3000 A (primary) and the relay is set to operate at
200 A (primary),
then PSM = 3000/200 = 15
Consider the same example, but in this case the relay setting has
been given as a tap/plug setting of 50% and the CT ratio of 400/5, then to
calculate PSM it necessary to refer all currents to either primary or
secondary values. If currents are referred to the primary,
3000 3000
then PSM = = = 15 (as above)
(0.5 × 5) × (400 / 5) 200
However, if the same primary current was considered but with a
relay current setting of 200%,
3000 3000
then PSM = = = 3.75
(2.0 × 5) × (400 / 5) 10 × 80
5.5 RELAY DISCRIMINATION

To ensure that the protection will correctly select and isolate the faulty
section of the network only, thereby leaving the rest of the system
undisturbed, it is necessary to achieve proper relay coordination/
discrimination.
Discrimination can be achieved by:
1. time
2. current
3. combination of time and current.
5.5.1 Discrimination by Time

To ensure selectivity of operation under all circumstances in a radial feeder,
the operating time of the protection is increased by the grading margin
(usually 0.4 or 0.5 seconds) from the far end of protected circuit towards
the generating source. This grading is very conveniently achieved with
the help of definite time-delay relays (which consist of an instantaneous
overcurrent relay followed by a timing relay), the contacts of the latter

initiate the trip the breaker.
Let us consider a simple radial system as illustrated by Fig. 5.3. (Time
grading in this instance has been assumed to be 0.4 seconds).
A B C D E
1.45 1.05 0.65 0.25
FIGURE 5.3 Simple radial case (time grading assumed 0.4 seconds)
The protection is provided at the sending end of each section A, B,

C, D, and E, the relay at D is set to the shortest possible time to allow the
fuse to blow for a fault which is in the secondary side of the transformer E.
Hence, if there is a fault at F, the relay at D operates at 0.2 seconds, the
circuit breaker at D clears the fault. The protections at C, B and A provide
back up. For faults between C and D the relay at C trips breaker at C and
so on.
One of the greatest disadvantages of this system is that the longest
fault clearance time occurs for faults in the section closest to the source
where the fault level is highest.
Let us consider another simple radial system as illustrated in
Fig. 5.3(a). (Time grading in this instance has been assumed to be
0.5 seconds).
D C B A
200/5 300/5 300/5 400/5
Max. fault level

(3 phase)
2000 A 3000 A 5000 A 6000 A
FIGURE 5.3 (a) Simple radial case (time grading assumed 0.5 seconds)
Start with the relay at ‘D’

Choose TMS = 0.1 (say).
Assume that the load current does not exceed 80 A (CT is 200/5),
then the relay tap/plug setting can be set to (1.2 × 80)/(200/5) = 2.4 A
select say 2.5 A (i.e., 50%), then
Fault current 2000 (primary)
PSM = = = 20
Relay current setting 2.5 × 40 (primary)
If reference is made to the standard IDMT curve, Fig. 5.2(a) (at TMS
= 1.0), it can be seen that with PSM equal to 20, the corresponding Tm is
2.3 seconds.
When the formula for the static/digital relays is used, the same result
is obtained.
Tm = 0.14/((PSM)0.02 – 1.0) = 2.3 seconds
T
as TMS =
Tm
T = 2.3 × 0.1 = 0.23 seconds
T = K × [TMS]/[(I/IS)α – 1.0] = (0.14 × 0.1)/[200.02 – 1.0]
= 0.23 seconds
Continue with the relay at ‘C’
Now, for the same current of 2000 A the relay at ‘C’ must be set to
operate at 0.5 seconds (time delay step) longer than the relay at ‘D’, i.e.,
0.23 + 0.5 = 0.73 seconds.
If one assumes that the load at ‘C’ is higher than the load at ‘D’, then
the current setting at ‘C’ must be increased as compared to that of ‘D’.
Assume that the setting at ‘C’ needs to be 100% and uses 300/5 CT, then
Fault current 2000 (primary)
PSM = = = 6.67
Realy current setting 5.0 × 60 (primary)
If reference is made to the standard IDMT curve (at TMS= 1.0), it
can be seen that with PSM equal to 6.67, the corresponding Tm is 3.6 seconds.
is obtained.
Tm = 0.14/((PSM)0.02 – 1.0) = 3.6 seconds
Hence for the required operating time of 0.73 seconds, the value of
T
TMS = = 0.73/3.6 = 0.20 seconds
Tm
is obtained for TMS.
TMS = (T × [(I/IS)α – 1.0])/K =(0.73 × [6.670.02 – 1.0])/0.14
= 0.20 seconds
When faults occur close to ‘C’ the fault current is 3000 A, the
corresponding PSM value for the relay at ‘C’ is,
3000
PSM = = 10
5 × 60
With PSM equal to 10, the corresponding Tm is 3.0 seconds and,
with TMS = 0.20 seconds the actual operate time for the relay is
T = 0.20 × 3.0 = 0.60 seconds

is obtained for T.
T = K × [TMS]/[(I/IS)α – 1.0] = (0.14 × 0.2)/[100.02 – 1.0]
= 0.60 seconds
Continue with the relay at ‘B’
Now at ‘B’ the required operating time for a fault at ‘C’ is 0.6 + 0.5
= 1.1 seconds. Because of increased load assume that the current setting at
B needs to be increased to 150%, then
3000
PSM = = 6.67
7.5 × 60
With PSM equal to 6.67, the corresponding Tm is 3.6 seconds.
T
TMS = = 1.1/3.6 = 0.31 seconds
Tm
TMS = (T × [(I/IS)α – 1.0])/K = (1.1 × [6.670.02 – 1.0])/0.14
= 0.30 seconds
When faults occur close to ‘B’ the fault current is 5000 A, the
corresponding PSM value for the relay at ‘B’ is,
5000
PSM = = 11.1
7.5 × 60
With PSM equal to 11.1, the corresponding Tm is 2.84 seconds and,
with TMS = 0.30 seconds the actual operate time for the relay is
T = 0.30 × 2.84 = 0.85 seconds
is obtained for T.
T = K × [TMS]/[(I/IS)α – 1.0]
= (0.14 × 0.3)/[11.10.02 – 1.0]
= 0.85 seconds
Continue with the relay at ‘A’
Now at ‘A’ the required operating time for a fault at ‘C’ is 0.85 + 0.5
= 1.35 seconds. Because of increased load, assume that the current setting
at ‘A’ needs to be increased, but because of the higher CT ratio, sufficient

margin is achieved by having the relay tap/plug setting on 150%, then
5000
PSM = = 8.34
7.5 × 80
With PSM equal to 8.34, the corresponding Tm is 3.23 seconds.
T
TMS = = 1.35/3.23 = 0.42 seconds
Tm
TMS = (T × [(I/IS)α – 1.0])/K = (1.35 × [8.340.02 – 1.0])/0.14
= 0.42 seconds
When faults occur close to ‘A’ the fault current is 6000 A, the
corresponding PSM value for the relay at ‘A’ is,
6000
PSM = = 10
7.5 × 80
With PSM equal to 10, the corresponding Tm is 3.0 seconds and, with
TMS = 0.42 seconds the actual operate time for the relay is
T = 3.0 × 0.42 = 1.26 seconds
The summary of corresponding operating times for the above
example are shown in Fig. 5.3(b).
A B C D E
1.35 1.1 0.73 0.23
FIGURE 5.3 (b) Summary of operating times for example
5.5.2 Discrimination by Current

Grading by current alone relies on the fact that the fault current along the
length of the protected circuit decreases as the distance from the source to
the fault location increases. The relays controlling the various circuit
breakers need to be set to operate at suitable values such that only the
relay nearest to the fault trips its breaker.
This method, however, relies on significant differences between the
fault levels at the consecutive stations, for most practical applications this
is very rarely achieved. In addition it would not be possible to use current

grading as the sole protection scheme of a primary plant because the
protection operation zones of such a scheme would not overlap and large
‘blind’ sections of the network would be created, and these in turn would
remain unprotected. This method, therefore, is used mainly in a
supplementary capacity in some schemes.
Let us consider a simple radial system as illustrated by Fig. 5.3(c).
0.12 W/km 0.02 W/km

11 kV 2000 metres 2000 metres 4 MVA
2 2
250 MVA 240 mm 120 mm 11/3.3 kV
source P.l.C. cable P.l.C. cable 7%
F1 F2 F3 F4
C B A
FIGURE 5.3 (c) Radial system using current and time
Consider the three phase fault levels at F1

Vf
IF = 1
1 (Z S + Z11 )
where Vf = Source phase to ground voltage
1
Zs = Positive sequence source impedance (ohm)

= (kV)2/MVA = 112/250 = 0.484 ohms
Z L1 = Positive sequence line impedance (ohm)
= 0.12 × 2 = 0.24 ohms

6350
∴ IF = = 8.77 kA
1 (0.484 + 0.24)
The relay controlling the CB at ‘C’, if set to operate at 8.77 kA, would
in theory protect the whole of the cable section between ‘C’ and ‘B’.
However, two important practical aspects which affect the method of
coordination can be demonstrated here, they are as follows:
1. The relay would not be able to distinguish between faults at F1
and F2 (as the difference between the distance may be very small
(a few metres)). Hence the corresponding change in IF would
1
be negligible.
2 . The relay settings would have assumed that the source fault level
is constant, whereas for the practical operating environment, it
is possible that the source fault level is reduced from 250 MVA
to 130 MVA.
Under these conditions the source impedance would increase
and the fault levels would reduce accordingly.
Zs = 112/130 = 0.93 ohms
6350
∴ IF = = 5.43 kA
1 (0.93 + 0.24)
At the lower fault level IF = 5.43 kA, the relay if it had been set
1
for 8.77 kA would not protect the cable section. Hence, in this
case current grading would not be a practical solution.
Consider the three phase fault levels at F4 (assume the source fault level to
be 250 MVA)
6350
∴ IF =
4 (Z s + Z L1 + Z L2 + Z T )
where ZL = Positive sequence line impedance = 2 × 0.02 = 0.04 ohms
2
ZT = Positive sequence transformer impedance
= 0.07 × [(112)/4 = 2.12 ohms
6350
∴ IF = = 2200 A
4 (0.484 + 0.24 + 0.04 + 2.12)
The relay controlling the CB at ‘B’, if set to operate at 2.2 kA, would
in theory protect the whole of the cable/transformer section between ‘B’
and ‘A’. However, it would not discriminate with the relay controlling the
CB at ‘A’.
If discrimination were to be achieved, a safety margin of say 30%
would need to be allowed for (assume 20% for relay error and 10% for
variation in the system impedance values).
∴ Choose a relay setting of 1.3 × 2.2 kA = 2.86 kA (relay at ‘B’)
However, the increased setting for the relay at ‘B’ would create a
‘blind’ spot where no protection was being provided. Again this would
not be a practical solution.
Consider the three phase fault levels at F3 (assume the source fault level to
be 250 MVA)
6350 6350
∴ IF = =
3 (Z s + Z L1 + Z L 2 ) (0.484 + 0.24 + 0.04)
= 8.3 kA
If the source fault level was 130 MVA then,

6350
IF = = 5.25 kA
3 (0.93 + 0.24 + 0.04)
Therefore, for this application with the relay at ‘B’ set to 2.86 kA, the
section of line from ‘B’ to the transformer would be protected.
5.5.3 Discrimination by the Combination of Time and Current
As demonstrated in earlier sections, discrimination by time alone will cause
the more severe faults to be cleared in the longest operating time,
discrimination by current alone can only be applied when there is
appreciable impedance between the two circuit breakers concerned,
however, even in these cases ‘blind’ spots would be created. Therefore,
there are severe limitations with both of these methods.
Because of these limitations, the inverse time overcurrent relay
characteristic has evolved where both the functions of time and current
are considered for achieving discrimination. If the fault level falls
substantially from substation to substation with increasing distance from
the source, then much faster fault clearance than that in the time grading
method may be achieved through grading by time and current.
By proper selection and setting of the inverse characteristics, a
discriminative plan of relatively fast tripping can be developed as detailed
in Fig. 5.3(d). As can be seen, even with the proper grading margin of at
least 0.4 seconds being maintained, fault clearance times near the source
are almost as quick in this scheme as clearing a fault a long way from the
source. The 0.4 seconds grading margin is required to cover such delays as
circuit breaker clearance time, relay overshoot time and relay timing errors.
The advantages of this method of relay coordination are
demonstrated by considering the system as shown in Fig. 5.3(d), and the
following calculations.
It should be noted that the graph in Fig. 5.3(d) illustrates the use of
‘discrimination curves’ which are an essential aid to ensure correct
protection coordination.
For the example, being considered, the 200 A fuse is the first curve
to be plotted as it is assumed to protect the largest outgoing 3.3 kV circuit.
Note that all fault currents have been referred to a 3.3 kV voltage base so
that all the protective devices to be considered can be plotted on the same
graph. The protection relays for this system have been assumed to be of
the Extremely Inverse type (CDG14 type characteristic).
Once the 200 A fuse has been plotted, then the grading of the
overcurrent relays from the remote station ‘E’ to the source station ‘A’ are
carried out progressively.
1000
Relay Relay
100 C A
Relay Relay
D B
10
Time (seconds)
200 A
fuse
0.1
35.7 98.7123 1540 3500 MVA

0.01
100 1000 10,000 100,000
Fault current (amperes) 3.3 kV Base
3500 MVA 1540 MVA 123 MVA 98.7 MVA 35.7 MVA
15,000 metres 2000 metres 200 metres

2 2
240 mm 30 MVA 2
132 kV 240 mm P.l.C. 120 mm P.l.C. Fuses
overhead line 132/11 kV cable cable
3500 MVA 6.2 ohms 200 A
22.5% 0.24 ohms 0.04 ohms
source
150 A
500/1 A 150/1 A 500/5 A 250/5 A 4 MVA
A B 11/3.3 kV
C D E
7%
132 kV 11 kV 3.3 kV
FIGURE 5.3 (d) Discrimination plan as detailed in 0.4 seconds grading

In addition for the purposes of this example, all the fault calculations
demonstrated below are all referred to a common 10 MVA base.
Per cent impedance of the system elements on l0 MVA base
4 MVA transformer = 7 × 10/4 = 17.5%
10
11 kV cable (ED) = 0.04 × × 100 = 0.33%
112
10
11 kV cable (DC) = 0.24 × × 100 = 1.98%
112
30 MVA transformer = 22.5 × 10/30 = 7.5%
10
132 kV overhead line = 6.2 × × 100 = 0.36%
132 2
10 × 100
132 kV source = = 0.29%
3500
Consider SUBSTATION ‘D’
The protection at this substation is the first relay that must
discriminate with the 200 A fuse at fault levels up to:
10 × 100
= 35.77 MVA
17.5 + 0.33 + 1.98 + 7.5 + 0.36 + 0.29
35.77 × 1000
i.e., If = = 6258 A at 3.3 kV
3 × 3.3
35.77 × 1000
or If = = 1877 A at 11 kV
3 × 11
For satisfactory coordination between the fuse and the relay, the
primary current setting of the relay should be approximately three times
the current rating of the fuse and, the time grading margin, when expressed
as a fixed quantity should not be less than 0.4 seconds, or when expressed
as a variable quantity should have a minimum value of (0.4t + 0.15)
seconds, where ‘t‘ is the nominal operating time of the fuse.
For the above fault condition the fuse operating time at 6258 A is
approximately 0.01 seconds, therefore the required operating time of the
relay should be equal to or greater than (0.01 × 0.4 + 0.15) = 0.154 seconds.
The relay current setting must also accommodate a safe load of
4 MVA (the rating of the 11/3.3 kV transformer). By selecting a 100% plug
setting i.e., 250 A and 4.76 MVA at 11 kV this requirement will be satisfied
(19% safety margin), also the requirement of three times the current rating
of the fuse is satisfied (i.e. [250 × (11/3.3)]/200 = 4.2). As a consequence
PSM = 1877/250 = 7.51.
From the above data, the time multiplier setting can then be selected
using the following formula:
TMS = T/Tm for electromechanical relays
= T/Tm = (T × [(PSM)2 – 1.0])/80
(for static/digital based relays)
From the characteristic curve of the extremely inverse relay for
1.0 TMS, the Tm for PSM of 7.51 is 0.9 seconds, therefore
TMS = 0.154/0.9 = 0.17
Select the next highest available time multiplier setting, in this case
0.2, this setting will ensure that suitable discrimination with the fuse is
achieved.
Consider SUBSTATION ‘C’
The relay in this case must discriminate with the relay in substation
‘D’ at fault levels up to:
10 × 100
= 98.7 MVA
1.98 + 7.5 + 0.36 + 0.29
i.e., 17.27 kA at 3.3 kV or 5180 A at 11 kV
In addition to providing primary protection to the section ‘C’ to ‘D’,
relay ‘C’ must also provide backup protection to relay ‘D’ protection
covering the section ‘D’ to ‘E’.
By selecting 100% plug setting i.e., 500 A and 9.33 MVA at 11 kV
(1667 A at 3.3 kV ) the sensitivity of the relay meets both the primary and
backup requirements.
Following the same procedure as was detailed for SUBSTATION
‘D’ by selecting a time multiplier setting of 0.7 for relay ‘C’, the required
discrimination with relay at substation ‘B’ is achieved.
Consider SUBSTATION ‘B’
‘C’ at fault levels up to:
10 × 100
= 123 MVA
7.5 + 0.36 + 0.29
i.e., 21.5 kA at 3.3 kV or 6.46 kA at 11 kV or 537 A at 132 kV.
In addition to providing primary protection to the section ‘B’ to ‘C’,
relay ‘B’ must also provide back up protection to relay ‘C’ protection
covering the section ‘C’ to ‘D’.
Following the same procedures as was detailed for SUBSTATION

‘C’, by selecting at 132 kV, 100% plug setting (i.e., 150 A and 34.3 MVA at
132 kV), and a time multiplier setting of 0.7 for relay ‘B’, the required
discrimination with relay at substation ‘C’ is achieved.
Consider SUBSTATION ‘A’
‘B’ at fault levels up to:
10 × 100
= 1540 MVA
0.36 + 0.29
i.e., 269.4 kA at 3.3 kV or 16.1 kA at 11 kV or 6.75 kA at 132 kV.
In addition to providing primary protection to the section ‘A’ to ‘B’,
relay ‘A’ must also provide backup protection to relay ‘B’ protection
covering the section ‘B’ to ‘C’.
Following the same procedures as were detailed for SUBSTATION
‘C’, by selecting at 132 kV, 100% plug setting (i.e., 500 A and 114 MVA at
132 kV), and a time multiplier setting of 0.9 for relay ‘A’, the required
discrimination with relay at substation ‘B’ is achieved.
If a comparison is made of the fault clearance times between the
‘time’ and ‘time and current’ grading approaches, as summarised in the
following table, it is evident that the ‘time and current’ (inverse time
characteristic) is far superior to the definite time overcurrent relay approach.
Relay Fault Time Time and Advantage

(MVA) Grading Current of Time and
Concept Grading Current
Concept Concept
(Seconds) (Seconds) (Seconds)
D 98.7/35.7 0.25 0.07/0.17 0.18/0.08

C 122.7/98.7 0.65 0.33/0.42 0.32/0.23
B 1540/122.7 1.05 0.07/0.86 0.98/0.19
A 3500/1540 1.45 0.25/0.39 1.20/1.06
Consequently for faults close to the relaying points, the inverse time
characteristic can achieve SIGNIFICANT reductions in fault clearance
times. Even for faults at the remote end of the line, considerable time
reductions are also achieved.
5.6 GRADING MARGIN

The time interval between the operation of two adjacent relays is dependent
on a number of factors, these include the following:
(i) Circuit Breaker Time to Interrupt Fault Current
(ii) Overshoot Time of the Relay
(iii) Tolerances/Errors
(iv) Final or Safety Margin
5.6.1 Circuit Breaker Time to Interrupt Fault Current
The circuit breaker must have completely interrupted the fault current
before the discriminating relays cease to continue operating. Therefore,
the circuit breaker total operating time must be included as part of the
time interval that is to be allowed for between discriminating relays.
5.6.2 Overshoot Time of the Relay
Even after a fault has been cleared by the remote protection, the local relay
may continue to operate until all the stored energy from the system fault
condition has been dissipated e.g., an induction disc relay has stored kinetic
energy in the motion of the disc, static relays have energy in its capacitors.
Although relay design is directed to minimise and absorb these
energies, the problem in general still remains. As a consequence some
allowance for overshoot must be included as part of the time interval that
is to be allowed for between discriminating relays (overshoot is more
predominant with electromechanical relays).
5.6.3 Tolerances/Errors
All components (i.e., relays, current transformers, system parameters used
to calculate fault currents, etc.) have errors. The operating time
characteristic of relays may have positive or negative errors, the current
transformers may have phase and ratio errors and the system parameters
such as impedances have positive or negative errors.
Therefore, it cannot be assumed that the information on relay
characteristics, etc. as supplied by manufacturers are without error.
Manufacturers generally supply tolerances for their equipment.
As a consequence some allowance must be made for these
tolerances/errors when applying the settings of discriminating relays.
5.6.4 Final or Safety Margin

In spite of the aforementioned allowances the discriminating relay must
ensure that the remote relay has been given the opportunity to clear the
fault and therefore not complete its own operation and unnecessarily trip
a section of line/plant that is not faulted.
Some extra allowance is generally required to ensure that a

satisfactory safety margin remains.
The total recommended time interval between discriminating relays
depends on the operating speed of the circuit breakers, relay performances
and accuracy of the system data.
In the past, 0.5 seconds was the normal grading margin. However,
with the advent of faster circuit breakers and reduced relay overshoot times,
it has been possible to reduce the total time for the grading of successive
relays to 0.4 seconds. This margin allows up to:
• 0.15 seconds for circuit breaker to interrupt the fault current,
• 0.05 seconds for overshoot time,
• 0.10 seconds for the tolerances in the relays and CTs, and
• 0.10 seconds for safety margin.
However, instead of using a fixed grading margin (0.5, 0.4 seconds
or other), it is preferable to adjust the time to satisfy the particular
application. In general it is preferable for the grading margin to use a fixed
time value for the operating time of the circuit breaker and relay overshoot
and, a variable time which takes into account the relay and CT errors as
well as the safety margin. On this basis the following could then apply:
• Select a fixed time value for the circuit breaker fault clearance
time (say 0.10 seconds),
• Select a fixed time value for the relay overshoot time (say 0.05
seconds),
• Select a fixed time of 0.1 seconds for the safety margin, and
• Allow 15% as total effective error and 10% as CT errors (total
error 25%).
Hence the time interval t’ between inverse time overcurrent relays
is given as:
t’ = (0.25t + 0.25) seconds
where, t is the nominal operating time of relay nearest to the fault.
The above demonstrates that each case study should be evaluated
on its own merits, however, the factors that need to be considered to
determine the grading time between discriminating relays remain
unchanged.
5.7 EARTH FAULT PROTECTION

Earth faults are the most frequent of all faults. The magnitude of earth
fault current is usually low when compared to the phase fault currents,
which is due to the fault impedance path. The fault impedance path may
include earth resistance or neutral earth impedance, this will depend on

whether the system is solidly earthed, insulated or earthed through some
resistance and/or reactance.
Hence, as a general rule more sensitive protections need to be
provided for the earth faults when compared to the multi-phase type faults.
Therefore, the connections made for the earth fault protections will need to
be different when compared to those used for the multi-phase fault
protections. More sensitive earth fault protection can be provided by the use
of relays which respond only to the residual current of the system (as the
residual component only exists when fault current flows to the earth). This
then allows these protections to be set independent of load current
(balanced or otherwise) and settings below normal load current can be
achieved.
Therefore, the earth fault relays have low settings which are generally
between 20% to 80% of CT rated current. However, in the case of
electromechanical relays with low current setting values, the relay could
impose significantly higher burdens on the CT. Hence, special
considerations may apply when low settings are to be applied on
electromechanical type earth fault relays. In some applications time/current
grading of earth fault relays may not be practicable unless earth fault
currents are limited or special CTs with a higher output are used.
Figure 5.4 illustrates the principle of earth fault protection. The
residual current component is extracted by connecting the line CTs in
parallel and having the earth fault relay connected in the neutral return
path of the CTs.
IA
CTs
A
ia CT = Current transformer
IB E/F = Earth fault relay
B
ib
IC
C
ic
if
E/F
FIGURE 5.4 Principal of earth fault protection
Figure 5.5 is basically the extension of Fig. 5.4. Here the phase
overcurrent elements are connected in the individual phases and the earth
fault relay is connected between the star point of the relay group and the
neutral of the CTs. A current will flow through the relay winding only
when a fault involving earth occurs.
It is normal practice to use overcurrent relays only on two phases as

these will detect any interphase fault; the connections to the earth fault
relay are unaffected by this consideration (Fig. 5.6). This is done purely
from an economical point of view. If all the CTs were ideal then under
normal operating and interphase fault conditions no current would flow
through the earth fault relay. However, with commercial CTs, due to the
difference in errors and amount of residual magnetism, some current may
flow through the relay. The magnitude of such unbalance or false residual
current is usually in the order of 0.01 to 0.1 Ampere at rated primary current.
It could be higher for heavy phase-fault currents.
CTs
CT = Current transformer
O/C = Overcurrent relay
E/F = Earth fault relay
O/C
E/F
FIGURE 5.5 Extension of Figure 5.4
O/C = Overcurrent relay

E/F = Earth fault relay
O/C O/C
E/F
FIGURE 5.6 Connection of overcurrent relays only on two phases for

phase fault and on neutral for ground faults
The grading of earth fault relays is normally limited to one system

voltage (due to the use of ∆/Y step down transformers), as an earth fault
in one section (Y) may not cause earth fault current to flow in the other
section (∆). Therefore, the earth fault on the Y side of the transformer will
not be seen by the earth fault relays on the ∆ side of the transformer, hence
the grading between earth relays on different voltage systems may not be
required (grading would be required if Y/Y transformers with neutrals
earthed are involved). However grading between the phase fault relays
on the ∆ side of the transformer and the earth fault relays on the Y side of
the transformer will be necessary, since phase currents are present in the ∆
side of the transformer.
CHAPTER
Fuses
6.1 INTRODUCTION
Fuses are the best known electrical devices and most of us at sometime or
the other have been made aware of their presence when one must be
replaced because it has blown/operated. They are basically simple and
relatively cheap devices although their behaviour is much more complex
and quite deceiving. Fuses perform the important duty of protecting the
equipment cheaply, efficiently and quickly and ensure that the effects of
faults are limited and that the continuity of supply is retained at all times.
Although the construction of fuses are not all that complex, but their design
and construction must be carried out with utmost care and precision so
that its reliability of operation is ensured.
The International Electrotechnical Commission (IEC) defines fuses
as “a switching device that, by the melting of one or more of its specially
designed and proportioned components, opens the circuit in which it is
inserted and breaks the current when this exceeds a given value for a given
time.” Australian Standards 1033 and 1034 cover the high voltage range of
the fuses.
The fuse is said to comprise all the parts that form the complete
device. The complete device being: (a) the fuseholder, which carries the
base and carrier and (b) the fuselink. The fuselink is a device comprising
of a fuse element or several fuse elements connected in parallel enclosed
in a cartridge, usually filled with an arc extinguishing medium and
connected to terminations. The fuselink is the part of a fuse which requires
replacement after the fuse has operated/blown.
FUSES 127
The fuse is a weak link and hence it has advantages over circuit
breakers. Generally the circuit breaker interrupts the current one to five
cycles after initiation at a natural current zero, whereas the fuse can operate
faster at any time and at any point on the current wave depending on the
magnitude of the overcurrent and the fuse characteristic. As element of
fuse is smaller in cross sectional area than the cable it protects, hence it
reaches melting point prior to the cable (provided both cable and fuse
element are of the same type of material). As the current increases the
element melts fast. Unlike circuit breaker, on deterioration it has tendency
of still faster operation. Hence ensuring its ‘fail-safe’ nature at all times.
The operation of the fuses can be made tamper proof when it is in sealed
and the cartridge type makes it silent and non-flammable. Hence the sealed
cartridge type which have element of silver (say) can be non-deteriorating
and gives consistent and reliable performance. The biggest advantages fuses
have is the devices are much cheaper than the circuit breaker of similar
rating and breaking capacity. Also the maintenance cost is much lower.
The disadvantage of fuses is the replacement time, although modern
cartridge fuses can easily be replaced, but the time to detect the faulty fuse
may be an added liability. Also, it has no switching ability at normal currents
and must be replaced after each fault operation. Hence if the fault in the
system occurs quite frequently, its replacement cost may make the device
quite expensive. However, a fuse-switch combination may overcome the
poor protection performance against small overcurrent.
6.1.1 Basic Definitions
The minimum fusing current is the minimum current at which a fuse
element will melt, which is the asymptotic value of current shown by the
time/current characteristic. (refer Fig. 6.30)
A conventional current rating is normally used under classification
of fuses which are reproducible under standard conditions.
The service rating is used for special type of fuselink which are to
perform unusual duties. These types of fuses must be used with the advice
of the manufacturer.
The current rating is the rating less than the minimum fusing current,
stated by the manufacturer. This is the current that the fuselink will carry
continuously without deterioration.
Fusion factor is the ratio of minimum fusing current to that of the
current rating.
6.2 CATEGORIES OF FUSES

Fuses can be classified into three categories:
(i) high voltage
(ii) low voltage
(iii) miniature.
The distinction between the high voltage and the low voltage occurs
at 1 kV and the miniature designation is associated with physical
dimension.
Fuses are produced in four constructional forms:
(a) cartridge
(b) semi-enclosed
(c) liquid
(d) expulsion.
6.2.1 Cartridge Type of Fuses
The cartridge type of fuse can be of all the three categories:
(a) low voltage cartridge fuse
(b) high voltage cartridge fuse
(c) miniature type.
A typical construction of the low voltage type is shown in Fig. 6.1.
The powder filled cartridge category of fuses have the most advanced type
of fuselink. Normally, fillers like sand which is pure and free of iron or
quartz is used and both are functions of silicon dioxide. When the fuse
blows, it creates a tube of melted sand around it which withdraws energy
from the arc and extinguishes it. If fine or coarse filler is used it causes
excessive pressure, however, intermediate grain produces optimum
cooling. It is desired to reduce the volume of metal in an element as a
result reducing the pressure on the cartridge.
FIGURE 6.1 Low voltage cartridge fuse
The advantages of filler is that heat is conducted away from the

element more rapidly than in air. Hence smaller element can melt at a larger
FUSES 129
minimum fusing current. Thus a thinner wire may be used for a given
current rating. When the wire is flattered into a tape the heat dissipation is
much faster. If the width at sectors along its length is increased, the heat
dissipation from the constriction is still faster. By such means the cross-
sectional area of the constrictions can be considerably reduced for a given
rated current which makes the fuse operate much faster than on a uniform
element as shown in Fig. 6.2.
Closing angle g
Supply
Voltage zero voltage
Cut-off current
Prospective current
5,000 A 22x
5000 A Current
Approx. Recovery
1000 V voltage
Volts across fuse terminals
5,000 kW Sand melted is proportional

to this area
Power dissipated in fuse cartridge
2
Total i t
Pre-arcing
2 (limits heating of ò i2dt
it conductors etc.)
Actual pre-arcing Joule integral
time
Actual arcing time

Element heating up to melting point
at constrictions during pre-arcing time
FIGURE 6.2 Representation of a sand-filled cartridge fuse with typical oscillogram

for the operation on short circuit (Approx. 100 A rating)
The dissipation of heat is further improved by using a number of

thinner elements in parallel instead of one thicker element. This gives
further advantages, firstly, the arc is distributed uniformly and secondly,
it improves performance of fuses on small over current. However, parallel
element may/may not blow out simultaneously.
Melting time of fuse element may be further adjusted by means of

special techniques of time delay.
1. M-effect (Metcalf or Melting point effect): This method can be
achieved by the addition of low melting point alloy to the surface,
where it will dissolve the element material when it melts. This
has the effect of opening the element at the point where the
dissolution takes place at a chosen point along the length of the
element. As this process takes longer time to complete than the
melting of the unmodified element of smaller cross-section, in
spite of the fusing current to be the same in both cross-section,
the use of M-effect can produce a time delay in the small to
medium over current region and thus prevent unnecessary
blowing of fuses by surges of current in normal service (Fig. 6.3).
2. Heat sink effect: In this process by inserting a large fusible metal
in the element a more extended time delay can be obtained.
Because of the large thermal capacity of the insert it acts as a
“heat sink”. Larger the volume of the fusible metal, longer is the
time taken to melt. Hence longer the time delay.
Tape
Wire
Low melting
point high thermal
capacity insert.
Melts on small
overcurrent
if sustained
Dual element
time tag
M-effect additions
(fuse on small overcurrent if sustained)
Region which melts on high overcurrent
FIGURE 6.3 Techniques of time delay on a section of a fuse element
Besides the time delay provision, both the above methods have other
advantages. It reduces the temperature at which the element opens the
circuit. Therefore, it reduces the temperature of the ‘blown’ cartridge
immediately after clearing a fault at small overcurrent. Under high over
current conditions, the fuse will melt at the most constricted points and by
designing these carefully and manufacturing them to high precision, the
pre-arcing time can be controlled.
FUSES 131
The high voltage powder-filled cartridge type (rating >1 kV) is of

current limiting type and uses similar basic principles as to its low voltage
counter part. However, it is further refined to produce special characteristic
and additional features. The length of fuse element and the number of
constrictions is roughly proportional to its voltage rating. Therefore, fuse
may tend to become quite long and difficult to use. Thus it is wound in the
form of a helix on a star core in a barrel, as shown in Fig. 6.4. However,
such construction requires great care, precision and careful design in order
to ensure that the arc from one turn does not merge with the arcs from the
next turn or sustained arcing may cause entire destruction. During
productions manufacturers perform non-destructive testing. If used within
prescribed rating and breaking capacity it can be a safe and reliable product.
Striker coil Striker assembly
Outer cap Starcore Silver elements
Body Sand filter
FIGURE 6.4 Construction of typical hv fuselink
The miniature cartridge type of fuselink may be of two types (i) filled
and (ii) unfilled type. They are basically small in size and used for the
protection of electric appliances, electronic equipment and component
parts, normally intended for use indoors. Many such varieties are available
as shown in Fig. 6.5. Powder is used as filler for the filled type and it has
high breaking capacity (in order of 1500 A). However, in the unfilled type
the element is supported in air inside the glass cartridge. It has low breaking
capacity (in order of 35 A). It is also possible to enclose huge variety of
element shapes, time delay springs, thermal sinks etc. in the glass cartridge.
The categories of speed of operation are signified by internationally
accepted letters that are marked on the fuselink bodies, viz.,
FF = super quick acting
F = quick acting
M = medium time lag
T = time lag or anti surge
TT = super time lag.
FIGURE 6.5 Selection of miniature fuselink
Fig. 6.6 provides an illustration of the different operating speeds of

various miniature categories.
1000
100
10 FF F M T TT
Time
0.1
0.01
1 2 3 4 5 6 7 8 910
Current
FIGURE 6.6 Time/current characteristics of miniature fuselinks

FUSES 133
The quick acting type of fuselinks are F and FF. The F type have
single wire construction e.g., silver, copper and its alloy, nickel and nickel
chromium. The FF type is similar to the F type except that the elements
have a restricted section in them. The time lagging types are M, T and TT.
These are all of low-breaking capacity type and have glass bodies and of
the unfilled type.
6.2.2 The Semi-Enclosed Type of Fuse
Fig. 6.7 illustrates the semi-enclosed type of fuse. This type is very widely
used and they are in the low voltage category. The re-wirable fuse consists
of a base, a carrier, the fuse element and an arc resistant tube (to limit the
expulsion of flame). The element material is tinned copper, which has a
melting point of 1083°C, but it cannot be run for any length of time at
temperatures approaching the melting point and rapid oxidation takes
place above 250°C. Hence, the semi-enclosed fuse requires a large
overcurrent to blow it. The fusing factor is about 1.75. The performance
depends on how the fuse is wired (may/may not touch the tube sides)
and on the state of the tube. Care should also be taken as to ensure that it
is free of kink and the correct size of fuse wire should be used.
Fuse carrier Fixed
terminal Fixed contact
Fuse
element
Fuse contact Arc resistant

tube Fuse base Cable socket
FIGURE 6.7 Semi-enclosed fuse
A lower fusing factor with reduced time-lag can be achieved by using

a wire of silver, which is used in many cartridge fuses. The higher
temperature at which the wire would have to run might, however, cause
over-heating at rated current.
There may be some form of emission of flame from the tube, in
particular with larger fault currents, however, careful screening is necessary
to avoid flashover to adjacent metal work or to other fuses.
When a number of fuse wires are used in parallel in a re-wireable
fuse, the resultant rating is less than the arithmetical sum of ratings of the
individual wires. Its effective value is influenced by the type of design of
fuse carrier being used, hence this practice is discouraged.
6.2.3 European Categorisation of Semi-Enclosed Fuses

The three systems used are described in terms of contact arrangements of
the fuselinks and these are
(a) Blade contact — NH type
(b) End contact/screw type — D type
(c) Cylindrical cap contact — B type
The blade contact type referred to as NH fuses, where NH is the
abbreviation of Niederspannungs Hochleitungs which is German for low
voltage with high breaking capacity. They find applications in factory
distribution systems and also in distribution cabinets of the electric supply
industry. Fig. 6.8 shows a range of such type of fuses. The fuse elements
are generally made of copper strip and the body is made of ceramic but in
recent years high-temperature thermosetting plastic materials have gained
popularity. Their bodies have a rectangular outside cross-section with a
circular longitudinal hole through them, and end plates complete with the
blade contacts are attached to the body with screws. To allow the fuselinks
to be mounted in close proximity to each other, even in the absence of
insulating separators, the end plates are normally confined within the
outside dimension of the fuselink body. The blade contact surface are
usually silver plated to assist in obtaining low-resistance connections even
when the forces applied by the spring contacts into which they fit are
relatively low. Some fuselinks are provided with cylindrical bodies and
these are allowed in the standard specification provided they meet the
dimensional requirements.
FIGURE 6.8 NH fuselinks

FUSES 135
FIGURE 6.9 NH feeder pillar showing use of handles
FIGURE 6.10 NH fuse switch

The NH fuselinks are inserted into their fuselinks by a detachable

handle which is made of plastic, a particular example is shown in Fig. 6.9.
Alternatively, in a widely used simple design of fuse-switch, the cover of
the switch acts as the fuse handle, the fuse links replacing the normal switch
blades and being withdrawn when the cover is opened or removed. This
arrangement is shown in Fig. 6.10.
The end contact or screw-type fuses
are a very old fuse system. The abbreviation
D standing for Diazed type. It is also called
‘bottle’ type due to the shape of the fuselinks
(Fig. 6.11). The ‘D’ type also indicates that
they are for domestic use. They have reached
wide acceptance and it is being continuously
used as replacement. Even new domestic FIGURE 6.11 ‘D’ type fuselink
installations tend to use such types of fuses.
The fuselink contain strip elements of copper or silver plated copper
and are filled with granular quartz. The bodies are made of ceramic
material. Each fuselink is fitted at the ends with cylindrical contacts made
of brass, usually nickel-plated and are often of two different diameters.
They are usually provided with grooves at ends to ensure good contacts
when they are fitted in the carrier. The
fuselinks are fitted with indicators which have
a button head form and is pushed out through
the end contact by a weak spring when the
fine wire device in parallel with the main X
Plan 'xx'
X
element melts and no longer provides
restraint. The button head is visible through
glass window in the fuse socket cap. Fig. 6.12
shows a standard holder. A range of gauge
rings with various internal diameters and
coloured ends to indicate the maximum
ratings of fuselinks which will fit into them, is
available. The appropriate ring is placed into
a fuse socket to ensure that a fuselink of too
great a rating for the circuit being protected
may not be installed. The fuselink is inserted
before the fuse socket cap is screwed to the
fuse socket which produces forces between the
fuselink end contacts and the spring contacts
FIGURE 6.12 Holder for
in the fuse socket and fuse socket cap.
‘D’ type fuselink
FUSES 137
The cylindrical cap contact or ‘B’ type fuses are widely used in France
for both domestic and industrial applications. The fuselinks have copper
wire or wire elements and are filled with quartz and have ceramic bodies.
These fuselinks are available with visual operational indicators, if required.
For domestic applications, the fuselinks are produced in a range of
diameters and lengths, each rating having its own unique dimensions to
prevent incorrect replacement after operation.
Fuse switch units having ‘B’ type fuselinks are available for industrial
use. Strikers, operating in a similar way to operation indicators may be
incorporated in these fuselinks. When a fuselink melts, the striker moves
out through the end cap and actuates a micro-switch which may initiate
an alarm.
6.2.4 North American Fuses
The various fuses are divided into classes, the main one being:
Class K — up to 600 A, 250 V and 600 V
Class H — up to 600 A, 250 V and 600 V
Class J — up to 600 A, 600 V
Class L — above 600 A and up to 6000 A, 600 V.
Fuselinks with current ratings up to 60 A are generally fitted with
cylindrical end contacts as illustrated in Fig. 6.13. Fuselinks for higher
ratings are provided with either blade-type terminations for mounting in
spring contact or blade terminations containing holes or slots which allow
bolted connections to be made (Fig. 6.14).
FIGURE 6.13 North American cylindrical fuses

FIGURE 6.14 North American tag type fuses
The Class K are cartridge fuselinks having a high breaking capacity

typically about 200 kA. The fuselinks are made of organic materials viz.,
fibre to glass cloth impregnated with resin. The Class K is the most popular
industrial fuse in USA. The fuselink operates slowly at currents in a range
above the minimum fusing level but clears rapidly when carrying large
currents. The way this is achieved is by splitting the fuse element into
three different sections. The two outer sections use conventional strip and
are surrounded by arc-quenching material such as quartz or calcium
sulphate. It is these sections which operate when high short-circuit currents
flow. The elements are generally made of copper but silver is sometimes
used. The centre section is unusual as it has two metal portions of large
thermal mass connected together with low melting-point solder (Fig. 6.15).
Granular quartz Body
Short circuit Thermal cut-out End cap

elements element
FIGURE 6.15 Dual-element fuselink
At currents above the minimum fusing level the solder melts, its
temperature rise being slowed by the adjacent masses of the metal. To
provide an adequate break, the two metal parts are pulled away from each
other by springs when the solder melts. The centre section does not contain
FUSES 139
filling material as this would prevent movement of the metal parts and it
is not required for arc-quenching purposes at that current levels. Other
types of Class K fuselink use the M-effect principle.
The Class H fuselinks have a low breaking capacity (about 10 kA).
They employ cylindrical fuse bodies in the high-breaking capacity fuselinks
and the connection arrangements and dimensions are also interchangeable.
There are two types of Class H fuselinks being non renewable and
renewable. The former uses copper elements and fillers whereas the latter
do not contain filling and use replaceable zinc elements with restricted
section. The end cap of the renewable type are removable (Fig. 6.16) so as
to allow the elements to be replaced when necessary.
Screw on end cap
Element
Inner cap Body
FIGURE 6.16 Renewable fuselinks
Both the Class K and H fuselinks are mounted in simple unshrouded

bases as illustrated in Fig. 6.17. As both classes are interchangeable, care
should be taken to ensure that the two fuselinks are never interchanged as
one has a lower breaking capacity than the other. Hence in order to
distinguish between them, Class K type of fuse have recently been replaced
FIGURE 6.17 North American fuse bases

by Class R fuselink. The Class R type of fuselink is shown in Fig. 6.18 and
they contain features which prevents the insertion of Class H or Class K
fuselinks.
FIGURE 6.18 Rejection features in class ‘R’ fuselinks
The Class J fuselinks are more compact than both the Classes K and
H. They have a breaking capacity of 200 kA and a high speed operation
under higher fault levels. Recently Class J has been replaced by Class T
fuselinks which has smaller dimensions.
The Class L fuselinks are used for all applications for current ratings
above 600 A. They have a breaking capacity of 200 kA.
In North America for domestic, commercial and light industrial
distribution applications Class G fuses are available. They have a rating of
300 V up to 60 A with a breaking capacity of 100 kA. In domestic premises,
the fuses are incorporated in the sockets from which the supplies are
obtained. As the fuselinks are screwed or plugged into their holders they
are known as ‘plug’ fuses. They generally have ratings up to 36 A and
operate at 125 V. Three types of such plug fuses are used:
(i) the ordinary
(ii) the dual type
(iii) the S type.
The ordinary plug type is as shown in Fig. 6.19 and contains an
element either in strip or wire form. The element can be seen through a
mica window set in the end of the body. The body has no filling material
and is made of glass. It has a metal contact in one end and a threaded brass
section around it which acts as a second contact and a means for screwing
the fuselinks into its base or holder. The fuselink can be replaced after
operation but there are some which allow replacement of the element.
Mica cover
Edison screw base
Contact
Fuse element
FIGURE 6.19 Plug fuse

FUSES 141
The dual element plug fuse has the same basic construction as the
ordinary plug fuse except that it contains two copper strips of which each
have one of their ends soldered to the other. A spring is incorporated to
separate the strips when the solder has melted as a result of the passage of
overloads. The operation is not produced by high transitory surge currents
but rapid clearance is obtained in the event of very high currents flowing
during the short circuits.
Both the plug type of fuse is fitted into standard Edison screw fuses
(Fig. 6.19). It is possible for users to make contact with live parts when the
fuselink is removed and a fuselink of any rating may be fitted into the
base.
The type S plug fuse is similar to the dual-element fuse but it is
designed for use with an adaptor which performs a role similar to the
gauge rings as mentioned earlier with the D type fuses. The adaptor has
an external screw thread which will mate with
the internal thread of a standard Edison screw Diaphragm
fuses and when fully secured it can be locked.
The adaptors are produced with several
different internal threads, one for each fuse
rating, hence the insertion of an incorrect Element
fuselink is prevented.
6.2.5 Liquid Type of Fuse
The liquid fuses are used for high voltage
application and utilises spring tension to assist
with arc extinction and to provide an adequate
physical break in the circuit to withstand the
Arc extinguishing
service voltage after operation. It consists of liquid
a toughened glass tube with metal end caps,
to the upper of which is secured by the short Spring
element (Fig. 6.20). This element has a high
tensile strain wire in parallel with the silver
fuse wire, as the latter is not strong enough to
withstand the tension of the spring securing Glass tube
the bottom of contact of the element to the

lower end cap. The wires are enclosed in a
fibre tube and the bottom contact of the
element also carries a conical liquid director.
The body of the fuse is filled with non
inflammable insulating liquid such as carbon FIGURE 6.20 Sectional view of
tetrachloride to a level just below the element. a liquid fuse
When the element melts, the bottom contact is retracted downwards

by the spring and the liquid director forces a stream of liquid into the arc
path. The vapour pressure generated de-ionises the arc path and is then
vented to the atmosphere by the rupture of a thin aluminium diaphragm
in the top-end cap. The element and liquid may be replaced after operation
and the fuse returned to service.
Liquid quenched fuses are in wide use on overhead line networks.
They are basically used outdoors and provision is made for removing from
and replacing the fuselinks into their mountings by pole operation from
the ground, a bayonet-fixing arrangement being utilised. They are mostly
used for the protection of 11 and 33 kV pole or pad-mounted transformers
on rural system and also for spurs feeding a number of transformers.
In new installations such types of fuses are not being recommended
and it is being replaced by the expulsion type of fuses. They have given
and are till now giving good/satisfactory service. However, problems have
been experienced with metal to glass seals which fail and cause leakage of
the liquid. A subsequent fault clearance results in the disintegration of the
tube.
6.2.6 Expulsion type Fuses
Expulsion fuses consists of a tube of insulating material into which the
fuse element is inserted, in some cases one end of the tube is closed and in
other both ends are open. (Fig. 6.21). When element melts and arcing takes
place the resultant gas pressure causes the arc to be blown out of the ends
of the tube and extinguish. In certain designs the process is assisted by
lining the interior of the tube with material such as boric acid, which
produce gas when heated by the arc. In order to accelerate the process of
arc extinction, the element is held under spring tension and when the
element melts the spring rapidly separates the two sections. The
disadvantages of the system is the clearing of low overcurrents. This is
due to small pressure generated but it is overcome by using a small inner
fibre tube to enclose within the main tube.
The operation of expulsion fuse is violent, in particular with large
fault currents and it is usually pole mounted out-of-doors, increased phase
spacing is employed to avoid flashovers. Walls of tube may be contaminated
by carbon and other arc products after blowing and in order to prevent
leakage along this path the tube is arranged to be isolated from the circuit
contacts after operation. This is achieved by utilising the spring normally
holding the element under tension effectively to shorten the length of the
fuse carrier when released. Also, it allows the carrier tube to disengage
FUSES 143
from the upper contact and to fall, under the influence of gravity, about
the lower hinged contact.
Screwed terminal
cap
Upper Eye for
contact operating rod
Element
Inner tube
S.R.B.P. tube
Standard
flexible
lead
Lower
contact
Pivot pin
Hinge pin
Spring loaded
quadrant
FIGURE 6.21 Expulsion fuse
The expulsion fuse is not manufactured for 415 V or other low

voltages. It is essentially high voltage fuse use on system up to 33 kV. It is
also used for the protection of overhead line network. Hence both fast and
slow blowing elements are available. Also a number of automatic reclosing
units have been designed, so arranged that if one blows it is automatically
replaced after an interval of a few seconds.
A cross-section through a typical fuselink is shown in Fig. 6.22(a)
and three popular physical connection arrangements are illustrated in
Fig. 6.22(b). These are designated:
(i) button head
(ii) double tail
(iii) universal.
Fibre Copper
sleeve sleeve Copper
Flex
sleeve
Solder Tin or copper

Washer
Strain element
tape Copper button
head rivet
(a) Sectional view of expulsion fuselinks
T-slow
T-slow
K-fast
(b) Terminations of expulsion fuselinks
FIGURE 6.22
It can be seen that all of them have braid at the lower end, the only
difference being the upper termination. The fuse carrier has pins at the
lower end which act as a hinge when it is mounted on the lower contact of
the fuse unit. In the service position the fuse carrier is tilted from the vertical
as seen in Fig. 6.23.
The main advantage of expulsion fuses is that they are cheap and in
most cases can be reused by fitting replacement links. In addition, as they
interrupt at current zero, they do not chop and do not produce excessive
switching voltages.
Their disadvantages are that they do not have the high breaking
capacities of powder filled fuses and are not current limiting.
FUSES 145
Terminal
Mounting Expendable cap

bracket (often cap only)
Fuse element
Fusible section
(enclosing tube
often omitted)
Fuse link
(includes fusible
section and tails)
Fuse base Fuse-carrier Fuse-carrier
Fuse-base contact
contact
FIGURE 6.23 Schematic arrangement of an expulsion fuse
In some applications they have been known to cause fires due to the
expulsion of incandescent particles during operation.
6.3 FUSE OPERATING OSCILLOGRAMS

Typical oscillograms of operation of fuses under small overcurrent and
large overcurrent conditions are shown in Fig. 6.24.
Prospective
Current passed by fuse Cut-off current current
substantially equal to
prospective current
Current zero
Peak arc
Relatively long voltage
Recovery
Pre-arcing time
Voltage zero Arc voltage voltage Recovery
voltage
Voltage zero
a b c
b a c
d.c.
d.c.
a—Initiation of current; b—Initiation of arcing; c—Final clearance
(a)
Current passed by fuse Cut-off current

substantially equal to
prospective current Prospective current
Current zero
Current zero
Peak arc
voltage
Voltage zero Voltage zero Recovery

voltage
Relatively long
Arc
Pre-arcing time a b c
voltage Recovery
a b c votage
a.c. a.c.
(b)
(a) Small overcurrent; (b) Large overcurrent
FIGURE 6.24 Typical oscillograms of fuse operation
6.3.1 Peak Arc Voltage

It was a characteristic of certain early types of filled cartridge fuse that the
voltage rise when clearing large fault currents was considerable, most fuse
specifications now prescribe a limit to this voltage which, if exceeded,
constitutes a disqualification.
A simple wire fuse breaks up into globules along its length, with an
arc between each globule and the next, thus producing a chain of arcs,
which add up to produce a voltage proportional to the length of the element
consumed and which may rise to several kilovolts.
Modern fuses employ elements with constrictions. This results
effectually in a number of short fuse elements in series, each so short that
the overvoltage is limited. Problems do not arise until the fuse is operated
very close to its breaking capacity and above its rated voltage, where the
excessive burn-back of the element can produce conditions leading to very
high voltages, as the wider parts of the element break up into molten
globules. This is seen to be a further reason why fuses should never be
used in a circuit above their rated voltage.
6.3.2 Fulgurite (Roping)
These are terms normally used for the residues of sand in the fuse which
were melted by the arc during clearance. These residues give much evidence
of the processes which caused the fuse to blow. For example, the arc energy
in the fuse may be calculated by weighing the fulgurite. If the weight of
the fulgurite is w, then the arc energy was
E = 2100 w Joules (w in grams)
FUSES 147
It is also possible to tell whether the fuse operated on a small or a

large overcurrent by breaking open the blown fuse and studying the
fulgurite. (Fig. 6.25.)
Before
After
Large overcurrent Small overcurrent
FIGURE 6.25 Fulgurite structure-fuse without M-effect (with M-effect the fuse will
operate at the M-effect on small overcurrent)
Peak arc voltage is dependent upon the number of constrictions in

an element, because of the arcs in series. This gives a minimum value of
peak arc voltage irrespective of the applied voltage up to a certain point.
When this point is exceeded, the extra applied voltage can force the arcing
to persist and produce burn-back and other effects which may increase
with each incremental voltage, thus causing a larger peak arc voltage.
This is illustrated in Fig. 6.26.
Maximum arc voltage
1000 V
500 V
250 V Voltage rating
Applied voltage
FIGURE 6.26 Arc voltage characteristic
It is, therefore, clear that a fuse of higher voltage rating should not
be used to replace a blown fuse of lower voltage rating unless due
cognisance is taken of the fact that its peak arc voltage will be greater. Peak
arc voltage must not exceed the dielectric withstand of the system in which
the fuse is placed.
Fuses for 11 kV use are frequently designed to produce low arc
voltages, in order that they may also be used on 6.6 kV systems. It should
not be assumed that this is the case without first consulting the
manufacturer on this point.
6.4 TIME vs CURRENT CHARACTERISTIC

Fuses operate by resistance heating of the fuse element. The heat dissipated
t
in time t is equal to z
0
i 2 R dt , where R is the resistance of the fuse element.
Further, at sometime t1 the element will reach the boiling point of the
t1
material and for large currents the total heat generated will be z
0
i 2 R dt
from this it can be deduced that, for a given element, under adiabatic
conditions, the total amount of heating required to fuse the element can be
t1
precisely determined, i.e., z

0
i 2 R dt is a constant. This quantity is known as
the Joule-integral and is denoted by I2t, where t is the virtual time defined
as the value I2t divided by the square of the breaking current.
Joule-integral can be used to accurately predict the cut off current
for a given prospective short-circuit current.
The pre-arcing I 2t of a fuse element is directly proportional to the
square of its smallest cross-sectional area. Consider the section of a fuse
element shown in Fig. 6.27 of length l and cross-sectional area A; carrying
a current i.
Mass m
Temperature q°C
Specific Heat s joules/gm
Current i Resistivity r0 ohm mm at 0°C
Temperature coefficient of
I resistance = a per degree
Resistance R centigrade
FIGURE 6.27 Section of fuse element heated by current
The heat produced in time dt is

i 2 ρ0 (I + αθ)l
i2R dt = dt
A
where
ρ0 is the resistivity in ohm-mm at 0°C
α is the temperature coefficient of resistance in per °C
σ is the specific heat in joules/gram
θ is the temperature in °C
FUSES 149
As this quantity of heat will raise the temperature by dθ if no heat is

lost to surroundings and from the knowledge of thermodynamics this
quantity of heat is the product of mass, specific heat and the temperature
i2 R dt = m σ dθ
Also by definition density is the ratio of mass to volume.
Hence, m=DAl
3
where D = density (kg/m )
and m = mass (kg)
A × l = volume (m3)
i 2 ρ 0 (I + αθ)l
dt = DAl.σ dθ
A
Dσ
or i2 dt = A2 dθ
ρ0 α (θ + 1 / α)
On integrating both sides
Dσ F
(θ 2 + 1/α ) I
z i 2 dt = A
2
ρ0 α
ln GH
(θ 1 + 1/α ) JK
where θ2 = melting point of the element
and θ1 = 20 °C
z i 2 dt = KA2
where K is the constant of the metal and can be directly calculated from
the known values.
For typical metal like silver and copper, normally used as fuse
elements, the pre-arcing I2t is 6.6 × 106 A2 (Amp2 sec) and 9 × 104 A2
(Amp2 sec ) respectively, where A is the cross-sectional area of the conductor
at narrowest point in mm2. Silver is very commonly used as a fuse element
due to its non-deteriorating, low oxidation properties. It also has good
M-effect with tin and its alloy.
The usefulness of virtual time is shown in Fig. 6.28 which represents
a typical fuse. It will be seen that the time less than about 5 × 10–1 s, there
is a spread in the normal current-time curve (shown shaded) and with the
larger currents there is considerable variation in the time, the ratio of the
largest to the smallest being about 5. There is however, no spread in the
curves of virtual time, and discrimination will be achieved between two
fuses provided that the curve of virtual total operating time of the minor
fuse lines below the curve of virtual pre-arcing time of the major fuse.
Hence it is common practise for checking the discrimination between fuses

to use I2t characteristics.
1000
500
100
50
10
5
1
–1
5 × 10
Time (sec)
–1
Limits of actual
10 pre-arcing time
–2
5 × 10
–2
10
–3
5 × 10
–3 Virtual pre-arcing
10 time
–4
5 × 10
Virtual total
operating time
–4
10
–5
5 × 10
–5
10
50
100
200
300
400
500
1000
2000
3000
4000
5000
10000
20000
30000
40000
50000
Prospective symetrical r.m.s. current (amp)
FIGURE 6.28 Vertical time curves
I2t characteristics are extremely useful to protection engineers who

require rapid assessment of the degree of protection against short-circuits
in a fuse protected network. The total operating time is greater than the
time shown on the fore-arcing time/current characteristic, because the fuse
element continues to arc between the melted ends until the current is
reduced to zero. Hence
Total operating time = Pre-arcing time + Arcing time
The pre-arcing time is the time between commencement of a current
large enough to cause melting of the fuse element and the instant when
the arcing is initiated. Total operating times cannot be shown with accuracy
FUSES 151
since the arcing time varies with the power factor and transient
characteristic of the circuit, the supply voltage, the electrical angle at which
arcing commences and other factors. When operated within the voltage
limits prescribed by the manufacturer, the arcing time is only significant
for large overcurrents, where there are short operating times.
The current usually plotted on the time/current characteristic is the
prospective current i.e., the current which would flow in the test circuit if
the fuse were replaced by a link of negligible impedance. The time for
short-circuit operation is usually the virtual time, because of the variation
of actual time with the point on the voltage wave and other variables. On
small over currents, however, where the pre-arcing time is long and the
arcing time negligible by comparison, the times shown can be taken as the
total operating times at the currents shown.
A fuselink which is to protect Supply Connecting
a particular piece of equipment or cable A cable
circuit should ideally satisfy a Item of
number of criteria. This is illustrated B
equipment
by considering a simple example

based on the circuit shown in FIGURE 6.29 Circuit protected by a fuse
Fig. 6.29.
Firstly, the minimum fusing current of the fuse should be slightly
below the current which the cables and item of equipment are able to carry
continuously.
The item of equipment will usually be able to carry overload currents
for limited periods, and the fuse should operate at these current levels in
times slightly shorter than the corresponding equipment time ratings.
Clearly, the cables should also be able to cope with this duty without
suffering damage.
Higher currents may flow as a result of faults within the item of
equipment and in these circumstances the primary requirement is that
consequential damage to the remainder of the circuit should be prevented.
The extreme case will occur in the event of a short-circuit between the line
and neutral terminals of the equipment. Clearance must then be effected
before damage is caused to the cables.
A further possibility is a short-circuit between the conductors of the
connecting cable. The most severe situation would arise if the fault was at
the input end, i.e., between points A and B and in these circumstances the
fuse would have to interrupt the circuit before the source and supply cables
could suffer damage.
To achieve the criteria, fuselinks should have time/current

characteristics which lie close to the withstand curves of their associated
circuits as shown in Fig. 6.30. Clearly the operating time of the fuse should
always be less, at any current level, than the period for which the circuit
can withstand the condition. This seems to be self-evident but in practice
there are a number of less obvious limitations and factors which must be
taken into account.
Time, (sec)
Withstand curve of
protected circuit
Fuselink characteristic
Current
FIGURE 6.30 Time/current characteristics of circuit and fuse
A very important factor which must be recognised is that the

operating times referred to above are the total clearance times, i.e., the sum
of the pre-arcing and arcing times. Now ideally, a fuselink should be capable
of carrying a current just below the minimum fusing level indefinitely and
also of carrying a current just below the minimum fusing level indefinitely
and also of carrying any higher current for just less than the corresponding
operating time and thereafter be in its original condition. In practice, these
conditions cannot be achieved because there are changes of state before
operation occurs. Once arcing has commenced it is clearly impossible for
an element to return to its original form and even in the molten state, an
element may distort and not return to its initial shape on cooling.
Furthermore, unacceptable changes may occur in fuses with low melting
point materials on the elements, if overcurrents continue long enough to
initiate the ‘M’ effect diffusion process.
There is thus a head band below the time/current characteristic in
which a fuse should, if possible, not be called upon to perform. The
likelihood of high current flowing through a fuselink for just less than its
operating time is small, but if it should happen, the probable result, which
must be accepted, is that the fuse will operate more quickly than expected
on a future occasion. Difficulties would certainly arise if a fuselink carried
currents near its minimum fusing level for long periods, and, to avoid this
situation, a fuselink is assigned a rated current somewhat below the
FUSES 153
minimum fusing value. The ratio of the minimum fusing current to the
rated value, which is defined as the fusing factor, usually has values in the
range 1.2–2. The significance of this factor is that protected circuits must
be able to operate continuously at levels appreciably above the rated current
of the fuse in the first criterion above, namely that the circuit continuous
rating must exceed the minimum fusing current is to be satisfied. This is
an uneconomic situation which arises with many protective arrangements,
because it is necessary to have current settings above the full-load value of
the circuit unless discrimination is achieved by differential methods. It is
clear that it is particularly desirable to use fuselinks of low fusing factor
when the cost of the protected circuit and equipment rises significantly
with its current-carrying capacity. Another factor which must be borne in
mind is that many fuselinks do not provide full-range protection, i.e., they
will not operate satisfactorily, at all current levels from their rated breaking
capacities down to the minimum fusing values. As stated earlier,
satisfactory arc extinction may not be achieved in some fuselinks at
relatively low over currents. Care must always be taken to see that such
fuselinks are only used in applications where currents of these magnitudes
will not be experienced or, if this cannot be guaranteed, then an associated
protective scheme must be provided to interrupt these currents before the
fuse can operate.
6.4.1 Published Time/Current Characteristics

The data required for producing these characteristics are obtained by testing
fuselinks which are at ambient temperature (15–20 °C) when current flow
through them is initiated. The curves published by the fuse manufacturers
usually show the relationship of the virtual pre-arcing time to prospective
current.
It will be appreciated that the effects of factors such as current
limiting, the instant of fault occurrence and the current wave shapes do
not significantly affect the performance at the lower current levels where
the operating times are long and of course, the arcing periods are negligible
compared with the pre-arcing times. As a consequence, the variations in
operating times are insignificant at these levels and there is therefore only
one time/current curve for each fuselink.
In practice the passage of load current prior to the overcurrent
condition of a high temperature environment causes operation to be slightly
faster than shown on the characteristic. There is also an associated reduction
of the times for which the protected circuit can withstand given overcurrents
if it is in an environment of relatively high temperature or if load currents
have been flowing in it. The two effects will not usually be equal but they
nevertheless reduce the significance of this factor.
Pre-arcing time, (sec)
5000
1000
100
10
100A
40A
1 16A
0.1
0.01
10 100 1000 10,000 50,000
Prospective current in amps
[r.m.s. Symmetrical]
FIGURE 6.31 Time/current characteristics
It is common to indicate the approximate minimum-breaking current

on the characteristics by using a full line above this level and a dotted line
below it. This is illustrated in the typical characteristics shown in Fig. 6.31.
6.4.2 Cut-Off Characteristics
These show the highest possible instantaneous values of current which a
given current-limiting fuselink will pass under fault conditions for varying
values of prospective current. They are of use in calculating the peak
mechanical forces which the equipment in protected circuits must be able
to withstand.
A typical characteristic is illustrated in Fig. 6.32.
t
ui
irc
tc
or
Peak current
sh
Characteristics
al
ric
for different
et
m
rated currents
ym
As
Prospective currents r.m.s
FIGURE 6.32 Cut-off characteristic

FUSES 155
6.4.3 Operating Frequency

The characteristics of fuselinks normally relate to operation at frequencies
of 50 or 60 Hz and there is little difference between the behaviours obtained
at these frequencies. Certainly a fuselink tested at a frequency of 50 Hz
will be entirely suitable for use at 60 Hz. Even higher power frequencies
would present little difficulty for the fuselink. Lower frequencies do,
however, need careful consideration since extension of the duration of the
half cycles of the source voltage wave causes the total operating times at
very high current levels to be increased. As a result higher arc energies are
released in the fuselink. For frequencies below 50 Hz some de-rating in
terms of rated operating voltage is therefore necessary. In the extreme case
of dc applications the voltage rating may be only half that allowable at a
frequency of 50 Hz and high values of circuit inductance may necessitate
further voltage de-rating.
6.5 DISCRIMINATION
Most circuits contain several protective devices and some of these are
effectively in series. They must all be coordinated so that correct
discrimination is achieved under all fault conditions and only the minimum
of interruption should occur to clear any fault condition.
6.5.1 Discrimination between Fuse Links
In domestic, and in many industrial installations, it is usual for the supply
authority to provide the main fuses, and consumers’ branch circuits are
protected by fuses of smaller rating. In the event of a fault in a branch
circuit the branch-circuit fuse should blow but not the supply authority’s
fuse, that is to say there should be discrimination. It is usual to call the
fuses the major fuse and the minor fuse, respectively. If there were no arcing,
and the fuses were of the same design, the minor fuse could be 90% of the
rating of the major fuse and there would still be perfect discrimination. In
practice, however, it is possible for a fuse to pass more I2t during arcing
than is passed during melting (termed more correctly, the pre-arcing
period), especially with large overcurrents in inductive circuits. If the major
fuse is to remain intact its pre-arcing I2t must not be exceeded: Hence to
achieve discrimination the pre-arcing plus the arcing I2t of the minor fuse
must not exceed, and preferably should be less than, the pre-arcing I2t of
the major fuse. This usually means that the rating of the major should be
not less than twice that of the minor fuse.
For convenience in ensuring discrimination between fuses,
manufacturers publish I2t characteristics similar to those shown in Fig. 6.33.
In these curves the minimum pre-arcing I2t is measured for every fuselink
at high prospective current, and the maximum total let-through I2t, pre-
arcing and arcing, will often occur at the condition of maximum arc energy.
To select a minor fuselink which will discriminate with a major
fuselink under all conditions, following two examples show the use of
these curves.
7
10
Total operating
2
I t (at 415 V)
6
10
Total operating
2
I t (at 240 V)
5
10
I t (Amp sec)
2
2
4
10
2
Pre-arcing I t
3
10
2
10
12
32
25
125
315
10
16
20
40
50
80
100
160
200
250
400
500
630
800
63
Rated current (A)
FIGURE 6.33 Typical I2t characteristic
Example 1. Let a 160 A fuselink be the major fuselink in a system. It

can be seen from the curves (Fig. 6.33) that the minimum pre-arcing I2t of
this fuselink is 8 × 104 A2s. In the example, this is seen to be equal to the
let-through I2t of a 100 A fuselink at 415 V. Theoretically this would be just
FUSES 157
too large a current rating for the minor fuse, because the element of the
major fuse would just melt with the maximum let-through of the minor
fuse. In practice, discrimination would probably not be lost, because the
condition of maximum arc energy is very unlikely. To be sure that no
deterioration of the major fuse can take place, the 2 : 1 ratio is to be preferred.
We see that for the 80 A fuselink the I2t let-through is only 4.5 × 104 A2s,
giving such a wide margin that the major fuse cannot suffer any permanent
effect.
It is evident from the foregoing that, if the rating of the major fuse is
not too close to that of the minor, discrimination will be assured, provided
that the fuses are of the same design. If the supply authority fits cartridge
fuses at the incoming feeder and the user installs semi-enclosed fuses in
the branch circuits, a larger ratio than two may be necessary to ensure that
the major fuse will not blow in the event of a fault in the branch circuit.
Circumstances can occur in which the major fuse is at approximately
its working temperature when a fault occurs in a sub-circuit containing an
unloaded minor fuse. Obviously the pre-arcing I2t of a hot fuse is less than
that of a cold one, but if the minimum ratio of two is maintained the major
fuse would not normally be more than 50% loaded and the corresponding
reduction in I2t would be insignificant. Tests have shown that even if a
cartridge fuse is preheated for a considerable time by a current not sufficient
to melt the element, the reduction in I2t is usually small, in the context of
discrimination.
Example 2. It is very common to employ the radial system as shown
in Fig. 6.34 which may contain a number of fuses which must be chosen to
discriminate or alternatively a fuse may have to operate in series with a
circuit which is tripped by a protective relay.
PD 1 Load 1
PD 4 PD 2 Load 2
Supply
PD 3 Load 3
FIGURE 6.34 Simple network
Here the major fuse in the supply connection (PD4) and minor fuses
in the individual load circuits (PD1, 2 and 3).
Clearly each minor fuse must have the time/current characteristic
needed to protect its load circuit and a fault on a particular load should
only cause its associated minor fuse to operate. The major fuse (PD4) will
also carry the fault current but it must not operate or be impaired.
For faults which cause relatively small currents to flow, the arcing
times, as proportions of the pre-arcing times, are small and consequently
discrimination can be predicted by comparing the time/current curves of
the major and minor fuses. Provided that the curves for the minor fuses
are to the left of that for the major fuse, that is the minor fuses operate
more quickly, then discrimination should be obtained. A significant margin
should nevertheless be allowed for fuse tolerances and because the major
fuselink may be carrying currents fed to healthy circuits as well as the
fault current. If these load currents may be large, then calculations should
be done to determine the possible currents in the major fuselink for given
fault currents and adequate time differences should still be present between
the operating times of the fuses concerned. This situation is illustrated in
Fig. 6.35.
Time, (sec)
tb
a
ta
Ia Ib Current
a = Minor fuselink
b = Major fuselink
Ib = Current in major fuselink when fault
current Ia flows in minor fuselink
tb = Must exceed ta characteristic
FIGURE 6.35 Fuselink characteristics
At higher fault-current levels which will result in melting the minor

fuse in less than 100 ms, the arcing time of the minor fuse must be taken
into account. This is done, not by considering the actual values of time,
but by using the I2t values. The requirement is that the pre-arcing I2t of the
major fuse shall exceed the total operating I2t of the minor fuse by a
reasonable margin (say 40%). It will be appreciated that load currents
flowing in healthy circuits while a fault exists, have negligible effects on
the operation of the fuses when the fault current is very great. Pre-fault
conditions are important, however, and the I2t margin suggested above
should be increased if it is known that before a fault, the minor fuse is
likely to be much less loaded, as a proportion of its rated current than the
FUSES 159
major fuse. This is because the minor fuse will be operating at lower
temperatures than the major fuse at the instant of fault occurrence. A
suggested rough, but general guide, for the extreme case of the minor fuse
being on no load and the major fuse on full load when the fault occurs, is
that the rating of the major fuse should be increased by a further 25%.
6.5.2 Networks Protected by Fuses and Devices of other Types

Here the general requirement is similar to that for discrimination between
two fuselinks, in that only the minor device is required to operate. It is the
latter device which has to be chosen first, because its time/current
characteristic must provide the necessary protection for its associated
circuit. Thereafter, the major device must have a characteristic which will
ensure discrimination.
In practice two alternative arrangements are encountered, one in
which the major device may be a fuse whilst the minor devices may be
small or miniature circuit breakers incorporating overcurrent protective
features, and the other in which a major circuit breaker and minor fuses
are used. The characteristics associated with these two situations are shown
in Fig. 6.36.
u
d f a = Minor device
u = Major device
1000
f = Fuselink
100
Time, (sec)
a
10
0.1
Current
FIGURE 6.36 Discrimination between a current limiting

fuselink and other protective devices
With the first arrangement, there is always an actual or potential

upper limit to the fault current at which discrimination can be obtained.
This is because the circuit breaker or other minor device always has a
definite minimum-operating time resulting from the delays in the
overcurrent detection equipment and the circuit breaker itself plus its own
arcing time, of which the latter is not likely to be less than the duration of
one half-cycle. The operating time of the major fuse on the other hand,
decreases continually with increase in current and the upper-current limit
at which discrimination can be achieved is at the intersection of curves d

and f in Fig. 6.36.
With the second arrangement, the curves f and u in Fig. 6.36 are
relevant. It will be seen that they approach each other most closely at times
of the order of 1–3 s where the influence of arcing time is negligible and
there is usually little difficulty in choosing characteristics which enable
full discrimination to be obtained.
6.5.3 Coordination between a Current Limiting Fuse and a Directly
Associated Device of Lower Breaking Capacity
It has already been stressed that some current-limiting fuselinks are unable
to clear currents in a range above the minimum fusing level and that an
associated device of limited-breaking capacity is needed to interrupt these
currents. In these circumstances the requirements are quite different from
the preceding two cases. The current limiting fuse is used as a back-up for
the other device. Thus in the event of heavy faults only the current-limiting
fuse is required to operate while the associated device is required to operate
in the event of overloads or small faults. This is achieved by choosing
characteristics for the fuselink and other device so that they produce a
composite characteristic of the form shown in Fig. 6.37, and this clearly
must give sufficiently rapid clearance at all current levels to adequately
protect the associated circuit. The other criteria which must be satisfied
are as follows:
1000
a
Time, (sec)
100
10 Max breaking
current of
expulsion fuse
1
A b
0.1
Current
a—Expulsion fuse characteristic
b—Current-limiting fuse characteristic
FIGURE 6.37 Coordination of a fuse and a directly associated device
(a) The take-over point (A) at which the curves intersect must be at
a current level below the breaking capacity of the other device
and above the minimum value which the current-limiting fuse
FUSES 161
link can interrupt, unless it is fitted with a striker which operates

the other device.
(b) To deal with cases where the current-limiting fuses clears the
circuit, the other device must be able to safely carry the maximum
fault current and, where it may have to close on to a fault, a
making capacity adequate for the cut-off and I2t let-through
values of the fuselink.
6.6 TESTING OF FUSES

Fuselinks, unlike most other equipment, cannot be subjected to extensive
routine proving tests at the end of the production process because if they
are operated they cannot be used again. The behaviour of individual
designs must therefore be determined by very rigorous type tests and then
the subsequent component parts must be produced to within very close
limits of those used in the type-tested fuselinks. In addition, inspection
and quality assurance systems must be employed to ensure that the volume-
production output corresponds closely with the initial devices.
In service, the conditions encountered by fuselinks may be very
variable and, of course, their performances may be affected thereby. To
ensure uniformity between manufacturers, tests must be done in specified
and standardised conditions. The tests are conducted in laboratories and
because of the control which is available, the limits set on parameters such
as ambient temperature are much smaller than those specified in the
standard conditions of service.
Because the performances of fuselinks may be significantly affected
by factors, such as the impedances of the test circuits, the size and
disposition of attachments (including the cables) and, the proximity of
supports or enclosures, the standards specify the test arrangements in great
detail and this is particularly so for the time/current and short-circuit
breaking-capacity tests.
In general the following checks and tests are done on a number of
pre-production fuselinks.
6.6.1 Construction and Dimensions
Each fuselink which is to be used during type testing must be carefully
examined during manufacture to ensure that there is nothing abnormal in
its construction and the dimensions of the component parts are measured
accurately to see that they are within close tolerances of the values to be
used in the subsequent volume production.
6.6.2 Electrical Resistance

This value is measured for each fuselink to be type-tested. It must be done
at an ambient temperature, typically in the range 20–25°C.
6.6.3 Power Dissipation
Normally low-voltage fuselinks must be tested in a standard rig, as stated
earlier. The power dissipated by each fuselink must be measured at rated
current and the temperature rises of contacts or terminals are measured by
thermocouples.
Similar tests are done on high-voltage fuselinks but the power input
need not be measured and when miniature fuselinks are tested only the
voltage drop at rated current needs to be recorded.
6.6.4 Power Acceptance of Fuse Holders

This test which is only applied to low-voltage fuses is performed by taking
the fuselink found to have the highest power dissipation in the above test
and inserting it into a fuse holder. Rated current is then passed through it
again and the temperature rises at various points are measured by
thermocouples.
6.6.5 Insulation Levels
Extensive tests must be done on high-voltage fuses because of the
vulnerable positions in which they may be used. They are subjected to
impulse voltages viz., 1/50 is waves) between parts which will, in service,
be live and earthed respectively. They are also subjected to power-frequency
over-voltages under both wet and dry conditions.
Low voltage fuselinks in their fuse holders are mounted on a metal
panel and a power-frequency supply of 2.5 kV is connected for one minute
between the panel and normally-live fuse parts.
6.6.6 Conventional Fusing Currents
The terms ‘conventional fusing current’ and ‘conventional non-fusing
current’ have been introduced into standard low-voltage fuse specifications
to replace the term, minimum-fusing current, which has been in use for
many years. The latter current is strictly the one which will cause a fuse
link to operate in an infinite time and therefore its determination is
impractical. The minimum-fusing current is defined as that current which
causes operation in four hours, but even the determination of this value is
very time consuming. To simplify the situation, the standard specifications
require that all of a number of fuse links mounted in a standard type-test
rig should operate in less than the conventional time when they are carrying
the conventional non-fusing current. The conventional time, which may
FUSES 163
vary between values of one and four hours, depending on the current rating,
is specified in the standards.
The concept of conventional currents has been used on the continent
for several years and it is likely that the present usage of the term
‘minimum-fusing current’ will be phased out.
6.6.7 Breaking Capacity

Because fuses are usually the ultimate back-up protection in the circuits in
which they are included they must be capable of operating under the most
onerous conditions which may arise. For this reason the tests at maximum
breaking capacity are done under specified conditions in low-power factor
(typically less than 0.2 for low-voltage fuses), single-phase, inductive
circuits arranged as shown in Fig. 6.38. Equipment is included to enable
the test circuit to be closed at any desired point in the voltage cycle so that
conditions of varying severity may be produced. The fuselinks are mounted
in standard rigs of the form shown in Fig. 6.39. The tests determine not
only breaking capacity but parameters such as I2t let through arc voltage
and cut-off current and when the latter is being found, the circuit is switched
so that arcing commences just prior to an instant when the system voltage
Circuit Make
Variable impedors breaker switch
Source of
Fuse
voltage
under
test
FIGURE 6.38 Test circuit
FIGURE 6.39 Standard test rig

is at its peak value. To cater for three-phase applications the appropriate

source voltage is used in the test circuit, that is 87% of the rated voltage in
the case of high-voltage fuselinks or 10% more than the rated line voltage
when low-voltage cartridge fuselinks are being tested. In all breaking-
capacity tests the full recovery voltage is maintained for at least 30 s and in
the case of high-voltage fuses it is maintained for 60 s.
In addition to the above tests, the specifications require
manufacturers to do type tests at lower current levels, and for these, the
power factors of the test circuits are generally higher than those used at
the maximum currents. This represents the situation which tends to arise
in service, proportionally more resistance being present at the lower fault
levels. This is also recognised in the standard specifications which allow
power factors of 0.3 to 0.5 for low-voltage fuselinks. Minimum breaking
capacities are not usually quoted by manufacturers for miniature fuselinks
and for the majority of low-voltage, general-purpose fuses, but a number
of type tests are done at different currents to ensure that they will operate
satisfactorily at somewhat arbitrarily chosen multiples of the rated currents.
The minimum breaking capacities of high-voltage cartridge fuselinks are
not specified in the standard documents, but it is nevertheless an important
parameter which is usually quoted by the individual manufacturers after
conducting tests in the standard manner.
To conduct the maximum breaking capacity tests it is required that
the fuse be tested for breaking capacity at least three levels of current.
Test 1 is at the maximum prospective current. Current-limiting fuses
are capable of breaking extremely high prospective currents because by
their current limiting action they prevent the current from reaching the
peak of the prospective current. In fact, the cut-off current only increases
proportionally to the cube root of the prospective current.
The point on the voltage wave is very important in testing fuses.
For example, the conditions of the breaking capacity test are specified not
only in terms of the current, voltage and power factor at which the test
must be made, but also the instant of initiation of arcing within the fuse
after the element has melted after voltage zero as follows:
For one test arcing must commence between 40° and 65° (this is the
condition represented in Fig. 6.40 by the curve a1, b1, c1). This condition is
considered to correspond to the maximum thermal stress on the fuselink.
For two further tests, arcing must commence between 65° and 90°
(this is the condition represented in Fig. 6.40 by the curve a2, b2, c2). This
FUSES 165
condition is considered to correspond to the maximum electromagnetic

stress on the fuselink.
If the fuse is capable of breaking its maximum prospective current
under these conditions, it should be capable of breaking it under any other
breaking angle.
Circuit voltage Same prospective r.m.s.
Voltage & current
Limits of arcing ‘current’ but different

angle closing angles, resulting
40° 65° 90° in different degrees of
asymmetry
b1 b2
a2 c2
0 30 60 90 120 150 180 210 240 270 300 330 360

Circuit closing angle g (degress)
a—Initiation of current; b—Initiation of arcing; c—Final clearance
FIGURE 6.40 Fuses blowing on high current, low power factor within
prescribed limits for breaking capacity.
Test No. 1 at maximum breaking capacity
Test 2 is at the maximum arc energy condition for a current-limiting

fuselink. It is made at a current, chosen such that it produces a cut-off
current approximately equal to the r.m.s. value of the prospective current,
when the circuit is closed at an angle of between 0° and 20° after voltage
zero. The condition of maximum arc energy usually corresponds to an
inductive current equal to 3–4 times the current corresponding to a pre-
arcing time of 0.01 s.
Tests 3, 4 and 5 are made in order to verify that the fuselinks will not
fail when breaking a small overcurrent. It has been explained earlier that
fuses have special difficulties in this region, and may be unable to extinguish
a low current arc, resulting in overheating, burning through the end caps,
or even explosion. Tests 3, 4 and 5 ensure that this does not happen, under
the specified conditions.
In all tests, 1–5, it is essential that, after the fuse has interrupted the
circuit, it can withstand the recovery voltage which appears at its terminals,
and should not restrike or produce dangerous leakage currents.
For this reason the recovery voltage is left on for a specified period
after tests (as long as 5 min if the fuselink contains any organic materials),
and the insulation is tested within 3 min of completion of the test to ensure
that it is better than a specified level.
6.6.8 Other Parameters Tested

Apart from the afore-mentioned tests, fuses are tested at independent
testing stations for the following parameters:
(a) dielectric properties
(b) temperature rise and power loss
(c) accuracy of time/current characteristic
(d) oil tightness (if intended to be used under oil)
(e) weather proofness and thermal shock (if intended to be used
outdoors)
(f) effectiveness of strikers (if fitted)
(g) accuracy of cut-off characteristics and I2t characteristics (when
required)
All these factors are recorded in the type test reports, in which the
data are assembled to prove compliance with a standard. This is intended
to ensure that every fuse made of identical manufacture to the fuse tested
will be safe and suitable for the application corresponding to the tests.
6.7 FUTURE DEVELOPMENTS

Modern fuselinks are available in greater range than ever before. Special
fuses for semiconductor protection have very fast operation, and carefully
controlled over-voltages. Motor circuit protection fuses have characteristics
specially designed for short-circuit protection without spurious operation
on sustained starting currents, in compact dimensions and with special
characteristics designed for discrimination with motor starters. Special
back up fuses are also available.
New developments are the dual-mode expulsion fuse and the
permanent power fuse. The dual mode expulsion fuse employs a saturating
transformer to produce a shape of time/current characteristic which gives
closer discrimination with an IDMT relay, permitting lower settings.
The permanent power fuse is a Japanese self-healing device using a
column of sodium as a fuse element. This is contained in a ceramic capillary
attached to an arc chamber, and the liquid sodium is forced to flow back
into this space when the arc extinguishes after the circuit has cleared. The
device is intended to back up circuit-breakers of limited breaking capacity.
New ideas still to be developed include the vacuum fuse and the oil-filled
fuse (not oil immersed).
CHAPTER
Distance/Impedance Protection
7.1 OVERVIEW OF DISTANCE PROTECTION

The extra high voltage systems form usually a tightly interconnected
network with no radial feeders. The fault currents in the network depend
more on the operating conditions of the whole system than on the fault
location along the line. The requirements of protection operating speed
and selectivity cannot be satisfied by grading by the current magnitude or
by current and time. For a great majority of transmission and
subtransmission lines the overcurrent protection systems are, therefore,
not adequate. Due to economical and technical restrictions, mentioned in
the chapter of “Transmission Protection” the application of differential
schemes is also limited, especially for longer lines. The most common
protection of transmission and subtransmission lines is distance protection.
Comprehensive explanation of distance protection theory is readily
available from the literature [references in bibliography]. Some features of
this theory are discussed in the following paragraphs. A good brief
description of the basics is provided in the textbook by Kothari and
Nagrath. Subsequent paragraph offers only a concise overview theory and
application of distance relays.
7.2 PRINCIPLE OF DISTANCE PROTECTION

A distance protection system is a non-unit system and as such it operates
on information obtained from instrument transformers at only one end of
the protected line. In principle the relay discriminates between faults on
the protected line, which should cause fast disconnection of the line at the
relay end, and the faults on the adjacent plant which should be left to the
protection of that plant to clear. This discrimination is achieved by
measurement of the distance from the relay location to the fault location.
This distance, when compared with the length of the protected line, should,
ideally, determine whether the fault location is nearer than or beyond the
other end of the line.
7.2.1 Zones
Using voltages supplied from the voltage transformers and currents from
the current transformers, the distance relay measures the impedance of
the line section from its location, at A, (Fig. 7.1), to the fault at K1. This
impedance is proportional to the fault distance A-K1. The distance relay
trips the circuit breaker at A instantaneously if the impedance measured is
less than a reference impedance e.g., the known impedance of line A-B .
Such a measurement cannot discriminate between the faults K2 just before
the busbar B in the station at the other end of the line, and those on or
beyond the busbars, like at K3 and K4 respectively, because the distance
from A to these points is practically the same. The protection engineer has
to make sure that the relay does not over-reach i.e., does not trip at A for a
fault on or beyond bus B even if the measurement of the impedance is
laden with some error. To prevent over-reaching, the reference impedance
for the instantaneous tripping is selected so that it corresponds only to
80% of the line length with a 20% safety margin left for all possible errors.
This creates a zone of 80% of the line, such that faults in that zone are
isolated instantaneously from end A. This zone is called Zone 1 of relay A
and the reference setting which determines the length of that zone is called
Zone 1 setting.
B
A K4 C
Ct K1 K2
G
K4 D
Vt
Distance
relay
FIGURE 7.1 Fault locations in zones 1 and 2 of distance relay
The isolation from A of faults in the remaining 20% section of the

line is not instantaneous. To isolate these faults the relay at A also compares
the impedance measurement with another reference, Zone 2 setting, greater
than the impedance of the whole line A-B. The relay is designed to trip at
A for faults in Zone 2 with a predetermined delay, usually 0.4–0.5 second.
This delay allows for instantaneous isolation at B of faults K3, K4 and K5 by
DISTANCE/IMPEDANCE PROTECTION 169
operation of bus B protection or the protection of line B-C or line

B-D, respectively, without unnecessary tripping at A.
To make sure that all faults on line A-B are isolated at A not later
than with Zone 2 delay, the Zone 2 setting is greater than the impedance of
line A-B by a margin of at least 20%, often 50% or more, so that the Zone 2
reach extends well into the lines outgoing from bus B.
Most distance relays have also a Zone 3 with even longer reach than
that of Zone 2 and a longer delay in operation. Zone 3 is often useful for
back-up protection for a remote circuit breaker or protection failure.
Distance relays are usually installed at both ends of a transmission
line and isolate any fault at both ends. Faults in the central 60% of the line
are isolated instantaneously by both relays while faults close to the line
terminals are isolated instantaneously at the end close to the fault and
with a delay of about 0.5 second at the remote end.
All zones of protection systems of various plant items have to be
coordinated to ensure their discrimination. Fig. 7.2 shows coordination of
distance schemes on adjacent lines. The operating time of relay C is shown
inverted for clarity.
Time
Zone 2 > 120%
Zone 1 = 80% Zone 1 = 80%
Distance
Zone 1
60%
Stat. A Stat. B Stat. C
Time at C
FIGURE 7.2 Time/distance characteristic
7.2.2 Selection of Measurands

The measurement of the impedance to the fault in distance relays is carried
out by comparison of selected voltages and currents. For a particular fault
only one pair of voltage and current is suitable for that measurement. For
example, the correct measurement of the impedance for a phase-to phase
fault a-b is by determining the ratio: Vab / (Ia – Ib). For a phase to ground
fault b-n the phase to ground voltage Vbn and a combination of phase b
and neutral current (Ib+ KIn) have to be used. To respond to all phase to
phase and phase to ground faults as many as six voltage/current pairs

have to be monitored and a three zone relay has to make 18 comparisons
of the measurement results with the reference values (settings).
The fast and expensive distance relays for E.H.V. transmission lines
have therefore 18 comparators continuously monitoring all possible fault
loops. This is necessary for the required speed and communication with
the other end as discussed later. Such as 18 element relay is called
‘full-scheme’ or ‘non-switched’ distance relay.
The relays for subtransmission lines usually do not have to be so
fast and do not have to use signalling links. Their construction is made
more economical by switching the same comparator to different loops and
using it consecutively for comparison in all zones. The switching is
controlled by simple separate starting elements that only have to recognise
the type of the fault and by the timer controlling the zones. This system is
called a ‘switched distance’ scheme. Traditionally switched schemes used
to be far cheaper than full schemes but this difference is much less
significant in the digital construction of relays. It is expected that the
switched schemes will be phased out in the future.
7.2.3 Polar Characteristics
The ratio of the voltage and current supplied to a comparator of a relay
represents a vector of impedance which can be shown in the impedance
plane Z (R, X), Fig. 7.3. This impedance is often referred to as ‘seen’ by the
relay. The angle of an impedances seen during direct faults on the protected
line, e.g., Zd, is equal to the angle of the line impedance, usually 70° to 89°
for transmission lines and in the order of 60° for lower voltage lines. If the
fault involves an object of some resistance like a tree branch, or just an
electric arc then the impedance seen by the relay has a horizontal resistive
component Ra and its resultant angle is smaller. Faults beyond the end of
the line are seen to have a greater magnitude, say Z4.
jX
C
Z4
B
Ra Load
Zd impedance
Za
area
30°
Line angle
A R
M
FIGURE 7.3 Impedance plane

The polar characteristic of a distance relay determines the area of

the impedance plane which contains all impedances causing operation of
the relay if seen by the comparator. Fig. 7.4 shows characteristics of Zones
1, 2 and 3 of a typical relay. The relay operates instantaneously if the
impedance seen by a Zone 1 comparator of the relay appears inside the
smallest circle. A Zone 2 delayed operation will result from seeing the
measured impedance at point K2.
7.2.4 Directionality
Change of the current direction with respect to voltage reflects in the
impedance plane as a change of the angle of the impedance seen. Faults
behind the relay, e.g., M in Fig. 7.1, fed via the protected line, correspond
to points in the third quadrant of the plane. Zones 1 and 2 in Fig. 7.4 are
directional because no point of the third quadrant is inside Zone 1 or 2
characteristics. A circular characteristic passing through the origin like those
of Zones 1 and 2 in Fig. 7.4, is known as a ‘mho’ characteristic.
jX
C
3
ne
Zo
B
ne
Zo
K2
ne
Zo
O
A R
FIGURE 7.4 Three zone distance relay
7.2.5 Stability on Load

During normal service of the line the comparators of the full scheme
continuously monitor impedances made of the normal voltages and the
load currents of the line. For a phase-to-phase element a-b the ‘impedance
seen’ is:
Va − Vb Va
Zl = =
I a − Ib Ia
The magnitude of this impedance is usually much greater than that
of any fault on the line. The angle of the load impedance Zl is that of power
factor of the transmitted load and varies between –30° and +30° for
transmission from A to B. A transmission in the opposite direction is seen

as an impedance with an angle between 150° and 210°.
The characteristics must be so selected that no load impedance is
seen inside any of the characteristics otherwise the relay will trip the healthy
line on load.
The shape of the characteristics may be circular, oblong e.g., lens,
quadrilateral or more sophisticated. Fig. 7.5 shows some typical
characteristics of distance relays. The variety of available shapes enables
separation of all impedances seen under no fault conditions from those
faulty conditions that require a reliable positive operation of the scheme.
The application of a particular relay is decided by its characteristics
and other factors like accuracy, operating time, burden, method of setting
adjustment, etc. Above all the relay must be reliable.
jX
jX
jX
R R R
Impedance mho Offset mho
jX
jX
R
R
Lens Quadrilateral
FIGURE 7.5 Distance relay measuring characteristics
7.2.6 Methods of Impedance Measurement

Determination whether the fault impedance (impedance of the line between
the relay and the fault point) fits within the characteristic of the relay may
be done by analogue comparison of signals derived from voltages and
currents or by numerical computation of the fault inductance and resistance
from the voltage and current samples taken at regular intervals in every cycle.
The methods based on analogue comparison used in more traditional
relays have already achieved high performance development levels of
reliability and speed. The numerical methods are used in digital relays
which, have already won enough confidence in the market to be exclusively
purchased by more and more utilities who appreciate the additional benefits
of digital technology.
7.3 ANALOG AMPLITUDE AND PHASE COMPARISON

A protection relay in general is expected to sense the change between
healthy and faulted conditions and send a signal when fault occurs. This,
in analog relays, is achieved by comparing two quantities either in
amplitude, or in phase. The amplitude or phase relation depends on the
system conditions and for a predetermined value of this relation, indicative
of a particular type and location of fault, the relay operates. Except in relays
such as overcurrent relays, where only one electrical quantity overcomes a
mechanical quantity such as the restraint from a spring, it is usual to
compare two electrical quantities. As such the device performing the
comparison part is the heart of a protective relay and is known as a
comparator. It decides the operating characteristics of the relay.
Taking a very general case to cover the complete range of
conventional relay characteristics, let S1 and S2 be the two input signals
such that when the phase relationship or magnitude relationship obeys
predetermined threshold conditions, tripping is initiated. The input signals
are derived from the primary power system via current and voltage
transformers. These signals may be derived from primary voltage or current
or from both, the latter necessitating some form of mixing device such as a
current voltage transactor as shown in Fig. 7.6.
v K1 K3 v
S1 Comparator S2
I K2 K4 I
FIGURE 7.6 Comparison of mixed signals
S 1 = K1 V + K 2 I ...(1)
S 2 = K3 V + K 4 I
where K1 and K3 are scalar constants and K2 and K4 vector constants with
angles θ2 and θ4 respectively. Taking V as the reference vector and vector
I to lag V by an angle φ equation (1) reduce to
S1 = K1|V| + K2|I|{cos(θ2 – φ) + j sin(θ2 – φ)} ...(2)
S2 = K3|V| + K4|I|{cos(θ4 – φ) + j sin(θ4 – φ)}
7.3.1 Analysis for Amplitude Comparator

If the criteria for operation is given by | S1 | ≥ | S2 | , then at the threshold of
operation | S1 | = | S2 |, equating the moduli of expression (2)
{K1|V| + K2|I|cos(θ2 – φ)}2 + {K2|I|sin(θ2 – φ)}2 ...(3)
{K3|V| + K4|I|cos(θ4 – φ)}2 + {K4|I|sin(θ4 – φ)}2
Rearranging the terms
(K12 – K32)|V|2 + 2|V||I|{K1 K2 cos (θ2 – φ) – K3K4 cos(θ4 – φ)}
+ K22|I|2 cos2 (θ2 – φ) – K42|I|2 cos2 (θ4 – φ) + |I|2 {K22 sin2 (θ2 – φ)
– K42 sin2 (θ4 – φ)}= 0
or (K12 – K32)|V|2 + 2|V||I|{K1K2 cos (θ2 – φ) – K3K4 cos (θ4 – φ)}
+ (K22 – K42)|I|2 = 0 ...(4)
Dividing by (K22 – K42)|V|2
| I |2 | I | (K 1K 2 cos θ 2 − K 3 K 4 cos θ 4 ) cos φ
+2
|V| 2
|V| K 22 − K 24
(K 1K 2 sin θ 2 − K 3 K 4 cos θ 4 ) cos φ K 12 − K 23

+ + 2 =0 ...(5)
K 22 − K 24 K 2 − K 24
| I |2 |I|
or +2 {x cos φ + y sin φ} + z = 0
V |V|
where x = (K1K2 cos θ2 – K3K4 cos θ4)/(K22 – K42)
y = (K1K2 sin θ2 – K3K4 sin θ4)/(K22 – K42)
K 12 − K 23
z=
K 22 − K 24
Equation (5) represents the equation of a circle on the β-plane having
|I/V|cos φ and j|I/V| sin φ as coordinates represented as
|I V|p + j|I V|q.
This circle as shown in Fig. 7.7 has radius
{K 12 K 24 + K 22 K 23 − 2K 1K 2 K 3 K 4 cos (θ 2 − θ 4 )}
r=
K 22 − K 24
{K 12 K 22 + K 23 K 24 − 2K 1K 2 K 3 K 4 cos (θ 2 − θ 4 )}
and d=
K 22 − K 24
Similarly equation (4) can be plotted in the α-plane by dividing it by

(K12 – K22)|I|2.
7.3.2 Analysis for Phase Comparator

The two quantities to be compared are S 1 and S 2. If α is the phase angle of
input S 1 and β that of S 2, the relay operates when the product of S 1 and
S 2 is positive. The product is maximum when the two quantities are in
phase. All the conventional characteristics of relays can be obtained with a
symmetrical phase comparator with (α – β) = ± 90°. Therefore, the threshold
condition is
tan (α – β) = ± α
tan α − tan β
or =±α
1 + tan α tan β
i.e., when 1 + tan α tan β = 0
K 2|I|sin (θ 2 − φ) K 4|I|sin (θ 4 − φ)
1+ × =0
K 1|V|+ K 2|I|cos (θ 2 − φ) K 3|V|+ K 4|I|cos (θ 4 − φ)
K1K3|V|2 + {K1K4cos (θ4 – φ) + K2K3cos (θ2 – φ)}|V||I|
+ K2K4 |I|2 cos (θ2 + θ4) = 0 ...(6)
This equation is again similar to equation (4) and can be plotted on
the β-plane. Dividing Equation (6) throughout by K2 K4 |V|2 cos (θ2 – θ4),
we get
| I |2 | I | {K 1K 4 cos(θ 4 − φ) K K cos (θ 2 − φ)
+ + 2 3
V | V | K 2 K 4 cos (θ 2 − θ 4 ) K 2 K 4 cos (θ 2 − θ 4 )
K 1K 3
+ =0
K 2 K 4 cos (θ 2 − θ 4 )
| I |2 | I | {(K 1K 4 cos θ 4 + K 2 K 3 cos θ 2 ) cos φ

+
V |V| K 2 K 4 cos (θ 2 − θ 4 )
(K 1K 4 sin θ 4 + K 2 K 3 sin θ 2 ) sin φ} + K 1K 3
+ = 0 ...(7)
K 2 K 4 cos (θ 2 − θ 4 )
| I |2 |I|
or + {x cos φ + y sin φ| + z = 0
V |V|
where
K 1K 4 cos θ 4 + K 2 K 3 cos θ 2
x=
K 2 K 4 cos (θ 2 − θ 4 )
K 1K 4 sin θ 4 + K 2 K 3 sin θ 2
y=
K 2 K 4 cos (θ 2 − θ 4 )
K1 K 3
z=
K 2 K 4 cos (θ 2 − θ 4 )
The circle has
{K 12 K 24 + K 22 K 23 − 2 K 1K 2 K 3 K 4 cos (θ 2 − θ 4 )}
Radius r =
2 K 2 K 4 cos (θ 2 − θ 4 )
{K 12 K 24 + K 22 K 23 − 2 K 1K 2 K 3 K 4 cos (θ 2 − θ 4 )}
and c=
2 K 2 K 4 cos (θ 2 − θ 4 )
Values of r and c for plot on the α-plane can be obtained, similarly.
In most relays at least one of the constants K (i.e., K1, K2, K3, K4) is
zero and two of them are often equal. Also the angle of the two vector
constants is usually the same. This makes the practical case relatively
simple. If θ2 = θ4 the values of r and c in the two cases are tabulated as
shown in Table 1.
TABLE 1: Comparative values of r and c for amplitude and phase
comparators (Fig. 7. 7)
Quantity Amplitude comparator Phase comparator
K 1K 4 − K 2 K 3 K 1K 4 − K 2 K 3
r
K 22 − K 24 2 K 2K 4
K 1K 2 − K 2 K 4 K 1K 4 + K 2 K 3
c
K 22 − K 24 2 K 2K 4
7.4 RELAY TYPES AND THEIR APPLICATION

Using specifically selected constant vectors (K1, K2, K3 and K4), various
types of distance relay characteristics may be obtained. The principal types
of distance relays are:
(i) impedance
(ii) reactance
(iii) admittance (mho)
(iv) ohm
(v) offset mho.
The common types compare two input quantities either in magnitude
or in phase. Any of the relay characteristics can be obtained either by an
amplitude comparator or by a phase comparator as explained earlier.
7.4.1 Impedance Relay

This is a device which measures distance by comparing the fault current I
with the voltage V across the fault loop. It is usual in this case to have an
amplitude comparator and the balanced beam type structure is most
common. The equation for the amplitude comparator at threshold as
derived already in equation (4) is
I
V q
f I
V p
FIGURE 7.7 Threshold characteristics of a comparator
(K12 – K32)|V|2 + 2|V||I|{K1K2 cos (θ2 – φ) – K3 K4 cos (θ2 – φ)

+ (K22 – K42)|I|2 = 0
If the constants are so adjusted that the input signals are
S 1 = K1 |V|
and S 2 = K4 |I|
i.e., K2 = K3 = 0
Substituting these conditions in the above equation, we get
K12 |V| 2 = K42 |I| 2
| V | K4
or = ...(8)
|I| K1
i.e., Z = constant K
The characteristics when plotted Z=X
on the R-X plane is shown in Fig. 7.8
which is a circle with origin as its centre;
signifying that a simple impedance relay R
would operate for any value of
impedance lying within the circle. The
characteristic also depicts that the relay
is not directional and it is essential to
provide a directional relay along with an FIGURE 7.8 Characteristics of
impedance relay. The combined impedance relay
characteristics of an impedance and directional relay are shown in Fig. 7.9,

where DD′ represents the directional relay characteristic and the operating
region is the shaded portion.
X
D¢
FIGURE 7.9 Characteristics of impedance with directional relay
7.4.2 Reactance Relay

All other relays except the impedance relay are conveniently obtained by
a phase comparator. The basic equation for a phase comparator already
derived in equation (6) at threshold is:
K1K3|V|2 + {K1K4 cos (θ4 – φ) + K2K3 cos (θ2 – φ)|V||I|
+ K2K4 |I|2 cos (θ2 – θ4) = 0
Using the input signals in the following manner
S1 = – KV + K’I ∠ (θ – φ)
S2 = K’ I ∠ (θ – φ)
i.e., K1 = – K
K 1 = K 4 = K’ ∠ θ
and K3 = 0.
Substituting these conditions in the above equation, we have
– K K’ cos (θ – φ) |V| |I| + K’2 |I|2 = 0
K’
or Z cos (θ – φ) = ...(9)
K
Now if θ is ∠π/2 the above equation reduces to the reactance form,
K’
i.e., Z sin φ =
K
K’
or X= ...(10)
K
When plotted on the R-X diagram, the characteristic is represented

by a straight line parallel to the horizontal axis R as in Fig. 7.10.
X
X = K¢
K
O R
FIGURE 7.10 Characteristics of reactance relay
With some predetermined setting of value X, the relay will measure

any value of reactance below the setting. A reactance relay responds only
to the reactive component of system impedance; consequently it is
unaffected by fault arc resistance. However, when fault resistance is such
a high value that load and fault current magnitudes are comparable the
reach of the relay is modified by the value of the load and its power factor
and may either overreach or underreach.
Voltage-restrained starting relays are used in a reactance measuring
scheme to give directional response and to prevent operation on load. The
reactance relay as seen by equation 10 is a particular case of an ohm relay,
in which the angle of compensation θ is 90°.
7.4.3 Admittance (Mho) Relay
If the signals S1 and S2 given to the phase comparator are:
S1 = – K |V| + K’ |I| (θ – φ)
and S2 = K |V|
i.e., K1 = – K
K2 = K’ ∠ θ
K3 = K
K4 = 0
Substituting these conditions in equation (6), we have
– K2 |V|2 + KK’ cos (θ – φ) |V||I| = 0 ...(11)
I K
or cos (θ – φ) =
V K’
K
or Y cos (θ – φ) = ...(12)
K’
This represents the admittance or the mho characteristic and when

plotted on the R-X diagram is a circle passing through the origin and when
plotted on the G-B diagram is a straight line as illustrated in Fig. 7.11.
X Stable B
Stable
q G
K¢
K Trip
Trip
K¢
K
q
R
(a) R-X plane (b) G-B plane
FIGURE 7.11 Characteristics of mho relay
The circle passing through the origin makes it inherently directional.

With such a characteristic the relay measures distances in one direction
only.
From the expression (11) it is evident that the relay is inoperative if
the voltage falls to zero, because both terms contain V. A memory circuit
may be used to prevent the immediate decay of voltage applied to the
relay terminals when a close-up three-phase short-circuit occurs. This
enables high speed mho protection to operate correctly on close-up faults,
provided that the protected circuit is energised before the short-circuit is
applied.
7.4.4 Ohm Relay
As explained earlier a reactance X
relay is a particular case of an ohm

relay. Its characteristic is represented
by equation (9) when plotted on the
R-X plane is a straight line (Fig. 7.12). Trip
K¢
The ohm relay is used as a K
q R
supplementary element to modify
the operating region of the other
types of measuring elements. FIGURE 7.12 Ohm relay characteristics
7.4.5 Offset Mho Relay

Let the signals S1 and S2 given to the phase comparator be
S1 = K |V| + K2 |I| ∠ (θ – φ)
S2 = K |V| + K4 |I| ∠ (θ – φ)
i.e., K1 = – K
K2 = K2 ∠ θ
K3 = K
K 4 = K4 ∠ θ
Substituting these conditions, we have
– K2 |V| 2 + {– K1K4 cos (θ – φ) + K2K cos (θ – φ)} |V| |I|
+ K2K4 |I| 2 = 0
V
Let be Z
I
and we get
– K2Z2 + K2K4 + KZ (K2 – K4) cos (θ – φ) = 0
Also, as Z2 = R2 + X2
Therefore,
K 2 K 4 (K 2 − K 4 )
R2 + X2 (R cos θ + sin θ) = 0
K2 K
(K 2 − K 4 ) cos 2 θ ( K 2 − K 4 ) sin 2 θ (K 2 + K 4 ) 2
or R– × ≤
2K 2K 2K
This equation represents a circle, with centre at (K2 – K4)/2K∠ θ on
the R-X plane, the radius being of magnitude (K2 + K4)/2K. The offset
threshold characteristic is shown in Fig. 7.13.
X
K2 Ðq
K
q
R
K1 Ðq O
K
FIGURE 7.13 Offset mho relay characteristics
An offset form is extensively used to provide a measure of back up

protection in respect of faults behind the relay.
7.4.6 Some Other Forms of Relay

The reactance characteristic is often used in combination with the mho
characteristic to provide extra coverage in respect of fault arc resistance on
lines of un-earthed construction (particularly, but not exclusively
distribution feeds). Such a set up produces a resultant characteristic as
shown in Fig. 7.14.
x
Faults are
coverage
FIGURE 7.14 Mho reactance relay characteristics
Extra fault arc coverage may be obtained by means of an offset mho/

reactance/directional characteristic as shown in Fig. 7.15.
X
FIGURE 7.15 Offset mho/reactance/directional relay characteristics
When protecting long lines, the reach of the polar characteristic along
the resistive axis can be many orders greater than the maximum fault arc
resistance. Even under normal conditions, when the line is transferring
power, some impedance is measured by the relay and this is generally
resistive but outside the characteristic. However, for long lines and heavy
circuit loading, healthy circuit tripping may occur, and it is necessary to
restrict the characteristic as shown in Fig. 7.16.
er
Blind
Typical load
impedance
area
R
FIGURE 7.16 Restricted characteristics for long line application
7.5 DERIVATION OF SIGNALS FOR DISTANCE PROTECTION

We have seen that distance relays make use of the voltage and current at
the relaying point. In practice we have to cater for several different types
of fault, i.e. 3-phase-earth faults (a – e, b – e, c – e) and 3-phase-phase faults
(a – b, b – c, c – a), and care has to be taken in the choice of the voltages and
currents to cover each type of fault. Consider a single phase-earth fault in
phase a of a feeder fed from one end only, as shown in Fig. 7.17.
a z
b z
c z
e zg
FIGURE 7.17 Single phase to earth fault representation
If we feed the relay with a voltage Va – e and Ia it will measure an

impedance Z + Zg.
However, when we consider a line fed from both ends (which is the
usual case), the current in the earth return path incorporating Zg is usually
different from that in the faulty conductor (i.e., there are sound phase-
currents).
Under these conditions Va = IaZ + IgZg, and the relay measures
Va/Ia = Z + (Ig/Ia)Zg. Now Ig depends on the pre-fault loading and the
method of earthing at either end of the feeder, and we cannot use just Va
and Ia to cover this fault. In practice we feed the relay with a current Ia + KIg
(Ig = residual current = Ia + Ib + Ic) and so the relay measures an impedance.
Va I a Z + I g Z g Z(I a + I g Z g/z )
= =
I a + KI g I a + KI g I a + KI g
Now the ratio Zg/Z does not vary with the position of the fault, and
if we make the residual compensating factor K = Zg/Z the relay always
measures the line impedance to the point of fault irrespective of the current
actually flowing in the earth return.
Likewise care has to be taken in the relaying qualities used in respect
of phase-faults, as shown in Fig. 7.18.
Ia z
Vab
z
Ib
zg
FIGURE 7.18 Phase faults representation
Here we see Vab = IaZ – IbZ, and provided we feed a voltage Vab and
Ia – Ib to the relay the impedance measured is again the impedance Z to the
point of fault.
The various voltages and currents used in distance protection are thus:
For Earth faults a–e Va → Ia + KIg
b–e Vb → Ib + KIg
c–e Vc → Ic + KIg
For Phase faults a–b Vab → Ia – Ib
b–c Vbc → Ib – Ic
c–a Vca → Ic – Ia
In schematic form the signals are derived from the system as shown
in Fig. 7.19.
CT
VT
CT
VT
CT
Trip
VT
Relaying
signals
Replica Mixing Comparators

impedance circuits
FIGURE 7.19 Signals derivation
7.6 METHODS OF REALISING COMPARATORS

Before considering circuit details, it is important to note the duality between
pure phase and pure amplitude comparators.
Consider a phase-comparator in block form as shown in Fig. 7.20
which trips when
β ≤ S1/S2 ≤ α
Likewise consider a pure amplitude comparator as shown in
Fig. 7.21.
S1 S1
Phase comparator Output Amplitude comparator Output
S2 S2
FIGURE 7.20 Block diagram of FIGURE 7.21 Block diagram of

phase comparator amplitude comparator
Here we get an output if |S1| ≥ |S2|.

Suppose we feed a signal S1 + S2 and S1 – S2 to our phase comparator,
and take signal S1 as a reference. Let S1 = A, S2 = B + jC and we get an
output from the phase comparator if
β ≤ (S1 + S2)/(S1 – S2) ≤ α
But (S1 + S2) = (A + B) + jC

(S1 – S2) = (A – B) – jC
Thus
S1 + S 2
∠ = ∠(A2 – B2 – C2 + jAC) = α at the limiting conditions.
S1 − S 2
AC
Thus tan α =
A − B2 − C 2
2
but S1 = A1, S2 = (B2 + C2)1/2

(|S1|2 – |S2|2) tan α = |S1| C
If α – φ = π/2 we see that tan α and since |S1|, C are generally finite,
there is an output if |S1| ≥ |S2| i.e., |S1| = |S2| is the limiting condition.
Thus by feeding S1 + S2, S1 – S2 to a pure phase comparator (over the limits
+/– π/2) we effectively compare S1 and S2 in amplitude. The same clearly
applies in reverse i.e., by feeding S1 + S2, S1 – S2 to an amplitude comparator
there is an output if – π/2 ≤ S2/S1 ≤ π/2 i.e., we may effectively compare S2
and S1 in phase by means of an amplitude comparator.
Fig. 7.22 summarises the situation where a phase-comparator is used
to compare the amplitude of signals S1/S2.
Inherent Inherent
phase phase
comparator comparator
S1-S2 S1-S2
S1 S2
S2
S1
FIGURE 7.22 Equivalent comparator relationship
7.6.1 Amplitude Comparators

A very common form of amplitude comparator is the back to back bridge
arrangement as shown in Fig. 7.23(a). Here the signals are in the form of
currents, and the action may be seen in the following Fig. 7.23(b).
The output voltage = (I1 – I2)R and the average volts are seen to be
positive if |I1| > |I2| , where the average is taken by means of integrator
circuit. Note: That the operating criterion is independent of the phase of S1
and S2.
I1 I2
S1 S2
(a) Back to back amplitude comparator
+ +
– –
S1
I1
I2
S2
(b) Waveforms
FIGURE 7.23
There is another comparator as illustrated in Fig. 7.24 which is often

taken to be a pure phase-comparator, but which is really an amplitude
comparator having inputs (S1 + S2) and (S1 – S2).
S1 + S2
S2 1 : 1
To integrator
Va trips when
ò Va dt is + Ve
S1 – S2
2S1
1:2
S1
FIGURE 7.24 Another form of amplitude comparator

It can be seen that the average output volts are positive (i.e. trip)
when –π/2 < S2/S1 < + π/2 but the transformer arrangements are really a
crafty way of providing the signals (S1 + S2), (S1 – S2) which are compared
in amplitude therefore. Some proposed methods of providing inherent
amplitude comparators using transistors have been made, but with
reference to high speed methods problems clearly arise. It is necessary to
locate the maxima of the signals and compare, and this makes the circuitry
involved quite complex. Also as we shall see later there is a large dynamic
range of relay signals, and this aggravates the situation. We will not
therefore consider amplitude comparators further.
7.6.2 Phase Comparators

The measurement of phase is fundamentally easier than amplitude, since
zero-crossings are easy to detect and in theory independent of the
magnitude of the signals being compared. There are many methods (some
practical others less so) of performing phase-comparison using
transistorised circuits, and now using integrated circuits. Because of their
importance we shall consider some of the methods in detail in order of
practicability.
7.6.3 Pulse Type Phase Comparator

Here a short duration pulse produced at the maximum of one relaying
signal is compared with one half cycle of the other signal. Thus a short
duration pulse appears at the output (E) provided –π/2 < ∠ S2S1 < π. The
principle is illustrated in Fig. 7.25.
S1 Sine/square A
conversion
AND E
Gate
B C D
S2 p/2 Sine/square
Phase shift Differentiator
conversion
(a)
v S2
S1
t
A t
B t
C t
D t
E t
(b)
FIGURE 7.25 Principle of pulse type phase comparator
7.6.4 Fixed Time Reference Comparator

Here the coincidence period between the two signals is compared with a
pulse generated therefrom. The diagram in Fig. 7.26 illustrates the principle
and shows that tripping is obtained if – (π – ωT) < ∠ S2/S1 < (π – ωT).
Notice that the limits of comparison are symmetrical about the condition
of in phase relaying signals and that by making ωT = π/2 the comparator
may be used to produce the common mho and offset mho characteristics.
S1
AND Time AND
S2
gate A delay (T) B gate C
(a)
v S2 v S2
S1
S1
t t
v v
A t A t
v v
B t B t
w 0T = d w0T = d
v v
C t C t
S1/S2 £(p – w0T) S1/S2 £(p – w0T)

(Trip) (Block)
(b)
FIGURE 7.26 Circuit action of fixed time reference phase comparator
7.6.5 Block-Average Comparator

In this device the coincidence period between two signals is fed to a linear
integrator, so that the output voltage increases for all phase angles between
the signals in the range –π/2 < S2/S1 < π/2. The schematic in Fig. 7.27
shows the principle of operation.
Such comparators have an inverse type of time response i.e., the
operating time increases as the relaying signals approach threshold (± π/2).
Thus for faults near the boundary of a characteristic the device takes longer
to operate. This feature increases the immunity of such devices to
measurement errors caused by transients in the relaying signals. In practice
the characteristic boundaries may be preserved provided the pick up/drop

S1
Coincidence A Integrator
B Level C
S2
detector detector
(a)
v v v
t t t
v v v
A t t A t
v v v
B t t B t
VR VR VR
VS VS
VS
v v v
C t t C t
Inside threshold Near threshold Outside threshold

(b) (c) (d)
FIGURE 7.27 Block average arrangement
off ratio of the level detector exceeds about 2.5 (theoretically two is a
transient exists in only one signal). This comparator is extensively used
due to its desirable response. Its only disadvantage is that it is rather
inflexible from in that it is only possible to produce simple shapes of
characteristic (e.g., mho) which nevertheless are satisfactory for the majority
of lines. Other more flexible methods to cater for the requirements of long
and very important lines have been developed.
Block Average Arrangement
The relaying signals compared in a block average comparator arrangement
to produce the polarised mho characteristics are of the well-known form
of equation (14).
S1 = IZr – V ...(13)
S2 = V + Vp ...(14)
where I, V are the relaying current at CT secondaries and the relaying
voltage at the VT secondaries respectively. Vp is the sound phase polarising
voltage and Zr is the replica impedance, the former is obtained using the
polarising circuit and the latter is associated with a transformer-reactor

arrangement.
The signal S1 is of a standard form and varies according to the type
of fault, e.g., for an ‘a’-earth fault, is of the form as shown in equation (15).
S1a = (Ia + KI res) Zr – Va ...(15)
where Ires is the zero sequence current and K is the residual compensating
factor, as given by equation (16).
K=
F
1 | Z LO |
−1
I ...(16)
GH
3 | Z LI | JK
where ZLO and ZLI are the z.p.s. and p.p.s. line impedances.
It is common practice to place replica impedances in the secondary
of the current transformers in order to generate the relaying signal S1 as
given by equations (14) and (15).
The transactor which is basically a transformer with a core having
air gaps and is extensively used in protection equipment using the block
average comparator arrangement and has a circuit as shown in Fig. 7.28(a).
Rb is a burden resistance so that neglecting the secondary winding reactance
(which is relatively low because of the air gap) yields the equivalent circuit
of Fig. 7.28(b) referred to the primary. With reference to Fig. 7.28(b) it can
be shown that the magnitude of the Zr the replica impedance referred to
the primary of the transactor, is given by equation (17).
I IP¢ I : nt
Lm R1 Va
(a) Actual circuit

I
Rb Va
Vi Xm = Rb¢ = Va¢
n2t nt
(b) Equivalent circuit referred to primary
FIGURE 7.28 Transformer-reactor arrangement

Vo′ Xm
Zr = = ...(1)
I [1 + cot 2 φ T ]
where φT = tan–1 (Rb/nt2 Xm).
The relaying signal produced by the transactor (e.g., for an ‘e’-earth
fault) is therefore S = I Zr – Va or in the transduced form, it can be shown
that:
nc
S1′ = nt2 Z I – Va′
nv r
where nc, nv are the turns ratio of the CT and VT respectively, which are
typically 1200/1 and 500/0.11 (for a 500 kV system) respectively. nt is
normally taken as 1/1.
It can thus be seen that the Zr′ replica impedance referred to the
secondary is given by equation (18).
nv
Zr′ = Z ...(18)
nc r
7.7 SIGNAL S2
As regards signal S2, there are essentially three basic types:
(a) self-polarised
(b) fully cross-polarised
(c) using the memory circuit.
7.7.1 Self-Polarised Type
In this case, the polarising signal is obtained from the same faulted phase
and is not very satisfactory, mainly because for close-up earth faults the
faulted phase voltage is close to zero, which gives unsatisfactory S2 signal.
However, it can have certain advantages for other fault positions. For an
‘a’-earth fault, for example, the S2 signal is given by equation (19).
S2a = Va + kp Va ...(19)
where kp is a constant.
7.7.2 Fully Cross-Polarised Type
The fully cross-polarised type is the most commonly used. In this case, the
polarising signal is obtained from the healthy phases and for an ‘a’ earth
fault, for example, it is given by equation (20).
S2a = Va + kp Vbc ...(20)
where kp is the polarising constant and is equal to kp ∠ 90°.
7.7.3 Using the Memory Circuit

It has been found that the signal S2 obtained using the memory circuit can
have certain advantages. In this type a mixture of the faulted and/or healthy
phase voltage with a voltage from a memory circuit is used as the polarising
signal. As, for example, for ‘a’ earth fault, the signal is as expressed in
equation (21).
S2a = Ea
or S2a = Ea + kp Vbc ...(21)
where Ea is the prefault steady state component of voltage and kp is the
polarised voltage constant which is equal to kp ∠ 90°.
7.8 PRE-FILTERS USED IN POWER SYSTEM PROTECTIVE

RELAYING
Second order low pass filter pre-conditions the S1 and S2 signals.
A low-pass filter is a device which passes signals of low frequencies
and suppressed or attenuates those of high frequencies. Its performance
may be illustrated by its amplitude response, which is a plot of the
amplitude H(jω) of its transfer function H(s) versus frequency ω,
where ω = 2πf. In all cases H = vo/vi, where vo and vi are the output and
input voltages.
Transactor
CVT
Phase shift
circuit
Missing
circuit S1
Second order To
filter comparator
S2
FIGURE 7.29 Pre-filtering circuits used before entering comparator
It is common practice to prefilter the signals S1 and S2 in a second

order filter before going to a comparator. A second order approximation to
an ideal low pass filter is achieved by the transfer function.
vo K1
= 2 ...(22)
vi p + ap + b
where ‘a’ and ‘b’ are properly chosen constants and K1 is a constant. The
term ‘second order’ refers to the degree of the denominator polynomial.
The gain of the low pass filter is the value of its transfer function at p = 0
and is given in the case of equation (22) by gain = K1/b.
There are a number of ways of obtaining low pass filters using active
devices instead of inductors. The method commonly used is that of Sallen
and Kelly, in which the active device is an operational amplifier (op-amp).
Vc(t)
R1 R2 Cf
+ +
+
–
Vi(t) V(t) C1
V¢(t) R3 Va(t)
R4
–
(a) Actual circuitary
i2(t) Vc(t)
R1 R2 Cf
+ +
i1(t) +
–
Vi(t) C1
V(t) Va(t)
(b) Equivalent circuit
FIGURE 7.30 Sallen and Kelly second order low pass filter
A Sallen and Kelly second order low pass filter is shown in Fig. 7.30(a),
where the resistors and capacitors are properly chosen to realise given
values of ‘a’ and ‘b’ in equation (22). The op-amp, together with the resistor
R1 and R2, constitutes a voltage-controlled voltage source (VCVS) and hence
the Sallen and Key network is of the VCVS type.
The value of the constants are as follows:
µ × 10 6
K1 =
R 1 R 2 C 1C f
1 1 1
a= (1 − µ ) + +
R 2 C1 R 1C f R 2 C f
1
b=
R 1 R 2 C f C1
The quantity l = 1 + R4/R3 is the gain of the VCVS and is also the
gain of the filter since K1/b = µ
There are many types of low pass filter, but the two most commonly
used are the Butterworth and Chebyshev types.
7.9 EFFECT OF THE RATIO SOURCE IMPEDANCE TO LINE

IMPEDANCE (ZS/ZL)
Any system could be represented by a single line diagram shown in Fig.
7.31 representing the source and the line to be protected, R being the relay
location. This simple impedance loop has a voltage V applied to it which
actually depends on the type of fault whether phase fault of ground fault.
ZS and ZL are the source impedance and the line impedance respectively.
ZS is a measure of fault MVA at the relaying point, and for faults involving
earth, is also dependent on the method of system earthing behind the
relaying point. ZL is a measure of the line impedance of the protected
section. IR and VR are respectively the current and voltage applied to the
relay. The voltage VR applied to the relay is thus IR ZL for a fault at the
reach point, and this may be alternatively expressed in terms of ZS /ZL
ratio as
Source R Line
FIGURE 7.31 Power system arrangement
VR = IR ZL
V
where Ir =
ZS + ZL
ZL
Thus Vr = V
ZS + ZL
1
Vr = V ...(23)
( Z S /Z L ) + 1
Equation (23) is true for all types of faults with the following rule
being observed:
1. For phase-faults V is the delta voltage and ZS/ZL is the positive
sequence source impedance/positive sequence line impedance,
V∆
i.e., VR = ...(24)
(Z S1/Z L1 ) + 1
2. With earth-faults V is the star voltage and ZS/ZL is a composite

ratio involving positive and zero sequence impedance, i.e.,
VY
VR =
(Z S /Z L ) + 1
where ZS = 2ZS + ZS
1 0
ZL = 2ZL + ZL
1 0
and of course VY = V∆ / 3 .
Fig. 7.32 illustrates the effect of ZS/ZL ratio on the voltage at relay position
R. Three fault locations are shown: Fault outside the zone of protection,
fault at setting distance, and fault in the protected zone. The ordinate at R
represents the voltage applied to the relay VR for various fault locations,
for the relay to restrain VR should be greater than the preset value of
VR = IR ZL and for operation V should be less than the preset value of
VR = IRZL.
VS VL = VR
R IR
ZS ZL
V VR
FIGURE 7.32 Relation between source, line and relay voltage
7.10 TIME GRADING OF DISTANCE RELAYS

A typical distance relay has stepped characteristics in three stages (as shown
in Fig. 7.2). Zone 1 of the relay provides instantaneous tripping for any
fault within a predetermined distance from the relay (generally 80% of the
protected feeder). Zone 2 operates with a time delay and covers the
remainder of the protected feeder and backs up the protection on the next
feeder for a part of that feeder. After a further time delay the range of the
distance relays is further extended to Zone 3, which is used as backup
protection only. The time delay between the various zones is chosen
according to stepped principle (usual value of time delay step is 0.5 s).
Because of this only high-speed distance relays are widely used in practice.
Such relays can be arranged for measuring impedance or reactance and
are normally provided with built-in directional feature. Hence the basic
features involved in such a relay are: (i) impedance or reactance
measurement, (ii) direction and (iii) timing.
For finding the pickup value of the impedance in section n let us say
Zone 1 is 80% of the protected section.
Z(n)
pickup (1) = 0.8 ZL
(n)
where Z(n)
pickup (1) = Setting value of pickup impedance of the Zone 1
impedance element.
ZL(n) = Line impedance of the protected section n
Therefore, any phase-fault for which the impedance seen by the relay
is less than the setting value of pickup impedance for Zone 1 will belong
to Zone 1 and is cleared instantaneously.
The Zone 2 includes the whole length of protected section plus up
to 40% of the next adjacent section. The clearing time for the Zone 2 is
taken as
t2 = t1 + ∆t
where t1 = Clearing time of Zone 1

∆t = 0.5 (time delay step)
t2 = Clearing time of Zone 2.
Analog input
Sample hold
CLOCK
control signal
MPX Microprocessor
address
Input
transformer CPU
1 ch M
Filter SH
U
2 ch L Setting
Filter SH
T Parallel
3 ch DI
Filter SH I ROM condition
P Trip
L DO output
E A/D DMA RAM
X
n ch E
Filter SH R
2 ch 1 ch n ch 1 ch 2 ch n ch
1 2 1 1 1 1 ch Data
ch n ch ch 0 0 0 2 ch Data
0 0 0
10
1 1 1 12 bit
10 11 n ch Data
0 0 0 1 ch Data
1 1 1 2 ch Data
11
t
10
FIGURE 7.33 Principle of digital relay

To ensure the reliability of relay operation for faults close to the

remote end of the line and discrimination with the protection of the adjacent
feeders the pick up impedance for Zone 2 should meet the requirements:
(n) (n) (n) (n + 1)
kdZL 〈 Z pickup (2) ≤ ZL + 0.4 ZL
(n)
where Z pickup (2) = Setting value of pickup impedance for Zone 2
impedance element
(n)
ZL = Line impedance of protected section n
(n + 1)
ZL = Line impedance of the next adjacent section (n + 1) to
the protected section n
kd = Dependability factor of at least 1.2, preferably 1.5.
Zone 1 reach is limited, because of the indeterminate nature of the
fault resistance which may disturb the selective operation for faults at the
end of the protected section and at the beginning of the next section. This
is why Zone 1 is kept nearly 80% of the protected section and Zone 2 protects
only a little portion nearly 20% of the protected section, but this provision
helps in selective operation.
7.11 REQUIREMENTS OF DEFINITE-DISTANCE SCHEMES

Two feeders BO and OA are represented on the complex plane in Fig. 7.34
considered from the point of view of relays at O. The length OA is thus
proportional to the impedance of feeder OA and θ is the impedance angle
of the feeder.
X¢
P P¢
q
O O¢ R
B
FIGURE 7.34 Ideal tripping area
Zone 1 of the protection at O is required to reach up to say P. The

relays at O therefore must operate for all faults between O and P only.
The effect of fault arc resistance is also shown in Fig. 7.34. The fault
resistance increases slightly for faults towards the remote end of the feeder
because less of the total fault current flows through the relay. The shaded
area OPP’O’ is known as fault area. The impedance seen for differing fault
positions will lie inside the shaded area. To protect against all faults the
relay must operate if the measured impedance lies within the fault area.
Hence relays at O with characteristics which completely enclose the fault
area are suitable for protection of feeder OA.
Fault Resistance
Fault resistance consists of two components, the resistance of the
arc and the resistance of earth. The second component is present only when
it involves an earth-fault. In such a case the resistance of earth would mean
the resistance of fault path through the tower, tower footing resistance
and earth return. An approximate value of arc resistance is to be obtained
by empirical relations. Warrington gives the expression
Rarc = 2.9 × 104 L /I1.4 ohms
where L = length of arc in metres in still air
I = fault current in amperes
The arc resistance formula developed in Russia is:
L
Rarc = 1050 ohms
I
L will initially be equal to conductor spacing in the case of phase-
faults and distance from conductor to tower in case of earth faults. With
cross wind, when there is a time delay in fault clearance such as in Zones 2
and 3, the arc is extended considerably, and the resistance is increased. For
Zone 1 where the tripping is instantaneous, the effect of arc resistance is
small and may be neglected except on very short feeders; but for Zones 2
or 3 high velocity winds may cause the relay to underreach seriously.
When both wind and time are involved, Warrington’s formula for
arc resistance is:
50
Rarc = (VL + 47 vt) ohms
I
where VL = nominal system interphase voltage, kV
v = wind velocity in km per hour
t = time in second
I = fault current in amperes
7.12 REACH OF DISTANCE RELAY

A distance relay is set to operate up to a particular value of impedance; for
an impedance greater than this set value the relay should not operate. This
impedance, or the corresponding distance is known as the reach of the
relay.
To convert primary impedance (impedance of the line referred to

the line voltage and current) to a secondary value (line impedance referred
to the relay side) for use in adjusting a distance relay the following relation
is used:
Zsec = Zprim × CT ratio/VT ratio
where the CT ratio is the ratio of the HV phase current to the relay phase
current, and the VT ratio is the ratio of the HV phase-to-phase voltage to
the relay phase-to-phase voltage all under balanced conditions.
The tendency of a distance relay to operate at impedances larger
than its setting value is known as overreach and similarly the tendency to
restrain at the set value of impedance or impedances lower than the set
value is known as underreach.
An important reason for overreach is the presence of d.c. offset in
the fault current wave, as the offset current has a higher peak value than
that of a symmetrical wave for which the relay is set.
The transient overreach is defined as:
Per cent transient overreach
Z os − Z sy
= × 100
Z sy
where Zos = The maximum impedance for which the relay will operate
with an offset current wave, for a given adjustment.
Zsy = The maximum impedance for which the relay will operate
for symmetrical currents for the same adjustment as for Zos.
The transient overreach increases as the system angle tan–1 X/R
increases.
Fig. 7.35 shows the variation of overreach with system angle.
Overreach (per cent)
20
10
10 20 30 40 50 60 70 80 90
System angle degrees
FIGURE 7.35 Overreach characteristics

A distance relay may underreach because of the introduction of fault
resistance as illustrated in Fig. 7.36. Relay at O is set for protection up to P.
Now if a fault at P occurs such that fault resistance (PP’) is high and by adding
this resistance the impedance seen by the relay is OP’ such that P’ lies outside
the operating region of the relay, then the relay does not operate.
X¢
A
P
P¢
R
O O¢
FIGURE 7.36 Underreach of distance relay
Occasionally a fault may occur in the Zone 1 and the relay may begin
to operate. If the fault impedance now increases due to the arc resistance
the total impedance seen by the relay will be the sum of the line impedance
up to the fault and the arc impedance. This sum may be more than the
impedance setting of the relay in which case the first stage operation will
stop and the fault will be cleared as if it is located in the second or third
zone. In order to prevent this, impedance relays are locked once they begin
to operate on the basis of the true impedance up to the fault. Any increase
in impedance due to arc will not affect the relay once it is locked.
7.13 DIGITAL COMPUTATION BY A MICROPROCESSOR

Digital distance relays estimate the location of a fault using instantaneous
values of the voltage and current simultaneously measured (sampled) at
fixed time intervals. The sampling is done several times per cycle. The
sampling rates of various relays may be different, the practical range being
from 4 to 40 samples per cycle. The most typical sampling rate, 12 samples
per cycle, corresponds to the electrical angle of 30° and , for 50 Hz systems,
the time interval of 1.67 ms. At this rate the sampling frequency is 600 per
second. A variety of algorithms have been proposed for processing samples
of current and voltage and several are utilised in the digital relays offered
by the major relay manufacturers. Some algorithms used in distance relays
calculate the fault impedance from the magnitudes and angles of the
fundamental frequency components of input voltage and current. These
fundamental frequency component values are estimated from several
samples of the pre-filtered input voltage and current.
If the measured signal is purely sinusoidal it is enough to have just
two samples taken at a known time interval (other than half a cycle or its
multiple) to be able to find the magnitude and the phase angle of the signal.
However, the practical voltage and current signals available at the relay
location during faults are often heavily distorted and laden with high
frequency noise. Filtering of the signals and mathematical processing of a
greater number of samples are, normally, necessary to make sufficiently
accurate estimate of the fault distance.
The error signals superimposed on the input quantities come from
several sources and are only partly predictable. DC and low frequency
transients are produced by instrument transformers and free oscillations
between system inductances and capacitances, higher harmonics are
generated in transformers , generators, static VAR compensators and non-
linear loads, non-harmonic high frequency random noise is produced by
multiple reflections of travelling waves initiated by faults and switching
operations. On top of these primary distortions there are also non-linearities
of the instrument transformers and interferences generated in the secondary
systems. The ability of the relay to extract the fundamental frequency
components from the noisy signals greatly increases with the time available
for filtering and the number of samples available for mathematical
processing. The rule known from application of all previous technologies
that the high accuracy and the high operating speed are conflicting
requirements applies also to digital technology and the practical solutions
are always results of compromises between these requirements.
In all practical digital relays the input quantity signals, voltage and
current, are passed through analog filters to eliminate any components of
frequencies higher than half of the sampling frequency to prevent errors
called aliasing. After the anti-aliasing filters the signals are sampled and
converted to digital form.
A simplified block diagram of a digital relay is shown in Fig. 7.33.
The sampling is controlled by a clock. The sampled values are held constant
for a short-time by sample-hold modules in order to provide for calculation
the voltage and current values corresponding nearly to the same instant
although the sampling takes place at a slightly different time. Sampling
exactly at the same time is possible but requires multiplication of A/D
(Analog to Digital) converters which is more expensive than using one
converter fed from multiplexed inputs as shown in the diagram.
Further processing of the digital values representing voltage and
current at the sampling instants may lead to estimation of magnitudes
and angles (vectors) of their fundamental frequency components and
subsequently, to determination of the fault impedance from these
components. These methods may utilise fast but less accurate algorithms
based on a short data window e.g., three samples, or more accurate longer
window algorithms. Some relays have both fast and slow algorithms
running in parallel to increase the speed for faults deep within Zone 1 and
to provide accurate borders between the zones.
A simple way of determination of the distance to the fault from three

samples may be illustrated in Mann-Morrison algorithm as follows:
For any sinusoidal wave-form the peak value can be determined
from two instantaneous values measured at 90° interval. Assume the period
of sinusoid v(t) to be T = 20 ms and that three samples v(t – ∆t), v(t) and
v (t + ∆t) have been taken at short intervals ∆t, say ∆t = 1.67 ms.
If v(t) = Vp sin (ωt + φ)
then v (t + T/4) = Vp sin (ωt + φ + 90°) = Vp cos (ωt + φ)
and Vp2 = [Vp sin (ωt + φ)]2 + [Vp cos (ωt + φ)]2
= [v(t)]2 + [v(t + T/4)]2
Note, however, that the second component (t + T/4) can be estimated
from the samples v(t – ∆t) and v(t + ∆t) measured, respectively, just before
and just after the instant t. These samples may be used for calculation of
derivative of function v(t):
v’(t) = ω × Vp cos (ωt + φ) = ω × v (t + T/4)
which, using trapezoidal rule, can be estimated as
v’(t) = [v(t + Dt) – v(t – Dt)]/(2 × Dt)
and may be used for finding
v (t + T/4) = Vp cos (ωt + φ) = v’(t)/ω
Eventually the peak value can be determined as:
2 2
Vp = [ Vp sin(ωt + φ)] + [Vp cos (ωt + φ )]
2 2
= [ V (t )] + [v’(t )/ω ]
[v (t + ∆t) − v(t − ∆(t)]
where v’(t) = .
2 × ∆t
The phase angle of the sinusoid at instant t is:
Vp sin (ωt + φ) v’(t) × ω
ωt + φ = arc tan =
Vp cos (ωt + φ) v(t)
It can be seen from the above that, without excessive calculations,
the magnitude Vp and angle ωt + φ of the signal V can be determined within
time interval of (t + ∆t) – (t – ∆t) = 2 × ∆t, in this case 2 ∆t = 3.34 ms provided
the signal can be assumed to be undistorted and sinusoidal.
Calculating the square root for voltage and current separately is not
necessary if the impedance is the required result. Having evaluated the
squares of voltage and current magnitudes one can calculate the square of
impedance magnitude by division:
Vp2 v(t )2 + [v’(t)/ω]2
Z2 = =
I 2p i(t )2 + [i’(t)/ω ]2
The impedance phase angle can be found from the formula:

φ = arc tan [ω × i(t)/i’(t)] – arc tan [ω × v(t)/v’(t)].
The above algorithm is very sensitive to distortion of the wave-form
and its use would require thorough filtering. Off course a determination
of the magnitude and angle from just two samples would be much more
inaccurate.
7.13.1 Full Cycle Fourier Algorithm
Some of the popular algorithms of higher accuracy are based on Fourier
Filters. The Full Cycle Fourier Algorithm correlates the fault voltage and
current signals with sine and cosine weighting factors. This technique
generates two orthogonal components for each input voltage and current
signal.
Note that for any harmonic
In sin (nωt + φn) = Inr cos (nωt) + Inx sin (nωt)
2
where In = ( Inr 2
+ I nx ) ; tan φn = Inr/Inx ; ω = 2πf = 2π/T;
nω = ωn = 2π/Tn
where n = the order of the harmonic, for n = 1 (the fundamental frequency
of 50 Hz) T = 20 ms, for n = 2 (second harmonic) Tn= 10 ms etc.
The coefficients Ir and Ix can be calculated from the following
formulas:
Fundamental frequency:
T
Ir = 2/T zt=0
i(t ) cos (ωt) dt
T
Ix = 2/T zt=0
i(t ) sin (ωt) dt
The nth harmonics:

T
Inr = 2/T zt=0
i(t ) cos (nωt) dt
T
Inx = 2/T zt=0
i(t ) sin (nωt) dt
The digital approximation of the integrals for the fundamental

frequency above may be:
K
Ir = 2/K ∑ i(k) cos (2πk/K)
k=1
K
Ix = 2/K ∑ i(k ) sin (2πk/K)
k =1
where K is the number of samples per cycle.

It can be shown that these components can be combined in each
phase according to the following equations to yield estimates of the
resistance and reactance to the fault location.
R = (Vr × Ir + Vx × Ix ) / (Ir2 + Ix2)
X = (Vr × Ir – Vx × Ix ) / (Ir2 + Ix2)
where R is the line resistance to the fault point
X is the line reactance to the fault point
Vr is the real voltage component
Vx is the imaginary voltage component
Ir is the real current component
Ix is the imaginary current component
This algorithm is evaluated after each new sample i.e. the data
window slides along the input data stream producing a new estimate of R
and L for each new set of input samples.
This algorithm is insensitive to any harmonic content of the input
signals and to the d.c. component of the signals. This ‘rejection‘ of all
harmonics is only obtained if, at least, one whole cycle of the fundamental
frequency is included in the sample data window. It may be noted that a
noise of high frequency which is not an integer multiple of the fundamental
frequency is not rejected by Fourier algorithm and may cause error in
distance estimation.
7.13.2 Differential-Equation Algorithms
This group of algorithms is based on a single phase model of the faulted
line. The differential equation relating the voltage and current to the line
parameters is:
v(t) = Ri(t) + L[di(t)/dt]
Solving this type of equation would produce the required resistance
R and inductance L which are both proportional to the distance to the fault.
McInnes and Morrison propose solving the equation by integrating the

voltage over two consecutive intervals : t0 – t1 and t1 – t2
t1 t1
z t0
t2
v(t) dt = R
z
t0
t2
i(t) dt + L[i(t1) – i(t0)]
z t1
v(t) dt = R
z
t1
i(t) dt + L[i(t2) – i(t1)]
The integrals above may be estimated from the samples taken at

instants t0, t1, t2 using the trapezoidal rule of integration:
t1
zt0
v(t) dt = ∆t/2[v(t1) + v(t0)]
For the samples of voltage and current taken at any three consecutive
instants k, k + 1, k + 2 the line equations may be written as:
LM ∆t (i k+1 + ik )/2 (i k + 1 − i k ) OP LR O = LM ∆t (v k +1 + v k )/2 OP
N ∆t (i
k+2 + i k + 1 )/2 (ik + 2 − ik + 1 ) Q MNL PQ N∆t (v k+2 + v k + 1 )/2 Q
The fault resistance and inductance from the above equations are:
( v k + 1 + v k ) (i k + 2 − i k + 1 ) − ( v k + 2 + v k + 1 ) (i k + 1 − i k )
R=
(i k + 1 + i k ) (i k + 2 − i k + 1 ) − (i k + 2 + i k + 1 ) ( i k + 1 − i k )
∆t (ik + 1 + ik ) ( vk + 2 + vk + 1 ) − (ik + 2 + ik + 1 ) (vk + 1 + vk )

L= ×
2 (ik + 1 + ik ) (ik + 2 − ik + 1 ) − (ik + 2 + ik + 1 ) (ik + 1 − ik )
Similar formulas for R and L can be obtained in terms of the signal

values and their first order derivatives or in terms of the first and second
derivatives of the signals. For example, if the values of signals v1 and i1
and their derivatives v1′ and i1′ at instant t1, and also values v2 and i2 and
their derivatives at instant t2 are available then the resistance and inductance
are:
v . i ′ − v2 . i1′
R= 1 2
i1 . i2′ − i2 . i2′
v2 . i1′ − v1 . i2′
L=
i1 . i2′ − i2 . i1′
With use of the first and second order derivatives the formulas
become:
v . i ′′− v2 . i1′′
R= 1 2
i1 . i 2′′− i2′ . i1′′
v2 ’. i1′ − v1′ . i 2′
L=
i1 ’. i2′′− i 2′ . i1′′
The derivatives of the signal used in the above equation may be

determined from three or more samples. The method of this determination
affects the frequency response of the relay. The differential equation
algorithms reject the d.c. components of the input signals and can be made
reject a particular frequency but, generally, higher frequency distortion
may cause considerable errors in their measurement. The use of second
order derivative method makes the relay more tolerant to lower frequency
interference which, together with the low pass antialiasing filters makes
the sensitivity window of the relay reasonably small.
Protection design engineer should be aware of the performance
characteristics and sensitivity of the relays resulting from their principal
signal processing method and should use it to the advantage of the
protection scheme. The characteristics of a practical relay result from the
combination of the analog and digital filters, impedance algorithms and
the operating criteria programs. This combination may mask to a degree
the generic performance characteristics of the algorithms but it has been
confirmed by numerous simulation tests in the laboratory that the
knowledge of the relay algorithms makes protection operation a lot more
predictable.
CHAPTER
Differential Protection
8.1 INTRODUCTION
Differential protection is a unit type protection scheme that uses a
comparison between the current entering the protected zone and the current
leaving the protected zone, to determine if there is a fault. The comparison
can be made on the magnitude of the difference between the currents, the
phase difference or a combination of both.
Differential protection, in its various forms, is widely used in
protection systems for generators, transformers, busbars and lines.
The objective of this chapter is to examine the operating principles
and design of the various schemes.
The principle forms to be discussed are:
• circulating current differential protection (low impedance),
• biased differential protection,
• high impedance current differential,
• pilot wire, and
• phase comparison.
8.2 CIRCULATING CURRENT DIFFERENTIAL (LOW

IMPEDANCE)
The low impedance circulating current scheme is the basic form of
differential protection. It is applied widely in subtransmission systems and
to a lesser extent in transmission systems as busbar protection.
The scheme is simply a summation, on a per phase basis, of the

current transformers associated with each item of plant in the protected
zone. An overcurrent relay is connected across the CT summation junction
and is set to operate when an internal fault results in current flowing in
the relay circuit. Fig. 8.1 shows current flows in the CT and relay circuits
for an internal and external fault. It can be seen that, for normal load
currents or an external fault, currents in the CT secondary circuits ‘circulate’
around the secondary loop and summate to zero at the junction. For an
internal fault, currents will flow into the fault from the primary sources
and the secondary currents will add to flow through the relay. Note the
polarity marks shown for the CTs and the connections to give the correct
summation of the currents.
R R R
(a) External fault
(b) Internal fault
FIGURE 8.1 Low impedance current differential
For the scheme to operate satisfactorily, it is necessary to ensure that

any errors that cause unbalance of the CT secondary summation, for an
DIFFERENTIAL PROTECTION 211
external fault, will not result in operation of the relay. All current
transformers must be connected on the same ratio, have similar
performance specifications and similar lead burdens between the CT and
the secondary summation point. To achieve the latter, it is usual to make
the CT summation as close to the primary plant as possible (e.g., in the
switchyard close the CBs). The summation is then cabled to the protection
panel.
In practice, there will always be some ‘spill’ current under external
fault conditions, particularly during the sub-transient and transient periods
of the fault. The precise level of these currents is difficult to predict so it is
usual to introduce a time delay into the relay and ensure that the minimum
operating level is not too sensitive. An inverse overcurrent relay, set on a
Short Time Multiplier Setting, gives a satisfactory performance. Even with
the use of an inverse overcurrent relay, the operating times that can be
achieved with this protection are far superior to what would be the case if
remote overcurrent or distance protections were relied upon to clear the
fault.
One difficulty that can arise with the application of low impedance
schemes is CT saturation. The most likely situation where this will occur is
where one of the primary circuits in the protected zone contributes a higher
level of current to an external fault than other circuits in the zone. If
saturation does occur, the CT will not contribute to the summation and
the magnetising impedance will appear as a zero impedance. As a result,
the circuit containing the saturated CT (including its cabling) will appear
as a resistance across the summation, in parallel with the relay. The
summated current from the unsaturated CTs will divide between the relay
and the saturated CT circuit according to the relative resistances of the
relay circuit and the saturated CT circuit i.e., spill current will flow through
the relay and most likely cause incorrect operation.
Referring to Fig. 8.2, it can be seen that the current through the relay
can be represented by:
I sec (Rlb + Rctb)
IRelay =
R Relay
From this formula, it can be seen that if RRelay is small compared
with (Rlb + Rctb) incorrect operation is likely to occur . One technique to
overcome the problem is to connect a ‘stabilising’ resistor series with the
relay. In the extreme, this approaches the high impedance scheme described
in section 8.4 and is not frequently seen in Australian practice; the preference
being to use the high impedance scheme.
a b
Rcta Rla Rlb Rctb
Relay
FIGURE 8.2 Circulating current scheme (one CT saturated)
In making the decision between low impedance and high impedance

schemes, however, it must be remembered that the high impedance scheme
requires dedicated CTs and these are not always available, particularly in
a subtransmission system.
8.3 BIASED DIFFERENTIAL PROTECTION

To improve the stability of a differential scheme to unbalances in CT
secondary currents that can occur under external fault conditions, and to
allow the operating time delay to be reduced, the concept of ‘bias’ is
introduced.
Biased differential protection, in it’s usual form, is designed for two
or three inputs and is most commonly used for transformer protection.
The bias feature overcomes the difficulty with power transformers of
achieving a good match between CTs on the different voltage levels and
the spill currents in the differential circuit that arise from tap changing.
There are also some forms of proprietary biased schemes available
for bus bar protection.
I1 I2
Bias Bias
Operate
(a) Biased differential relay

tes
ra
8
ope
n
tio
ec
ot ea
6 Pr ar
e
tiv
Current, I1 ra
ope
4 In
tes
e ra
op
on
2
ecti
ot
Pr
0 2 4 6 8
Current, I2 (% rated secondary current)
(b) Typical biased characteristic relay
FIGURE 8.3 Biased differential protection
The concept of the biased relay is as shown in Fig. 8.3(a). This form,
known as percentage bias, operates on the principle that if the two currents
entering and leaving the protected zone are I1 and I2, then:
The operating quantity = K1 (I1 – I2)
The biasing quantity = K2 (I1 + I2)
Suitable choice of constants K1 and K2 ensures stability for external
fault conditions despite measurement errors, while still retaining adequate
sensitivity for operation on internal faults. Fig. 8.3(b) shows a typical
characteristic.
In electromechanical terms the relay can be thought of as having
bias coils in the circulating current path that exert a restraining torque on
relay operation and on coil in the operating leg that produces an operating
torque.
Where the protection is used on transformers, the CTs associated
with star connected windings are usually connected in delta and those for
delta connected windings in star. This corrects the power transformer line
current phase shift and eliminates the zero sequence currents on the star
side which might otherwise upset the stability due to the lack of a
corresponding zero sequence component on the delta side. Bias settings
are chosen to ensure that the relay will remain stable with the transformer
at the extremes of the tapping range.
8.4 HIGH IMPEDANCE CURRENT DIFFERENTIAL

With the low impedance circulating current scheme, described in section
8.3, it was seen that it is not practical to maintain stability under external
fault conditions, if CT saturation occurs. High impedance current

differential protection was developed to overcome this problem and to
achieve high speed operation is required for internal faults. In principle,
the scheme involves increasing the stabilising resistor to a value that results
in the operating voltage having a margin over the voltage developed across
the circuit with the saturated CT.
High impedance schemes are widely used for busbar protection and
zone protections (e.g., the connections to a power transformer) on EHV
and subtransmission systems and for generator differential protection.
The scheme makes use of the principle, described earlier, that the
secondary winding of a saturated CT will not contribute to the CT
summation and its secondary winding will represent a resistance connected
in parallel with the CT summation and the differential relay
A voltage measuring relay (usually an instantaneous overcurrent
relay with a series resistor) is connected across the CT summation point.
To ensure stability for external faults, the relay is set to operate at a voltage
approximately 1.1 times the voltage that will appear across the summation
with one of the CTs which is saturated under external fault conditions.
The setting calculation is performed on the highest CT + lead resistance,
connected to the summation i.e., the CT furthest from the summation point.
On an internal, fault, the summation will see the relay impedance
and CTs will develop a voltage across the differential relay. The operating
voltage is usually set to 0.5 Vk , to ensure definite operation.
Under internal fault, conditions, the conditions in the CT secondary
circuit is similar to that of an open circuit operation and results in the
potential for very high secondary voltages to be developed. The voltage
waveform will be distorted but the peak value may be many times the
nominal saturation value.
A guide to the possible peak voltage can be obtained from the
following formulae:
Vp = 2 2 Vk ( Vf − Vk ) ...(1)
where Vp = peak voltage developed

Vk = saturation voltage (knee point voltage)
Vf = prospective voltage in the absence of saturation
If
Vp = 2 Vk ...(2)
I ek
where If = fault current

Iek = exciting current at knee point voltage
Vk = knee point voltage
Equation (1) applies when there is a finite burden resistance
(although high) but it does not hold for the open circuit condition.
Equation (2) applies for the open circuit secondary condition. Any
burden across the secondary will reduce the voltage.
It is usual to limit the voltage across the relay circuit with the
connection of a non-linear resistor. A ceramic resistor (commonly referred
to by the trade name (Metrosil) having a characteristic V = Clβ is used,
where C is a constant depending on the dimensions and β is a constant in
the range 0.2–0.25.
Because of the risk of high voltages appearing on the secondary
circuits it is necessary to ensure that the secondary wiring and isolation
equipment has sufficient insulation and that procedures are in place to
ensure that staff access to the equipment for testing and maintenance is
safe.
8.5 PILOT WIRE PROTECTION

The limit to the application of the differential schemes described above is
the length of cabling required to summate the CT secondary currents. In
practice, this limits the application of these schemes to plant and
connections within stations.
To remove this limitation pilot wire protection schemes were
developed, where the CT secondary currents are transformed to a level
that can be transmitted over telecommunications cables. Comparison of
the phase and magnitude of quantities representing the local and the remote
currents are then made at each terminal, i.e., comparisons are made at each
end of the protected zone in contrast to the direct comparison of the current
differential schemes described earlier. Bias is included in the scheme design
to provide stability for differences in CT characteristics.
Insulation, current carrying limitations and capacitance of
telecommunications cables, limit the length of cable that can be used to
around 30 km. As a result, this form of protection is usually used for short
transmission and subtransmission lines. These schemes are well suited to
this application as they provide a high speed unit type scheme compared
with the alternative of distance or overcurrent protections which can be
difficult to apply to short lines and have longer operating time.
Two basic schemes are generally used:

• Circulating current Fig. 8.4(a), where current circulates in the pilot
circuit under load or external fault conditions and through the
operating circuits of the relays at each end of the protected zone
for internal faults.
• Balanced voltage scheme Fig. 8.4(b), where the voltage on the pilot
circuits are opposed, under load or external fault conditions,
resulting in zero current in the pilot circuit and relay operating
circuits. An internal fault results in a phase reversal in the voltage
at one end of the pilot circuit with the result that current flows in
the pilots and the operating circuit of the relay.
It will be noted that the schemes shown in Fig. 8.4 use one pilot
circuit to cover all fault conditions. A summation transformer, Fig. 8.4(c),
is used to combine the three phases into a single quantity for the
comparison. This is done to minimise the use of pilot cable circuits, as they
are relatively expensive to provide.
Line
I I
R R
Summation V Pilots Summation
transformer V transformer
Relay Relay
(a) Circulating current scheme
Line
I I I
Summation V Pilots V Summation

R R
transformer I I transformer
Relay Relay
(b) Balanced voltage sheme

A
T1
B
T2
N
(c) Summation transformer
FIGURE 8.4 Pilot wire protection
The interphase sections of the winding A-B and B-C are often given
equal numbers of turns, the neutral end of the winding C-N will generally
have a greater number of turns.
Unbalanced fault currents will then energise different number of
turns according to which phase is faulted. This leads to relay settings which
are in inverse ratio to the number of turns involved. If the relay has a setting
of 100% for an A-B fault, the following proportionate phase settings will
apply.
Phase A-C 100%
Phase A-C 50%
Phase A-B 100%
Three phase 58%
The earth fault settings will depend on the relative number of turns
in section C-N but will also depend on which phase is faulted.
8.6 PHASE COMPARISON PROTECTION

Phase comparison is a term used to describe a differential scheme that
compares the phase relationship of the currents at each end of the protected
zone and initiates tripping when the phase difference exceeds the tripping
criteria. These schemes were principally designed to use high frequency
bearers such as power line carrier on longer transmission lines. The carrier
signal is modulated to represent the power system current and transmitted
to the remote end where the phase relationship is compared with the local
current. The schemes were used only to a limited extent in Australia
because of their relative complexity and difficulties with maintaining
reliable performance compared with distance schemes.
As shown in Fig. 8.5, modulation is of the all or nothing type giving
half-cycle pulses of carrier signal interspersed with half periods of zero
signal.
With an external fault, the transmission from one end fills in the
gaps at the other end resulting in a continuous signal on the line, when
observed by the measuring equipment at one end. Signal attenuation does
not affect the measurement. The continuous signal blocks operation. For
an internal, fault the resultant signal on the line has half-period gaps during
which the tripping function is initiated. Time delay circuits are included in
the protection to compensate for the transmission time of the carrier (0.33
milliseconds or 6 electrical degrees per 100 km of line for power line carrier).
Starting relays using current level, rate of change of current or
impedance measurement are used to initiate the comparison
(telecommunications regulations usually do not allow continuous
transmission of power line carrier) and to interlock the tripping circuit.
The tripping function is set so that operation will not occur until the
gap corresponds to a specific phase difference between the currents at each
end of the zone - usually in the range 18°–30°.
As with pilot wire schemes, separate comparison for each phase is
usually uneconomical and some arrangement for combining the phase
currents are made.
Modern developments, with the advent of digital transmission
systems and particularly the availability of fibre optic transmission systems
on transmission lines, has generated a renewed interest and there are
schemes available on the market.
Internal fault Internal fault
30° tripping 180° tripping
Through fault angle angle
Fault current at station A
Effect of squaring circuit
Hf signal transmission at A
Fault current at station B
Hf signal transmission at B
Combined hf signals on line

at station A
Blocking voltage from detector

stage
Current in trip relay
FIGURE 8.5 Phase comparison protection

CHAPTER
9
Unit, Remote and Back Up
Protection
9.1 UNIT AND NON-UNIT PROTECTION SCHEMES

Protection schemes fall into two basic categories, these are the ‘Unit
Schemes’ and the ‘Non-Unit Schemes’.
The basic features of the two categories are as follows:
♦ Unit Schemes — These schemes are used to protect a
defined/discrete zone, usually their
boundaries are defined by the Current
Transformers.
Advantages: No grading is required with
adjacent protections nor with the load
current, they are fast operating and
sensitive.
Disadvantages: Do not provide back up to
adjoining protection zones, therefore
duplication of protections is generally
necessary.
♦ Non-Unit Schemes— These schemes do not protect discrete zones
but they overlap with other protection
zones. As a result, they need to discriminate
with other protections and distinguish
between the load current and the fault
current.
Advantages: Ability to discriminate over
other relay groups allows back up to be
provided to the relay nearest the fault if that
relay has failed to operate.
Disadvantages: Time/current/impedance grading required to allow

for discrimination with other protections, therefore slowest fault clearing
times are going to be at the highest fault level locations.
The following compares in more detail some of the main features of
the two:
Feature Unit Protection Non-Unit Protection

Scheme Scheme
1. Principle of Operates if the input and Operates if a measured
operation output quantities of the quantity (volts, amperes,
protected zone are not ohms, etc.) is over/under or
equal (e.g., a differential departs from a set level (e.g.,
type protection, etc). under/over voltage
overcurrent, impedance
protection, etc.).
2. Sensitivity to The magnitude of the differ- In the case of simple over-
short circuits ence between the input/ current protection the sensi-
between output quantities required tivity is limited by maximum
conductors for operation can be less load, i.e., the fault which can
(i.e., Multi- than the maximum load be detected is greater than
phase faults) current (i.e., good sensitivity load current. Impedance or
is possible). distance protections do
allow sensitivities which
are better than maximum
load current.
3. Sensitivity to Virtually only limited As for the Unit Scheme
short circuits by the sensitivity which
between can be built into the
conductors and scheme—therefore good
ground (i.e., sensitivity is possible.
Phase-to-Earth
faults)
4. Speed of Generally can be made Generally to obtain
operation very fast and basically discrimination with other
is an inherent property protective devices it is
of the scheme itself. necessary for time delays
Operating times of to be introduced.
20–30 milliseconds are Operating times greater
common. than about 0.5 second are
usually appropriate.
5. Facility for By its nature it cannot Back up to protective
providing back- provide back up for equipment at other
up for other failure of protective locations is possible
UNIT, REMOTE AND BACK UP PROTECTION 221
protection equipment at other within the constraints

elements locations. imposed by loading, system
interconnection and earthing.
6. Cost Generally somewhat higher This varies with the sophist-
than for non-unit schemes ication of the device, from
but can be significantly being relatively cheap for a
higher if interstation sign- single quantity device (e.g.,
alling is necessary. overcurrent and overvoltage)
to quite costly for a complex
distance relay.
7. Range of May find application at all The simpler types of equip-
application system levels but mostly ment or schemes would not
at the main transmission generally be found on the
level. (For some types, the main transmission system
cost may be difficult to but the complex schemes
justify at the lower voltage (e.g., distance relays) are
levels). found at both main and sub-
transmission voltage levels.
From the above comparison, it can be seen that each category has its
advantages/ disadvantages for particular applications but overall cost is
generally the biggest factor in applying unit-type schemes, particularly
where it becomes necessary to provide interstation signalling.
Distance protection does overcome to a large extent one of the main
disadvantages of non-unit schemes (i.e., limited sensitivity) and therefore
it is becoming the most common choice for the protection of lines in the
sub-transmission system along with almost universal use, with associated
signalling, at the main transmission level for which it was originally
developed.
The typical application of Unit and Non-Unit type Schemes are
demonstrated in Figs. 9.1 and 9.2. The typical type of schemes used are as
follows:
Unit Type Schemes
• Pilot wire
• Transformer differential
• Bus protection (high impedance, medium impedance, low
impedance)
• Phase comparison
• Distance protection with associated signalling
• Directional comparison
• Restricted earth fault etc.
220 kV Bus
Microwave
PLC
optic fiber
66 kV Bus
Pilot
wire
Unit schemes
Non-unit schemes
FIGURE 9.1 Unit and non-unit schemes

G GEN
Transformer
protection
zone
GEN TR
High impedence
protection zone
No. 1 Bus
LINE BUSBAR protection

protection zones
zone
No. 2 Bus
LINE
FIGURE 9.2 Unit protection type schemes
Non-Unit Schemes
Feeder Overcurrent
Transformer H.T. Overcurrent
Bus Overload
Earth Leakage Protection
Distance Protection without Associated Signalling
Fuses etc.
9.2 REMOTE AND LOCAL BACK UP PROTECTION

With any protection system it is generally accepted that any device other
than a fuse can fail to operate. To reduce the risk of a fault not being cleared,
it is common practice to provide protection systems that cater (as a
minimum requirement) for the failure of at least one element and still ensure
that the fault is cleared. On a world wide basis, this has proved to be a
satisfactory assumption.
As a consequence back up protection is applied primarily to cover
for the:
• failure of the circuit breaker associated with the primary protection
zone,
• ‘blind spots’ of the system which are not included in any zone of
the main protection because of the location of the current or
voltage transformers and,
• failure of the main protection systems including the measuring
current transformers and voltage transformers.
On simple radial systems, each circuit breaker and its associated
protective relays provide back up for the failure of the circuit breaker or
the main protection at the next station. The failure could result from the
relay being defective, the station battery being flat, or the circuit breaker
being unable to interrupt the fault. This type of arrangement is termed
remote back up and is provided by time-graded protections (current or
impedance). Remote back up protection has the advantage that it provides
back up protection for both relays and circuit breakers.
However, remote back up has the following disadvantages:
— The protection providing the remote back up can be very slow to
operate since it is required to discriminate with other protections.
— It may not be possible to set the back up protection so that it can
detect the fault condition at the remote location.
For example:
• the load currents could be greater than the fault currents.
• in a power system with interconnections, parallel circuits and
infeeds from intermediate current sources could affect the
performance of time-graded type protections, (e.g., increasing or
decreasing the reach of the relays). In these cases, it may again
not be practical for the protection to be set to detect the fault at
the remote location.
Obviously remote back up may not be practical for some
applications, for such situations another form of back up has been
developed and this is termed local back up.
Local back up is provided in two parts:

• Back up to cover Protection Scheme Failure
— provision of two independent protections (In general local
back up is not duplicated, however, this is dependent on the
philosophy/practice of the relevant organisation).
— provision of independent current transformers, circuit
breaker trip coils, d.c. supplies, voltage transformer supplies
(generally by segregating the circuit near the voltage
transformer).
• Back up to cover circuit breaker failure or provision of ‘Blind Spot’
protection.
Local back up for circuit breaker failure or ‘blind spot’ protection, is
provided by determining that a circuit breaker has failed to trip or isolate
a fault, and then in turn initiate tripping of all circuit breakers that are
adjacent to the failed circuit breaker to clear the fault.
Briefly the advantages and disadvantages of the remote and local
back up are as follows:
Remote Back up
Advantages Disadvantages
• inexpensive • slow
• backs up all elements • less sensitive
including battery • difficult to set with infeeds and
• simple coordination with other
protections
• entire station tripped
Local Back up
Advantages Disadvantages
• faster clearance • more expensive
• more sensitive • more complex maintenance
• easier to set • two independent primary
• minimum system protections are required
interrupted • incorrect operation can shut
down other plant
Remote Back up
Remote back up as indicated previously can be provided by time
discrimination protections, i.e., overcurrent and distance are two typical
protections that are used for this function. Remote back up is generally
applied in radial systems.
A simple example of remote back up is demonstrated by the system
shown in Fig. 9.3. In this example, the protection at ‘B’ provides the back
up for ‘C’ and ‘A’ provides the back up to ‘B’.
Clearance
time
Dt Dt
A B C
I I I
R1 R2 R3
Overcurrent protection
Clearance
time
0.5 S
Dt Dt
Zone 1 Zone 1 Zone 1
0
A B C
Z Z Z
Distance protection
Dt-discrimination margin
FIGURE 9.3 Remote back up
Local Back up
Back up to Protection Scheme Failure
The back up scheme to be provided must ensure, as a minimum,
that the performance and reliability of the overall protection system to
clear a fault is maintained for the failure of any single element.
To achieve the above requirement, the practice of organisations is,
in general, to have a philosophy similar to the following:
• Provision of two independent protection relay equipments on all
protection zones (e.g., lines, busbars, transformers etc.).
• Where possible, duplicated primary protections on a plant item
should use, equipment operating on different measuring
principles and/or design. The exception being duplicate
electromechanical protection schemes where in-service experience
has shown that such relays have displayed a high level of stability
and reliability.
• Provision of independent inter-station signalling links, where
required, preferably on different routes for the two independent
protection schemes.
• Provision of independent d.c. supplies to the two independent

protections.
• Provision of independent current transformers to the two
independent protections.
• Provision of independent voltage supplies to the two independent
protections.
• Relay and other protective equipment performance requirements
required to comply with British, IEC or Australian Standards or
other as appropriate.
• Regular maintenance and testing of equipment.
Typical examples of design practices to provide back up for the
failure of protection are shown in Figs. 9.4 and 9.5.
Backup to Circuit Breaker Failure
In the case of circuit breaker failure, it is not economically viable to
provide duplicate high voltage circuit breakers for back up. Even with the
provision of duplicate trip coils on circuit breakers, electrical or mechanical
failure with the circuit breaker can prevent the clearance of a fault.
To ensure the clearance of the fault in the event of circuit breaker
failure, local back up is provided. Although there may be differences with
the specific circuit breaker fail schemes logics used from organisation to
organisation and from country to country, the principle of operation is in
general similar in all cases. It is normal practice to utilise the main protection
in conjunction with a current check relay (or in some applications the use
of the circuit breaker auxiliary contact) to detect the failure of the circuit
breaker and initiate a timing relay. If the fault is not cleared by the primary
scheme before the timing relay times out, then all the adjacent circuit
breakers are tripped to clear the fault.
The current check relay used to detect the failure of the circuit breaker
is an instantaneous type overcurrent relay. To ensure that the circuit breaker
scheme functions correctly, it is essential that the current check relay
sensitivity selected (setting) is below the minimum level of fault current.
Generally, two arrangements of incorporating circuit breaker failure
relays are used, these are:
• Arrangement ‘One’ (Refer to Fig. 9.6(a) for typical logic).
With this arrangement, the circuit breaker failure scheme is not
enabled until the primary protection, that initiates the trip of that
circuit breaker, has operated.
228
Battery
Line X Y Y CB fail
X
Relay Trip Relay Trip (back up)
Y
To other
X To other X prots
X prots
Comm
Comm
Seal in Seal in Seal in
relay relay relay
X line Y line CB fail
protection protection protection
Trip Trip Trip
POWER SYSTEM PROTECTION AND COMMUNICATIONS

X Bus
protection relay relay relay
Y CTs CB fail CB fail
CB fail initiate initiate
aux RLY aux RLY
If applicable
Line X Initiate tirp
protection Y of bus via
CTs X bus prot.
Trip relay
X CB CB ‘A’ SW
trip Initiate trip
coil of adjacent
Y CB CB ‘A’ SW CB for 1.5 CB
trip arrangement
coil via X trip coil
X protection Y protection Circuit breaker fail

Bus
FIGURE 9.4 Design practice providing back up for failure of protection-partial duplication
UNIT, REMOTE AND BACK UP PROTECTION
X Battery Y Battery
Line
X X Y Y
Y Relay Trip Relay Trip
To other
To other Y prots
X X prots
Comm
Comm
If applicable Seal in Seal in If applicable Seal in Seal in
relay relay relay relay
X line X CB Y line Y CB
protection fail protection fail
Bus Trip Trip

X relay relay
protection
Y CTs
CB fail Trip Trip
relay relay
Y Prot &
Line Y CB fail
protection
If applicable
If applicable
CTs X Prot & Initiate trip Initiate trip
X CB fail of bus via of bus via
X bus prot. Y bus prot.
X CB CB ‘A’ SW Trip relay
trip Trip relay
coil
Y CB CB ‘A’ SW
trip
coil Note: Note:
CB fail logic not included Initiate trip CB fail logic not included Initiate trip
for failure of CB for: of adjacent for failure of CB for: of adjacent
‘X’ bus protection operation CB for 1.5 CB ‘Y’ bus protection operation CB for 1.5 CB
Recieve X back up trip from arrangement Recieve Y back up trip from arrangement
remote end of line. via X trip coil remote end of line. via Y trip coil
Bus
229
FIGURE 9.5 Design practice providing back up for failure of protection-full duplication
DC supply
+ –
Timer
Protection
Current
check
Trip A
relay
Preferred
circuit Trip
appropriate
DC supply CBs or
+ – send remote trip
Current
check
Protection
Timer
Trip A
relay
Alternative
circuit Trip
appropriate
CBs or
send remote trip
FIGURE 9.6 (a) Circuit breaker failure scheme arrangement ‘one’
The advantage of this scheme is that the circuit breaker failure

relay settings can be below load current without impacting on
the security of the system.
However, the disadvantage of this arrangement is that the circuit
breaker failure scheme remains in a de-energised state until a
protection operation has occurred. The reliability of the scheme
could be an unknown quantity.
• Arrangement ‘Two’ [Refer to Fig. 9.6(b) for typical logic]
With the arrangement, the circuit breaker failure scheme is always
in an energised state. In this case, the operation of the overcurrent
check relay feature is independent of the primary protection.
The disadvantage of this arrangement is that the circuit breaker
failure relay settings usually need to be set above load current for system
security reasons. As a consequence it may become difficult to set the relay
to cover some contingencies for circuit breaker failure, in particular if the
fault current is less than the load current.
DC supply
+ –
Current
check
Timer
Protection
Trip A
relay
Trip
appropriate
CBs or
send remote trip
FIGURE 9.6 (b) Circuit breaker failure scheme arrangment ‘two’
Where the fault current is less than the load current, different
arrangements may be necessary, these would include the use of high set
current checks with other sensitive fault detectors interlocked by low set
overcurrent check relays.
In these applications the high set current check is set above the load
and would cover the high level faults, the low set current check arrangement
would cover low level fault conditions. Typical examples of local back up
protection used on EHV systems refer to Fig. 9.7.
Local back up EHV lines (Three pole trip)
RWB AND Timer

Line prot
Trip all
bus CBs
R OR
HOC
B
LOC N
AND Timer Trip adjacent CBs

RWB initiate remote trip
Bus prot
as required
(a)
Local back up EHV lines (Initial three pole trip future provision
for single pole trip)
AND
R
OR Timer
Line
W Trip all
prot
bus CBs
S
OR
R
HOC W
OR
S
Deleted when single
pole trip required
LOC N
RWB initiate remote trip
Bus prot
as required
(b)
Local back up EHV lines (Single pole trip-line prot,

three pole trip-reactor prot)
Reactor RWS AND

prot OR
AND OR Timer
R Trip all
Line OR bus CBs
W
prot
B
OR
R
HOC W
B
AND OR
R
LOC W
S
NEG SEQ
RWS initiate remote
Bus prot
trip as required
(c)
Auto-Transformer (Three pole trip)
Reactor RWS
prot
AND Timer
R OR Trip all bus
Line OR
CBs
prot W
B
R OR AND
LOC W
B
NEG SEQ AND Timer Trip

RWS adjacent
Bus prot
CBs
(d)
Local current check logic utilising remote signalling
Remote receive AND

R
remote sensitive current check
W
Local low set
current check B
(e)
Remote trip receive interlock logics
Timer Timer
Remote trip
receive AND
Under voltage
3f impedance
Zero sequence
Negative sequence
Trip appropriate
CBs
(f)
FIGURE 9.7 Typical examples of back up protection on EHV systems
A typical example of local back up protection is demonstrated in

Fig. 9.8. In this example, the protections at E (P1 or P2), apart from their
prime function of tripping the circuit breaker E for a fault at F, also in
conjunction with the current check relay CC, detect the failure of the circuit
breaker to trip. The timing relay, set to say 200 milliseconds, depending on
the circuit breakers and relays involved, will then initiate the trip of the
circuit breakers A, B, and C if the protection does not reset in that time,
indicating that the circuit breaker has failed to trip.
E G
CC F
B
P1 P2
C
(a) Schematic
X+ Y+ B/U+
P1 P2
+
CC
Trip relays – +
– –
Time
C.B. ‘E’ X Y
– delay
trip coils relay
Y– –
X–
Trip c.bs. A, B, C
(b) D.C. circuit
FIGURE 9.8 Examples of typical back up scheme

Provision of ‘Blind Spot’ Protection utilising Local Back up Protection

The term ‘Blind Spot’ as applied to protection schemes indicates
that if a fault occurs in this area it will not be cleared by the primary
protection unless some special protection is provided.
These ‘Blind Spots’ are created because of the current transformer
arrangements. To reduce costs it is normal practice to have current
transformers located on one side of the circuit breaker only, this
arrangement leads to a blind spot being created.
If one considers the example shown in Fig. 9.9, the current
transformer in this case is located on the line side of the circuit breaker
with overlapping line and busbar protection zones. However, if a fault
were to occur between the circuit breaker and the current transformer, the
fault would be detected by the busbar protection only, the fault would not
be detected by the line protection since it is outside its protection zone. As
a consequence the busbar circuit breakers would be tripped, however the
infeed from the other end of the line would remain unless some other form
of protection was provided. It is normal practice to utilise the circuit breaker
failure protection for this purpose.
Dead Bus
zone
Or
Blind
zone Circuit
breaker
Line
Fault
Y
Line CBF
protection
X
CBF
Y
1. X and Y protection operates and trips CBF isolating bus from fault is outside
line protection zone and line protection will not operate.
2. After a time delay the X and Y CBF protection in conjunction with the bus
initiates a trip to clear the breaker at the remote end of the line.
Current check BU Remote

AND trip
Bus protection TMR
OR CB Trip
Line protection
FIGURE 9.9 ‘Blind spot’ or ‘Dead zone’ protection
In this case, the busbar protection (which would remain in an

operated state until the fault had been cleared) would initiate the circuit
breaker failure protection. The overcurrent check of the circuit breaker
failure protection would also have operated and would remain operated
until the fault had been cleared. The operation of both the busbar protection
and the overcurrent check would cause the timing relay of the circuit
breaker failure scheme to operate, this in turn would initiate the trip of the
circuit breaker at the other end of the line, thus clearing the fault.
Generally the protections for ‘Blind spot’ are not provided with two
independent protections, because the probability of a fault in this area is
considered so low that duplication of protections to cover for this
contingency is, in general, not considered necessary. However, there are
instances where the ‘Blind Spot’ exposure is significant to justify the
provision of two independent protections. Each protection system needs
to be considered on its merits.
Circuit Breaker Failure Scheme Timer Settings
Whether arrangement ‘One’ or ‘Two’ is applied, is dependent on
the preferences and philosophies of the relevant organisation. In both
arrangements, the time delay that is set on the circuit breaker failure scheme
should include allowance for the following times:
• total circuit breaker operating time (from initiation of trip to fault
clearance),
• reset time of the current check relay (the reset time of some relays
could be affected by the performance of the current transformers),
• timing relay error, and
• safety margin.
The timer setting in some instances can be reduced by using a circuit
breaker auxiliary contact to de-energise the timing relay circuit. However,
this type of scheme has a number of disadvantages and, in general, the use
of auxiliary contacts should not be used as they are considered to be
insecure. The disadvantages of using circuit breaker auxiliary contacts are

as follows:
♦ cabling is required to the circuit breaker and this could be a source
of problems
♦ the circuit breaker mechanism could operate (and hence the
auxiliary contact) but the circuit breaker may not have interrupted
the fault current.
♦ erroneous initiation during testing/commissioning of the circuit
breaker.
Circuit Breaker Failure Scheme Current Settings
Whether arrangement ‘One’ or ‘Two’ is applied, the circuit breaker
failure schemes are required to operate for minimum fault conditions. The
setting should be set to less than half the minimum fault level to ensure
the reliability of the scheme. Typically the current sensing portion of
available relays have a continuous rating of twice rated nominal current.
In the case of arrangement ‘Two’ the current setting should also be
greater than or equal to 120% of the maximum load for normal system
conditions. In the event that the two setting constraints of fault current
and load current cannot be satisfied then an alternative design may need
to be considered arrangement ‘One’ may need to be used or alternatively
it may be possible to use high set current checks with other sensitive fault
detectors interlocked by low set overcurrent check relays as indicated
previously.
9.3 EXAMPLE DEMONSTRATING A METHOD TO

DETERMINE CURRENT SETTINGS OF CIRCUIT
BREAKER FAILURE SCHEMES
The system that is detailed in Fig. 9.10, will be used to demonstrate a typical
approach when determining the required settings of a circuit breaker failure
scheme. The logic assumed for the scheme is in accordance with
arrangement ‘Two’ as detailed in earlier section.
The supply system consists of:
• two radially fed 220 kV/66 kV YY∆ transformers without
220 kV bus,
• 220 kV circuit breakers located on the 220 kV incoming lines, and
• 66 kV circuit breakers located on the 66 kV side of the transformers,
66 kV bus tie and 66 kV feeders.
220 kV source
Terminal station
220 kV lines
220 kV Buses
Terminal station
F2
F1
C/C C/C C/C current check
66 kV feeders
FIGURE 9.10 System used to demonstrate the method

of determining circuit breaker failure settings
The failure of the 66 kV transformer circuit breaker is to be considered

for this example.
Assume that the most onerous contingency to be considered for the
failure of this circuit breaker is a fault occurring either on the 220 kV side
of the transformer or on the tertiary of one of the 220 kV/66 kV
transformers.
Assume that the circuit breaker failure protection scheme proposed
at this stage is as follows:
• high set current check with two phase elements and a neutral
element,
• proposed settings of 2100 A for the phase elements and 525 A for
the neutral elements.
The exercise will be to verify that the proposed arrangement is
satisfactory.
The other conditions that have been supplied are as follows:
• the maximum load expected to be taken by the transformers is
1200 A.
• the available phase settings for the relay are 1000 A, 1600 A, 2100 A
and 3000 A.
• the available neutral settings for the relay are 250 A, 400 A, 525 A
and 750 A.
• Equivalent fault currents seen by the back up protection (with
safety margins) for faults on the 220 kV bus or on the tertiary of
the transformer are as follows:
Fault Location Fault Level Safety Margin
220 kV Bus 3Φ = 3648A (R,Y, B) 3648/2100 = 1.74
Φ-Φ = 0A (R ) –
3158A (Y) 1.5
3158A (B) 1.5
Φ-g = 3657A (R) 1.74
595A (Y) –
595A (B) –
2467A (N) 4.7
Transformer 3Φ = 3884A (R,Y, B) 1.85
Tertiary Φ-Φ = 1942A (R ) –
3884A (Y) 1.85
1942A (B) –
Φ-g = Not Applicable
From the data provided above, the following conclusions can be

reached:
• the setting of 2100A for the transformer 66 kV circuit breaker
current check is too high to cater for this example of back fed
transformer faults. (A safety margin of greater than 1.8 must be
achieved. This condition is not met for the 220 kV faults).
• to cater for the phase-to-phase faults on the transformer tertiary,
a current check with three individual phase elements is required.
• based on the maximum transformer load of 1200 A, the current
check phase element can be set on 1600 A.
By selecting a current check with three individual phase elements
with settings of 1600 A, the required safety margins and the required
performance would be achieved for the operating conditions considered.
This page
intentionally left
blank
PART B
POWER SYTSEM
COMMUNICATIONS
This page
intentionally left
blank
CHAPTER
Introduction
1.1 INTERNATIONAL ELECTROTECHNICAL COMMITTEE

(IEC) TECHNICAL COMMITTEE (TC) 57
The International Electrotechnical Committee (IEC) Technical Committee
(TC) 57 was established in 1964 because of an urgent need to produce
international standards in the field of communications between the
equipment and systems for the electric power process, including telecontrol,
teleprotection and all other telecommunications to control the electric
power system. IEC did not only consider equipment aspects, but more
and more system parameters. This scope was modified to prepare standards
for power system control equipment and control systems, including
supervisory control and data acquisition (SCADA), energy management
system (EMS), distribution management system (DMS), distribution
automation (DA), teleprotection and associated communications.
The technical experts of twenty-two (22) participating countries have
recognized that the increasing competition among electric utilities is due
to the deregulation of energy markets. The integration of equipment and
systems for controlling the electric power process into integrated system
solutions is needed to support the utilities ‘core processes’. Equipment
and systems have to be interoperable, and interfaces, protocols and data
models must be compatible to reach this goal.
1.2 ELECTRIC POWER RESEARCH INSTITUTE (EPRI)

The Electric Power Research Institute (EPRI) was in existence since the
1970s to develop technologies for the benefit of electric utilities. It manages
research and development projects with funds supplied by those utilities

as group and other sources. Since the 1980s EPRI has recognized the
potential benefits of a unified scheme of data communications for all
operating purposes across the entire utility enterprise. They focussed on
the ease of combining a broad range of devices and systems; and the
resultant sharing of management and control information among all
departments of the utility organization. EPRI commissioned the North
American Utility Communication Architecture (UCA) project, which
identified the requirements, the overall structure and the specific
communications technologies and layers to implement the scheme. The
UCA initiative works under similar charter and recommends for
implementation of interfaces, protocols and data models. It is expected
that upon completion, the IEC TC 57 will adopt these recommendations
and make them a subset of the IEC 61850 standard.
The key to standardization is interoperability between vendors and
systems. Of particular interest to all are on-going discussions of functional
interoperability, hardware and software interfaces, protocols, data models
and interchangeability.
1.3 INNOVATIVE INTEGRATION DEVELOPMENTS (IEDs)

Innovative integration developments (IEDs) within multifunction
microprocessor based relays and other electronic devices have created new
ways of collecting and reacting to data and using these data to create useful
information. Power providers are facing demands to increase productivity
and make electric power safer, more reliable and more economical. This
can be done when electric manufactures provide innovative, simple to use,
robust technologies to protect, automate, control, monitor and analyse
power systems. An essential element of this strategy is the development
of important communications technologies and protocols. When integrated
together, relays and IEDs become a powerful, economical and streamlined
Instrumental and Control (I&C) systems, capable of supporting all aspects
of electric power protection, automation, control, monitoring and analysis.
Communications processors, RTUs and PLCs are used as integration and
automation controllers all over the world. The number of Utility
Communication Architecture (UCA) based IEDs for protection and control
available on the market is continuously growing and they are starting to
appear in installations around the world.
There are some significant differences between UCA based IEDs and
conventional microprocessor based protective relays. This requires good
understanding of the fundamentals of communications based substation
protection devices and at the same time the availability of proper
INTRODUCTION 245
configuration tools that will make it easy for the user to adopt the relay for
its application in the substation. This is especially important for peer-to-
peer communications based protection and control functions.
The material is organized in seven chapters. After Introduction of
the International Electrotechnical Committee and its objectives, the
initiatives taken by the Electric Power Research Institute are identified
followed by development of Intelligent Electronic Devices and the
development of Multifunction Microprocessor based relays. A separate
chapter is dedicated to communication principles where various
terminologies and architectures are identified. Numerous terminologies
are described so that the student or power network operator can have an
Engineering book that is easily understandable and which can be referred
to easily. Various architectures are considered that may very well be
appropriate for different scenarios. A separate chapter is included on
Protocols. In this chapter issues such as how communications processors
communicate to each other within the entire power network are identified
in detail. The application part of this book covers design and
implementation of Universal Middleware to support real-time
communication services over substation communication networks.
Intended to link various power networks via LAN & WAN
communication and information embedded power networks, this book is
authored from both university and power network operator’s perspective.
This book is aimed at the audience of application, design and Research &
Development Engineers in Power System Communications as well as
university graduate and continuous education students.
CHAPTER
Communication Principle
2.1 TERMINOLOGIES
2.1.1 System Automation and System Integration
System automation is the control of power system apparatus operations to
take the place of the human functions of observation, decision and action.
Substation automation refers to using Intelligent Electronic Device (IED)
data within the substation and control commands from remote users to
control the power system devices within the substation.
System integration is the act of communicating data to and from or
between IEDs in the Instrumentation and Control (I & C) system and remote
users. Substation integration refers to combining data from the IEDs that
are local to a substation so there is a single point of contact in the substation
for all the I & C data. This single point of contact then mediates remote
and local substation control. Since true substation automation relies on
substation integration, the terms are often used interchangeably. There is
often a need for multiple single points of contact to serve multiple user
connections or provide redundancy. The single point of contact is an I&C
device acting as a client/server, programmable logic platform, gateway,
router, dial-out device, communication switch, time synchronization
broadcaster, or a combination of these.
2.1.2 Substation Controller
Products from many industries are used to perform substation automation.
RTUs, port switches, meters, bay modules, and protocol gateways from
the Supervisory Control and Data Acquisition (SCADA) industry; PLCs
from the process control industry, relays and communication processors
from the protection industry, and PCs from the office environment.
COMMUNICATION PRINCIPLE 247
Substation controller and bay controller are the terms commonly

used to refer devices that perform data acquisition and control of IEDs
and contain local Input/Output (I/O). The communications processor is
the only substation controller that can perform all the substation automation
tasks. The communications processor is also the only device that is designed
to meet the harsh environmental conditions as the relays themselves.
SCADA and process control industry products are not designed to meet
these environmental standards.
2.1.3 Client/Server for Dynamic Data
The communications industry uses the term client/server for a device that
acts as a master or client, retrieving data from some devices and then acts
as a slave, or server, sending this data to other devices. The client/server
for dynamic data collects and forwards data frequently based on master
poll rate or by exception. These data include protection data, metering
data, automation data, control data and supervisory data.
2.1.4 Client/Server for Archived Data

A substation archive client/server collects and archives historical data from
several devices. These data include system profiles, event reports,
Sequential Events Recorder (SER) reports, power quality reports and
protection quality reports; they provide a clear picture of system
performance. The user retrieves data when it is convenient to do so.
2.1.5 Data Concentrator
A data concentrator creates a substation database by collecting and
concentrating dynamic data from several devices. In this fashion, essential
subsets of data from many IEDs are forwarded to a master through one
data transfer. The data concentrator database passes data from one IED to
another when they are not connected peer-to-peer.
2.1.6 Message Broker
A message broker collects and stores entire messages from several sources.
Rather than extract and concentrate only a subset of the data, the message
broker collects the entire message including header, content and error check
terminator. The message broker then acts as an agent for the message by
negotiating where and when to send the message. In this fashion, an entire
message can be exchanged between two devices that cannot be directly or
transparently connected.
2.1.7 Programmable Logic Platform
A programmable logic platform executes custom automation logic
equations. RTUs and PLCs have little or no default automation capability.
Therefore, RTUs and PLCs support flexible programmability so the end

user can create necessary automation from the ground up.
2.1.8 Protocol Gateway

A gateway converts conversations from one protocol or communication
language to another. Often RTUs or PLCs are used for the sole purpose of
acting as a gateway between substation data and a legacy SCADA or energy
management system (EMS) protocol.
2.1.9 Hubs
A hub is a relative simple multi-port device that rebroadcasts all data that
it receives on each port to all remaining ports. It operates at the physical
layer of the Open System Interconnection (OSI) network model (Fig. 2.1),
so it does not use any of the data to remaining routing actions. The
International Standards Organisation (ISO) created the OSI [1-3].
Application
Presentation
Session
Transport
Network
Data link
Physical
FIGURE 2.1 ISO OSI model
Layer Levels
• Application—Provides a set of interfaces for applications to use
to gain access to networked services.
• Presentation—Converts application data into a generic format
for network transmission and vice versa.
• Session—Enables two parties to hold ongoing communications,
called sessions, across a network.
• Transport—Manages the transmission of data across a network.
• Network—Handles addressing message for delivery, as well as
translates logical network addresses and names into their physical
counterparts.
• Data Link—Handles special data frames between the network
layer and the physical layer.
• Physical—Converts bits into signals for outgoing messages and

converts signals into bits for incoming messages.
2.1.10 Switches
A switch is an intelligent multiplexing device that monitors the data link
layer of the OSI network model (Fig. 2.1). If a data packet is incomplete or
indecipherable, the switch ignores it and does not broadcast it. If a data
packet is intact, the switch rebroadcasts it to another port, based on the
addressing data included in the packet and the addresses associated with
each port of the switch. New switches can operate on the Network (Layer 3)
or Transport (Layer 4) packet information.
2.1.11 Router
A router is an intelligent device used to connect two networks together. It
can be a complex model, with many features. It operates at the Network
layer of the OSI network model (Fig. 2.1). In another term, as used in the
communications industry it refers to a device that routes data in transit
between source and destination. The router intelligently transmits messages
received on one communications port to another communications port.
The destination port for the message is dynamically determined via the
content of the message. This is used to efficiently route SER and other
messages through multiple substation controllers without affecting
substation automation.
2.1.12 Servers
A server collects data from all of the local devices and creates a substation
database. Often a local human machine interface graphics package uses
data from this database. Servers function at the Application layer of the
OSI model (Fig. 2.1).
2.1.13 Dial-out
A dial-out device initiates conversations or triggers paging from the
substation to a remote user. Use for dial-out include ensuring connection
security, eliminating the need for a dedicated communications connection,
and performing unsolicited indication of a disturbance with fault location.
2.1.14 Communications Switch
A communications switch is the single point of contact for remote users to
dial in and make a direct connection to all substations IEDs individually. A
single communications connection form inside or outside the substation is
switched between several IEDs. The user initiates a dynamic conversation
with specific IED and the port switch merely ‘passes through’ the
conversation.
2.1.15 Time Synchronization Broadcast

In order to synchronize IED clocks, a device in the substation needs to
generate, or acquire from an external source, a time value and then
broadcast it to the IEDs. Millisecond accuracy and repetitive
synchronization are necessary.
2.1.16 Local I/O
Substation controllers often make use of the I/O within connected IEDs.
They also support local I/O terminated at the substation controller for
automation and alarm functions.
2.1.17 Eavesdrop Communications
Eavesdrop communications refers to monitoring a conversation between
two devices in the I & C system, for capturing and storing the transferred
data. This is useful for extracting the information without influencing the
flow of data between devices that may not be available.
2.1.18 Emulate Protocol Messaging with Settings
When choosing the best new and in-service devices a successful I & C
system, it is often necessary to select from multiple vendors and also
multiple vintages or generations of products. Many of these devices employ
proprietary communications and interfaces. Most substation controllers
must have embedded software written specifically to communicate via
proprietary interfaces.
2.1.19 Autoconfiguration
Some substation controllers simplify implementation through
autoconfiguration. This process automatically determines the proper baud
rate to communicate with the connected IED as well as start-up parameters,
device type and capabilities.
2.1.20 Device-Level SER
A device-level SER application creates and stores event data with a time
stamp. Predefined input contacts and logic elements are monitored as the
source of event records. The SER associates a time of occurrence with each
event and stores these data in a buffer. It forwards these data in the order
of event occurrence in an unsolicited fashion and/or in response to a
request.
2.1.21 Station-Level SER
A station-level SER application creates and stores event data with a time
stamp for local inputs. It also collects and stores SER messages from other
IEDs in the station in an unsolicited fashion and/or in response to a request.
The local SER messages and the SER messages from the other IEDs are
stored in a buffer and organised in order of occurrence. The station SER

are also forwarded in order of event occurrence in an unsolicited fashion
and/or in response to a request.
2.1.22 Tier-to-Tier/Peer-to-Peer
Some substation controllers accommodate substations of varying size as
well as redundant designs by supporting peer-to-peer and tier-to-tier
functionality. Peer-to-peer refers to the direct transfer of data between
devices functioning in a similar capacity. Tier-to-tier refers to devices that
can transfer data while connected in such a way that one is the client and
the other the server.
2.1.23 Substation Hardened
The inverse of the failure rate of a device, or mean time between failure
(MTBF) compares reliability of devices. Most devices that were traditionally
used for automation in the past, such as RTUs and PLCs were designed to
be operated in controlled environments like control rooms and generation
facilities. The average MTBF or RTUs and PLCs is 11 and 17 years
respectively. This low MTBF reflects a design philosophy based on frequent
replacement and maintenance. Protection devices are designed to be more
reliable, fail less frequently, be in service longer, and cost less to maintain
than PLCs and RTUs. The device MTBF can be used to predict how available
an automation system will be and how frequently the maintenance staff
will be replacing failed devices [1-2].
2.2 ARCHITECTURES
2.2.1 Types of Connections
Direct connect and multidrop are two types of data link connections
available to create networks. In a direct connection, there are only two
devices connected via network media, which can be metallic, wireless or
fiber. Each interface consists of a separate transmit and receive connection
at each device. Since there are only two devices, each of them can constantly
control the connection on which they are transmitting and both can know
implicitly to which other device they are connected. Several individual
direct connections to many IEDs would allow each of them to communicate
simultaneously. Many direct connections originating from one device is
called a star network topology (Fig. 2.2).
Network
controller
Relays/IEDs
FIGURE 2.2 Star topology
Any protocol, including those designed for multidrop applications,

can be used for direct connections in a star topology. Star network designs
support a wide range of IED capabilities. Simple, slow communicating
devices can coexist with more complex fast communicating relays. Devices
from different manufacturers with different protocols can coexist in the
same star network because each has a dedicated direct connection.
Open architecture is a term that refers to networks that are
interoperable between hardware and software interfaces and therefore
among vendors. The star topology is the only design that is truly open
architecture and will accommodate multiple protocols, baud rates and
network interfaces.
The most common communication architecture used today is the
multidrop network. In a multidrop network topology, several devices can
be physically connected in a bus or ring network and control of the transmit
and receive connection must be negotiated. Figs. 2.3 and 2.4 illustrate relays
connected in a bus and ring topology respectively. A multidrop connection
requires only one device to communicate at a time. There are often
additional components for terminations and network drop connections,
which are vertically down to the individual relays or IEDs. Because all
IED/relay share the cable, communications are usually controlled by the
network master or a token passing scheme in which IEDs/relays have
permission to communicate when they receive the virtual token and then
pass the token on when they are done. Peer-to-peer messaging may or
may not be available. Sequential polling of each IED/relay usually performs
data retrieval by the master [3-4].
Network
controller
Relays/IEDs
FIGURE 2.3 Bus topology

Network
controller
Relays
FIGURE 2.4 Ring topology
Software and hardware are used to determine which device has

permission to transmit so that data does not collide on the conductor. Since
several devices are connected, addressing is necessary within the protocol
to identify the source and destination of the data being communicated.
This addressing adds overhead in the form of processing time and amount
of information that needs to be transmitted thus reducing the data transfer
rate. Devices compensate this by increasing the speed at which they
communicate and increasing the amount of communications processing
that they perform.
Troubleshooting communication problems on a multidrop network
is difficult. Messages from many sources must be captured and deciphered.
Direct connections are quickly and easily verified using simple LED
indication.
Relays have varying memory and computational capacities and,
therefore varying protocol support capabilities. Interactions on a multidrop
network must be done at the lowest common denominator and all devices
must support the same baud rate and physical connection.
It is important to note that if the mediation of control of data
transmission should fail, none of the multidropped devices can
communicate. This can be caused by relay communications hardware
failing to release control, relay communications software failing to process
mediation schemes correctly, or corruption of the network [2-4].
2.2.2 Long-term Trend of Networks
The future trend is away from multidrop network and towards star
network, e.g., Ethernet. Originally this was conceived as a multidrop
network using expensive coaxial cable, however widespread use has shown
that a star network is far superior. Today all Ethernet networks are built
using hubs. A hub acts as a very short bus, which allows one to wire from
the hub using inexpensive cable in a star configuration.
With network traffic and use continuing to rise, smarter devices like
switches are replacing hubs. A switch can store and forward information
making the logical network a star also. Multiple nodes can transmit or
receive messages from the switch at the same time. Ethernet has now
completed the transformation both electrically and logically in a star
network architecture.
2.2.3 Peer-to-Peer Network Architecture

Protection data, for the purpose of security, reliability and speed are highest
priority and should be transferred via a single purpose conversation on a
channel dedicated to this purpose along (Fig. 2.5). These functions perform
optimally if this data can be transferred every device processing interval,
between 1-12 ms.
Relays/IEDs
FIGURE 2.5 Peer-to-Peer direct connection
2.2.4 Typical Integrated Digital Protection and Control System

A simplified diagram of a typical Integrated Substation Protection and
Control System is shown on Fig. 2.6.
It consists of a series of devices interconnected through an Ethernet
network. Such a system has a hierarchical structure with distributed
intelligence and different level of complexity. The hierarchy however is
only functional, as well at the same time it can be flat from the
communications point of view, i.e., all IEDs are connected to the same
Ethernet network. For large substations with several voltage levels and
multiple buses the number of hubs will increase and depending on the
requirements for protection performance the hubs can be replaced by
switches in order to limit the traffic on the different segments of the
substation network.
The primary functions of the IEDs is to protect different substation
and power system elements viz. transformers, buses, capacitor banks,
motors, lines etc. IEDs perform this basic function only under fault
conditions, which is a event with very low probability. However, they need
to have sufficient processing power and intelligence. Hence at lower
hierarchical level it allows their use for data acquisition, control, monitoring
and fault recording system.
At the next level the Bay Controller IEDs provide additional digital
and analogue interface with the substation environment and at the same
time provide protection and control functions.
At the top level the Substation Controller IED or the substation

computer it provides integrated protection and control. It provides
substation protection and control functions based on the exchanged high-
speed peer-to-peer communications messages and over the substation
LAN. It also provides the Human Machine Interface (HMI) functionality
with the different IEDs in the substation. It supports alarm and event
reporting, data archiving, analysis, monitoring etc., functions.
In Fig. 2.6 the architecture uses the Ethernet network with the
required Ethernet hub and a router connecting it to the utility Wide Area
Network (WAN) [2-4].
SCADA Engineering
master station
Ethernet RS232
HUB
Route WAN Terminal
server
Terminal
server
Substation
UCA HMI
gateway
RS485 Ethernet
Ethernet
RS232
Legacy IED IED Laptop Data
Front
IED computer concentrator
RS485 RS485
RS485 Back 2 Back 2
FIGURE 2.6 Communication processor centric hybrid network
2.2.5 Hybrid Network Architecture

Using the communication processor as the substation controller, a hybrid
system, as shown in Fig. 2.7, can be created to perform control, monitoring,
automation, protection, analysis, tests, maintenance and operation of the
power system.
Host and/or network

connections
Noninstrusive
eavesdrop
data link
Communication processor
Expert
Metering nodes
master
Met
Multidrop subnetwork
Subsystem Peer-to-Peer
connections
FIGURE 2.7 Communication processor centric hybrid network
2.2.6 LAN Design Consideration

A station LAN with all IEDs on one segment and a multiple segment process
LAN design is shown in Fig. 2.8. A merged station and process LAN is
shown in Fig. 2.9.
The example LAN designs in Figs. 2.8 and 2.9 are two of many
different ways to configure the network. Other design methods improve
reliability, speed and maintainability. Optimising reliability and speed
create conflicting substation LAN designs. Speed is important for
sophisticated distributed protection, synchro-check and time
synchronisation of IED clocks. Peer-to-peer speed is fastest when all IEDs
are connected on a single LAN segment but communication functions are
more reliable when systems are redundant and without a single point of
failure. The peer-to-peer communications are based on what is defined as
a GOOSE i.e., Generic Object Oriented Substation Event and it is based on
a multicast asynchronous reporting of an IEDs digital outputs status to
other peer devices enrolled to receive the configuration of the substation
integration process (Fig. 2.10). It is important to note that if the mediation
of data transmission control fails, none of the devices on a LAN segment
could communicate. This can be caused by the IED communications
interface failing in such a way as to corrupt the network.
Distributed
HMI
controller
Ethernet
switch
Primary Back up Primary Back up

protection protection protection protection
Feeder 1 Feeder 1 Feeder 2 Feeder 2
Process Ethernet Ethernet

LAN switch switch
Disconnect Circuit Merging Disconnect Circuit Merging

switch breaker unit switch breaker unit
Feeder 1 Feeder 1 Feeder 1 Feeder 2 Feeder 2 Feeder 2
FIGURE 2.8 Station LAN and multiple segment process LAN design
Station
LAN Distributed
HMI
controller
Ethernet
switch

Process Ethernet Ethernet

LAN switch switch

FIGURE 2.9 Merged station and process LAN design

Ethernet
IED
HUB
Sending Receiving
IED IED
GOOSE
Receiving
IED
IED
FIGURE 2.10 Sending and receiving IED on an ethernet LAN
The designer has to ultimately balance the needs to create isolated

LAN segments for security, redundant systems for reliability, and
monolithic and single segment LANs for high speed. The value of each
need will be compared against the cost in dollars and additional processor
burden within devices.
2.2.7 External Substation Connections
The IT products in the substation facilitate easy connection to other
corporate systems through WAN or Internet connections. These connections
possibilities highlight the importance of securing connections into the
substation LAN. Fig. 2.11 shows a previous substation network with the
addition of external connections [3-4].
Distributed
HMI controller To Internet
To WAN
Ethernet
switch
Router

Ethernet Ethernet
switch switch

FIGURE 2.11 External connections to substation communication network

A three-pronged third-party interface development approach can

be:
• connect relays as an RTU, PLC or HMI slave, by sharing interleave
protocol specifications with SCADA vendors and system
integrators, and through the communications processor.
• connect directly to substation networks as a Modbus slave in the
Communications Processors or as a Modbus Plus® slave in the
Communications Processors.
• for the long-term, use a local high-speed network in the substation
and this will be the method of choice for the third-party
interconnection such as an application protocol running on
ethernet [1-4].
CHAPTER
Protocols
A modern power system is one of the largest complexes constructed and

operated both in terms of geographical distances as well as generated and
transmitted power. Such a system needs precise and high quality control
with protection functions as primary due to the top priority safety reasons.
Traditionally protective relays have been electro mechanical devices whose
purpose was only to protect electrical power systems against system
failures. Application of microprocessors to power system relaying has
increased the functionality of protective relays and brought new concepts,
which considers control, protection and monitoring functions integrated
together. In the past decade, new communications schemes have been
designed and retrofitted into the substations by the utilities to integrate
data from relays and Intelligent Electronic Devices (IEDs) and capitalize
on the protection, control, metering, fault recording, communication
functions available in digital devices. This chapter describes substation
communications and the ongoing communication standardization efforts
discussing the IEC 61850 and the Utility Communications Architecture
(UCA) standards.
3.1 INTRODUCTION TO POWER SYSTEM

COMMUNICATION
Many of today’s electric utility substations include digital relays and other
Intelligent Electronic Devices (IEDs) that record and store a variety of data
about their control interface, internal operation and performance, and about
the power system they monitor, control and protect. Nowadays, digital
relays are widely replacing the aging electromechanical and solid-state
PROTOCOLS 261
electronic component—type relays and relay systems. Fig. (3.1) shows a

digital relay with its target interfaces. Digital relay’s popularity comes from
their price, reliability, functionality and flexibility. However, the most
important feature that separates the digital relay from previous devices is
its capability of collecting and reacting to data and then using this data to
create information.
Such information includes: [5, 6]
• Fault location and fault type
• Prefault, fault and post-fault currents and voltages
• Relay internal element status
• Relay control input and output status
• Instantaneous and demand metering
• Breaker operation data
• Relay operation data
• Diagnostic and historical data.
CT
VT
Digital rela
y
FIGURE 3.1 Digital relay with target interface
Instrumentation and Control devices, which are built using

microprocessors, are commonly referred to as Intelligent Electronic Devices
(IEDs). Microprocessors are single chip computers that can process data,
accept commands, and communicate information like a computer. IEDs
can also run automatic processes, and communications are handled through
a serial port like the communications ports on a computer. Some
examples [8] of IEDs used in a power system are:
• Instrument transformers
• Transducers
• Remote terminal unit (RTU)
• Communications port switch
• Meter
• Digital fault recorder

• Protocol gateway.
3.1.1 Power System Communications Initiatives

Communication systems have been used for decades to enhance the
performance of power systems. Without the use of a proper communication
channel, power system protection suffers from a major disadvantage of
not being able to accurately discriminate faults [27]. When voltages and
currents are analysed only from one terminal, it cannot be concluded
whether a fault near the far end terminal is internal or external to the
protected line segment. This requires delayed tripping for such faults, which
can endanger system stability or increase damage. At the far end terminal,
the decision whether the fault is internal or external is obvious, not from a
distance measurement but from knowledge of the direction of the fault.
This information can be transmitted to the other terminal with the use of a
proper communication channel enabling it to decide whether to trip or not
to trip [9, 10, 11, 36].
Power providers are focussed on increasing productivity and making
electric power safer, more reliable and more economical by providing
innovative, simple to use, robust technologies for power system protection,
automation, control and monitoring. Development of appropriate
communications technologies and protocols is at the heart of this strategy.
When relays and IEDs are integrated together, they form a powerful,
economical Instrumentation and Control (I&C) system to support all aspects
of electric power protection, automation, control, monitoring and
analysis [8].
Figure 3.2 shows how IEDs and relays can be interconnected together
to form protection schemes for power systems. Such a system also supports
the substation in terms of the monitoring, analysis and automation aspects.
The relaying and measurement tasks have been well understood
and standardized. On the other hand, the technical methods and operating
impact of data communications continue to evolve dramatically. There is
a wide variety of incompatible communications approaches and systems
in the marketplace. Competing manufacturers have been following unique
approaches when designing the communications interface circuits. Other
IED makers designed networks, which make it possible to connect a number
of devices in one substation to a single local host that could dynamically
address requests for data to any unit. However, the user could not directly
interconnect competing products since the protocol remained unique for
each system. Whilst the use of products from competing vendors offers
PROTOCOLS 263
SCADA Engineering
master station
RS232
Ethernet Terminal
Router WAN
HUB server
Terminal
server
UCA Substation
gateway HMI
RS485 Ethernet Ethernet

Legacy IED RS232 Laptop Data
IED IED
Front computer concentrator
RS485 RS485
RS485 Back 2 Back 2
FIGURE 3.2 Typical integrated substation protection

and control system [28]
users a variety of protection and monitoring capabilities [12, 13], the

variations in the communications system, and the need for a different
system for each vendor, has often frustrated users.
Nowadays, the desire and the need to merge the communications
capabilities of all the relays and IEDs in a substation into a standard that is
capable of providing data gathering and setting capability as well as remote
control, is clearly recognized. Furthermore, multiple IEDs can share data
or control commands at high speed to perform new distributed protection
and control functions [14].
In addition to standardization efforts within the substations, the
development of several powerline communication protocols such as CEBus
has also renewed the interest in powerline communications, which has
existed for use in the home as a networking medium for many years now
[37, 38].
The emergence of new devices with the recent technological
developments have enabled utilities to utilize the powerline to send and
receive control signals with some degree of reliability offering broadband
services. Powerline Communication (PLC) is the transmission of data along
the utility powerline network, which eliminates the need to rewire houses
and buildings with separate communication links. Although other options
such as satellite communications exist, the high cost and the unavailability
of these technologies in rural and suburban areas have also increased the
importance of powerline communications. The main advantage of
powerline communications is the fact that the physical network is already
installed over a wide area [38].
As mentioned earlier, some forms of PLC have been in use for quite
long time now. However, a new perspective has been formed in the recent
years with the proliferation of companies developing low-cost chipsets,
which enable new high-speed short-distance applications. In addition,
companies have high hopes for developing ways of communicating
Broadband Internet connections over long distances using the utility
powerline network. There are many existing and evolving high-speed PLC
technologies with further research work proceeding aggressively.
However, the fact that powerlines are inherently limited to transmission
at the highest 400 Hz brings some concerns [37]. In addition, the high
electronically contaminated structure of the powerline and the high signal
attenuation at the frequencies of interest makes it challenging to achieve
dependable results. Hence, there is still a long way to go!
3.2 PROTOCOLS IN GENERAL

A protocol is basically a set of rules that must be obeyed for orderly
communication between two or more communicating parties.
Communication between data processing systems from different
manufacturers has often been particularly difficult due to the fact that there
has been separate development of data processing and data
communications techniques, often resulting in complex and expensive
interfaces. With the International Standards Organization (ISO) model,
which is commonly known as the Open Systems Interconnection (OSI),
the communications process has been divided into seven basic layers as
shown in Fig. 3.3. These layers define how data flows from one end of a
communications network to another end and vice versa. Two devices can
only communicate if each layer in the model at the sending device matches
with each layer in the model at the receiving device [15, 16, 17].
Application layer Application layer

Information
Presentation layer processing Presentation layer
Session layer functions Session layer
Transport layer Transport layer
Network layer Communi- Network layer
cation
Data Link layer functions Data Link layer
Physical layer Physical layer
THE PHYSICAL MEDIA OF THE OSI
FIGURE 3.3 The OSI reference model

PROTOCOLS 265
The user can quite often make choices in any given layer. The
ensemble of choices made to implement a protocol is termed a profile. The
rules designed by a protocol profile are designed to organize operating
issues in the following areas:
• Framing
• Error control
• Sequence control
• Transparency
• Line control
• Time-out control
• Start-up control.
There are literally thousands of combinations of protocol agreements
that can be created with the large domain of existing pieces. The main
protocols that have found widespread use in the substation environments
are [18, 10]:
• MODBUS: A popular master-slave protocol with industrial users,
which has also become popular in substations. It issues simple
READ/WRITE commands to addresses inside an IED.
• Distributed Network Protocol (DNP): An increasingly popular
master-slave protocol mainly in North America. DNP can run over
multiple media, such as RS-232 and RS-485 and can issue multiple
types of READ/WRITE messages to an IED.
• IEC-870-5-101: It is considered as the European partner to DNP.
It differentiates itself from DNP with its slightly different
messaging structure and the ability to access object information
from the IED.
• UCA: It is the Utility Communications Architecture designed to
satisfy every possible requirement in substation equipment.
3.3 EXPAND ON DNP

DNP (Disturbed Network Protocol)-1 and DNP-2 which can be used to
construct parallel structures for simulating large Scale Neural (NN)
models [39].
3.3.1 Introduction to DNP
The DNP-1 was implemented in 1991, and the DNP-2 was implemented
by the VLSI chip of single-chip multiprocessor having four DNPs. The
DNP-2 has the performance 50 MCPS (Million Connections Per Second) at
50 MHz and the microprocessor (E-MIND/2) having 2D mesh structure of
1024 DNP-2’s accomplishes maximum 40 GCPS (Giga Connections Per
Second) [39].
3.3.2 The Structure of DNP

A processing element, the DNP-l is a 8-bit microprocessor devised with
instructions and data paths for efficient neural computation: a weighted
sum, a non-linear function etc. The DNP-1 consists of three block units: an
arithmetic unit, a control unit and a communication unit [39].
The DNP-2 is a 16 bit microprocessor and its architecture is similar
to DNP-1 exclusive of parallel arithmetic constructs and two pipeline
functions. The DNP-2 also consists of three units: An arithmetic unit, a
control unit and a communication unit [39].
3.3.3 Size/Speed Comparison of DNP
It is necessary to compare performance of the DNP-1 with the DNP-2. Table
3.1 clearly indicates how one can obtain the results that the function (an
arithmetic pipeline of an arithmetic unit and an instruction pipeline of a
control unit) and the parallel constructs of DNP-2 are cost efficient for
neural network emulation [39].
Table 3.1: Cost/Performance comparisons using AT of two processors
(AU: Arithmetic Unit; CU: Control Unit)
Data DNP-l DNP-2

Model
AU CU Processor AU CU Processor
(l6 bits) (l6 bits) (l6 bits)
Number 491(950) 456(456) 94(1306) 6646 1682 8328

of gate
Number 9(17) 4(4) 13(21) 1 1 1
of SC
AT 4419 1824 12311 6646 1682 8329
(16150) (1824) (27426)
3.3.4 Implementation and Performance

We make the DNP-2 as a processing element of large scale parallel
neurocomputer, and fabricate four DNP-2’s into one chip to increase the
density of system integration. The DNP-2 is implemented in 0.8 µm CMOS
process, with a target of 50 MHz clock rate. This chip features 299 pins
and a power dissipation of 2 watts at 50 MHz. The silicon size measures
11.5 × 11.5 mm2 and integrates 60,000 gates excluding on chip memory of
DNP-2. The DNP has a peak performance of around 50 MCPS [MCPS].
PROTOCOLS 267
3.4 STANDARDISATION DEVELOPMENTS

Study committee B5 (previously 34) of CIGRE comprises of advisory
groups, working groups and task forces covering Power System Protection
and Automation in terms of the principles, design, applications,
coordination, performance and asset management of:
1. system protection,
2. substation control and automation,
3. remote control systems and equipment, and
4. metering systems and equipment.
The working group 07 of the study commitee B5 is the one mainly
concerned with substation control and automation reporting on aspects
such as:
• substation control and automation,
• possible architecture of automated systems, and
• state of art in communication standards and applications.
The WG 07 reports that the power system industry is in a fast
competition to have an optimal management of the power system network
in all system levels. With the privatisation of the power system industry, a
new electricity market has been formed differing in all aspects from the
traditional old market. Therefore automating the existing substations is
very important for the utilities, who want to meet the existing challenges
of the future market and reliability of the existing equipment [19, 20].
Nowadays, most of the substation automation systems have similar
architectures usually including a central computer connected to
decentralized computers and protection relays, and also synchronization
and communication components. A few major alternatives to the
architecture that is usually a central computer and the database is
distributed. Protection and control can be combined in a single IED or
protection in a physical device separate from control.
The WG 07 reports that recent developments in technology have
brought new concepts requiring IEDs to function on a much broader level,
such as measurement devices or control devices. It also requires IEDs to
communicate with each other using a common software environment
involving standardized protocols and standardized object models for each
IED. The UCA 2.0 was the first attempt to meet such requirements. The
WG 07 report states that the IEC 61850 project has evolved further
incorporating the UCA 2.0 works and extending it towards the process
level[21, 22]. The harmonisation of data models between UCA 2.0 and the
IEC 61850 will eventually be an important step towards a worldwide-
accepted standard [19, 20]. IEC 61850 is a super subset of UCA 2.0 while
offering some additional features. IEC 61850 was published as standard
around the year 2003[23]. The following sections will detail the work done
by the IEC and EPRI groups towards standardization.
3.4.1 The UCA Substation Communication Project

The introduction of higher-level protocols in IEDs has only enabled
communications between like devices or in other words communications
between devices from the same manufacturer. In order to communicate a
variety of devices from different vendors, which enables the utilities with
a variety of protection, monitoring and automation capabilities, there is a
need to use protocol converters or gateways. Furthermore, the IED
protocols are also limited in capability including speed, functionality and
services making engineering harder, and increasing operation and
maintenance costs. Worldwide, electric utility deregulation is expanding
and creating demands to integrate, consolidate and disseminate real-time
information quickly and accurately within and with substations [25].
The Electric Power Research Institute (EPRI) has existed since 1970s
to develop technologies for the benefit of electric utilities. It manages
research and development projects with funds supplied by those utilities
as a group and other sources. The Utility Communications Architecture
(UCA) was commissioned by the EPRI, which identified the requirements,
the overall structure, and the specific communications technologies and
layers to implement the standardization scheme. UCA aims to dramatically
improve device data integration into the information and automation
technology in order to reduce the engineering, monitoring, operation and
maintenance costs increasing the agility of the whole life cycle of a
substation [14, 26].
Many relay and IED manufacturing companies and progressive
utilities showed their interest in UCA work and joined in the effort to define
and demonstrate a communications network stack. The approach adopted
defines the technical requirements for a system to control and monitor
substations of any size. With the use of the substation Local Area Network
(LAN), the mass of the dedicated wiring among the IEDs and power
apparatus is ultimately reduced or replaced.
With continued EPRI support, a long list of relay, meter, and IED
vendors have built UCA-compliant versions of products. The elaborate
specification for a communications protocol, which handles all the data
PROTOCOLS 269
collection and high-speed control functions, has been evolving quite

rapidly.
The equipment makers continue to modify and update the
implementations in each of the products. Many US and overseas utilities
have signed up to demonstrate UCA substation systems. The users can see
an impressive and elaborate demonstration of interoperability among a
broad variety of equipment from competing manufacturers in meetings
held several times a year by the UCA Substation Initiative Project. The
importance of achieving interoperable communications has forced collegial
cooperation among competitors, who see the individual-product features
and performance as the proper ground for competition [12, 19].
The layers of the seven-layer stack are logically separated to ease
troubleshooting and to allow levels to be replaced without affecting
neighbouring levels. The more sophisticated Application Layer provides
the services necessary to perform data acquisition and control in the
substation and also allows data sharing. This layer is expensive to develop
and needs to be maintained longer than the quickly changing Physical
Layer. The Physical Layer describes the signal transmission media
independent of the communications protocols. The middle five OSI layers
are often referred to as the Protocol Stack. The Protocol Stack describes a
combination of protocols that work together to achieve network
communications. Three common stack combinations are the OSI protocol
stack, the combination of the Transmission Control Protocol (TCP) and
Internet Protocol (IP), and the combination of the User Datagram Protocol
(UDP) and IP. The relationship of these protocol stacks within the
Application and Physical Layers is shown in Fig. (3.4) [26].
Possibilities:
· IEC 870-5
· DNP
· Profibus
· Other
Application layer FTP Telnet MMS Other
TCP UDP
Protocol stacks OSI
IP
Physical layer Physical connection
FIGURE 3.4 Parallel protocol stacks [8,20]

UCA’s main difference from the previously designed and used

protocols is its use of object models of devices and device components.
The common data formats; identifiers and controls for substation and feeder
devices can be defined with the use of object models of devices and device
components. The models specify standardised behaviour for the most
common device functions, and allow for significant vendor specialisation
for future innovation [19, 26]. The three levels of UCA are shown in Fig. 3.5.
Device
Data objects
Service interface How to describe data/devices

How to access?
Communication profiles Communication channels
Data Data
Data on the wire
FIGURE 3.5 The three levels of UCA [22]
As shown in Fig. 3.5, the UCA comprises the data object models
(forming the highest level), the service interfaces to these models (defining,
retrieving, reporting and logging of process data, controlling devices, file
transfer) and the communication profiles.
The direct data acquisition and control of field devices is an area,
which has been undergoing significant transition. Traditionally, the end
field devices were directly connected to Remote Terminal Units (RTUs),
which provided a network interface and performed initial processing of
the acquired data.
The introduction of the microprocessor technology has allowed for
the direct network access to the devices as well as more processing being
performed at the end device by leading to the development of Intelligent
Electronic Devices (IEDs).
The cost of integrating the end devices has been increasing rapidly
since the end devices (IEDs) became more and more complex in time due
to the advancements in the technology. Within the UCA framework, the
device object model is referred to the definition of the data and control
functions made available by the device, along with the associated
algorithms and capabilities [19, 26, 27].
Number of efforts has been initiated to develop detailed object
models of common field devices, including definitions of their associated
PROTOCOLS 271
algorithms and communications behaviour visible through the

communication system. Modelling efforts within the customer interface
area are also in progress.
These efforts include metering and interfaces to residential and
commercial customer devices. There has been active industry participation
in the customer interface modelling efforts. Significant work has been
accomplished as part of several UCA pilot projects and preliminary results
are available in draft form. The device models developed within the UCA
2.0 effort describe the communications behaviour of the devices by making
use of a common set of services. The detailed interoperable structure for
utility field devices can be fully specified by mapping these services onto
the UCA Application Layer Protocol (ALP) when used in conjunction with
the device models. The services and mapping to MMS are defined in UCA
Common Application Service Models (CASM).
An added benefit is that CASM simplifies the integration efforts
across functional areas of the utility. Moreover, device models can be
specified independent of the underlying protocol. Active participation of
groups outside the UCA activities has been encouraged due to this feature
of protocol independence, which also simplifies migration through the
construction of gateways to older existing protocols [19, 26, 27].
Vendors of utility field devices begin with their existing Vendor
Product Specification, which defines the functionality that the device
performs as shown in Fig. 3.6. The appropriate model of the field device
can be chosen from the various UCA Utility Standard Device Models, such
as Generic Object Models for Substation and Feeder Equipment (GOMSFE).
The vendor arrives at a product model by selecting from the optional model
components, adding their specialization to their existing product
specification. The product model defines the communications behaviour
of the vendor products in terms of the common application service models.
The mechanisms for representing the application services in the
underlying UCA application layer protocol is described by the UCA
Common Application Service Models (CASM) document. Next step is to
produce an application layer, which completely specifies the application
layer communications software required to support the product as a UCA
compliant device, by the specific mapping of objects and services into the
application layer protocol. Finally, the user selects the appropriate UCA
profiles to be supported from the UCA Profile Specification, Version 2.0
and determines the lower protocol layers by taking the expected operating
evironment of the device into consideration. The selected profiles, combined
with the Application Definition, forms the final Product Design. The Fig. 3.6
illustrates how the UCA Version 2.0 is implemented for field [27]. A device
is conformant to the UCA specifications only if it incorporates the following
three distinct specifications:
• The appropriate UCA object mode
• One or more UCA profiles
• The appropriate mapping of the Common Application Serivces
used in the Object Model to the application layer protocol.
UCA documents
Vendor documents Standardized
IEC 870-6-503: services,
Vendor product
Vendor product Selection of TASE.2 TASE.2 services mapping to
information,
specification building blocks and protocol application
user
layer
require-
ments
Specific
IEC 870-802: object types
TASE.2 TASE.2 object
TASE.2 Service Specific mapping to be
Services models
interface of objects supported
supported
UCA version 2.0

Specific UCA
Application profile
TASE.2 data Selection of Version 2.0
definition specification
objects profiles
Software Product
implem- design
entation
FIGURE 3.6 Definition of UCA field devices [23]
The number of UCA based IEDs for protection and control available
on the market is continuously growing and they are starting to appear in
installations around the world. However, the significant amount of
conventional microprocessor based protective relays installed earlier in
the substations need to be considered. There are some significant differences
between the UCA based IEDs and conventional microprocessor based
protective relays. This requires good understanding of the fundamentals
of communications based substation protection devices and at the same
time the availability of proper configuration tools that will make it easy
for the user to adapt the relay for its application in the substation [28].
3.4.2 lEC 61850 Project
IEC 61850 is based on the need and the opportunity for developing standard
communication protocols [30] to permit interoperability of IEDs from
PROTOCOLS 273
different manufacturers. Utilities also require IED interchangeability, which

is the ability to replace a device supplied by one manufacturer with a device
supplied by another manufacturer, without making changes to other
elements in the system. IEC 61850 makes use of existing standards and
commonly accepted communication principles, which allows for the free
exchange of information between IEDs. It considers the operational
requirements since any communication standard must consider the
substation operations functions. However, the communication protocol
standard IEC 61850 focuses on neither standardising the functions involved
in substation operation nor their allocation within the substation
automation systems which are used to define the impact of the operational
functions on the communication protocol requirements. Substation
automation systems normally incorporate functions, which can be assigned
to three levels: the station level (level 20), the bay level (level 1) and the
process level (level 0), for control, supervision, protection and monitoring
of the high voltage equipment and the grid. The physical mapping of logical
interfaces forms the communications between these levels as shown in Fig.
3.7 and forms the basis for the IEC 61850 standard series [25].
Remote Technical
control services
Station
FCT.A FCT.B
level
Bay
Bay unit
level
Prot. Cont. Prot. Cont.
Remote protection HV equipment Remote protection HV equipment
FIGURE 3.7 Logical interferences in a substation [31]
IEC 61850 identifies all the known functions in a substation

automation system and splits them into sub-functions or so-called logical
nodes. A logical node is a sub-function located in a physical node, which
exchanges data [29] with other separate logical entities. In IEC 61850, all
logical nodes have been grouped according to their most common
application area, a short textual description of the functionality, a device
function number if applicable and the relationship between logical nodes
and functions [25]. IEC 61850 decouples applications to design them

independent from communication so they are able to communicate by use
of different communication protocols. This is due to the fact that the vendors
and utilities have maintained application functions that are optimised to
meet specific requirements and that have reached a high degree of maturity
and quality.
Therefore, IEC 61850 provides a neutral interface between
application objects and the related application services allowing a
compatible exchange of data among components of a substation automation
system. Fig. 3.8 shows the basic reference model [28]. One of the most
important features of IEC 61850 is that it covers not only communication,
but also qualitative properties of engineering tools, measures for quality
management and configuration management. This is necessary since when
the utilities are planning to build a substation automation system with the
intention of combining IEDs from different vendors, they expect not only
interoperability of functions and devices, but also a homogenous system
handling.
Application
ACSI
Neutral Abstract
interface communication
service interface
SCSM 1 SCSM 2 SCSM n ‘‘Specific communication

service mapping’’
Specific
interfaces
AL 1 AL 2 AL n
Application layer 7
Layers 1.6
Communication stacks
FIGURE 3.8 The basic reference model [25]
Quality assurance for system life cycles is one of the important

aspects covered by the IEC 61850, which defines the responsibilities of
utilities and vendors [29, 31, 32]. Guidelines for environmental conditions
and auxiliary services with recommendations of the relevance of specific
requirements from other standards and specifications are also defined [25].
PROTOCOLS 275
For a long time, one of the major problems faced by protection

engineers to use IEDs to their full extent was the proprietary nature of the
communication interfaces. It was quite impossible to connect multiple IEDs
from multiple vendors without the use of special gateways and converters,
which tend to limit the functionality of the overall system. The concept of
logical nodes with a standardization of data contained within a logical
node in so-called data objects allows interoperability between IEDs or in
other words plug and play capability of IEDs in order to share information
and commands on a single network. Fig. 3.9 illustrates the relationship
between the process and the communication interface [25].
Object dictionary of a device contains all accessible information
Process
Application view
Binding
Logical node obejct
Data objects Data objects
Objects
According to
Data objects Data objects
7-4 and 7-3
Communication view
Services by which the information can be

accessed or manipulated
Communication objects and services according to

7-2 mapped to a SCSM
Network
FIGURE 3.9 Relation process and communication interface [25]
Applications can be defined using the standardized data without

knowledge about the actual device. This is because with the use of
standardized objects, the data contained in a device and the data available
on the network for further use is know up front and the naming of data is
independent of the actual device. As long as the logical nodes, data classes,
data object and data elements are implemented as specified in the standard,
we know up front where data will be present from a communication point
of view. The plug and play becomes possible after we add the capability of
self-description of logical nodes and therefore those of the devices.
Manufacturers need to provide devices containing extensions of functions
that are not yet modelled in IEC 61850. With the rules contained in EIC
61850 on how to model extensions, the data contained in these specific
extensions can be made available over the communication network in a
predefined way assuring interoperability [25].
With the plug and play capabilities embedded in the standard and
the immediate prove of concept in pilot projects, IEC 61850 promises to be
a great step forward in the development and acceptance of substation
automation systems world-wide. This will finally bring the real benefits of
automation and integration to utilities that were originally promised years
ago [25]. A non-proprietary, standard, high-speed protocol offering
sufficient services was required to enable a robust, integrated substation
communications network without protocol converters.
The introduction of IEC 61850 and the Utility Communications
Architecture has made it possible and justifiable to integrate station IEDs
through standardization. Using the standardized high-speed
communications between IEDs, the utility engineers can eliminate, many
expensive stand-alone devices and use the sophisticated functionality and
the available data to their full extend [25, 35].
3.4.3 DNP-3[153]
DNP-3 (Distributed Network Protocol) is a set of communications protocols.
It plays an important role in SCADA (Super Control and Data Acquisition)
system, where it is used for communications between system components.
The protocol was specifically developed for facilitating communication
between data acquisition and control devices.
DNP3 is an open, intelligent, robust, and efficient modern SCADA
protocol. It can:
• request and respond with multiple data types in single messages,
• segment message into multiple frames to ensure excellent error
detection and recovery,
• include only changed data in response message,
• assign priorities to data items and request data items periodically
based on their priority.
• respond without request (unsolicited),
• support time synchronization and a standard time format,
• allow multiple masters and peer-to-peer operations, and
• allow user definable objects including file transfer.
DNP-3 Provides Multiplexing, Data Fragmentation and More:
• DNP-3 is a layer 2 protocol. This means that it provides for
multiplexing, data fragmentation, error checking, link control, and
PROTOCOLS 277
prioritization. It also provides layer 2 addressing services for user

data.
• DNP-3 enables the various devices in process automation system
to communicate. The DNP-3 protocol is widely used in the
electrical and water industries by utility companies. It is also
possible for DNP-3 to be utilize in other areas, through it is not as
common.
DNP-3 Protocol Facilitates SCADA Communications:
• SCADA system use the DNP-3 protocol for communications
between various system components. The DNP-3 protocol for
communication between the SCADA system master, the system’s
RTU’s (Remote Terminal Units), and IED’s (Intelligent Electronic
Devices).
• DNP-3 was developed to meet the need for a standard
communications protocol that would allow for communication
between SCADA system components developed by differing
vendors. Using IEC 60870-5 as a base, DNP-3 was created as an
open protocol for use in these situations. This protocol was
available for immediate implementation within SCAD networks,
and catered to the specifications laid out by North American
organizations.
DNP-3 Provides Communication Reliability for Utilities:
• DNP-3 ensures the reliability of communications within the harsh
environments of electrical and water utilities. The protocol is able
to avoid being distorted by EMI, legacy system components, and
poor transmission due to DNP-3’s specifically designed
communications format. Although the protocol was created with
reliability in mind, DNP-3 is not secured, an important
consideration during SCADA planning.
• One can protect important DNP-3 communications and SCADA
system with a network alarm monitoring system. The most
advanced alarm masters can monitor in the alarms in a single,
convenient browser window. Sending one’s alarm notification via
email or page, the alarm master will inform instantly if there is a
problem with operation.
3.4.3.1 Typical Australian Practice (viz. SP Ausnet) of Connecting
Between IEDs and RTU
The most common connection is RS485 and RS422 (multi-drop) with
protocol. Each IED has an unique DNP3 address. Master sends massage
with destination address. All IEDs on the bus receive the message but only
the matching address IED responses. In SPAusNet, only a handful of
analogue circuits is used for SCADA communication between RTU and
Control Centre. SP AusNet Digital Network SDH/PDH (circuit switch) is
used to provide all communications requirements for Protection, Controls
and SCADA. It also builds a separate Ethernet GigaBit network (where
possible) to provide corporate, engineering access to IEDs at Terminal and
Zone substations. Future development will be Operational IP network to
provide SCADA on IP.
CHAPTER
Middleware
4.1 INTRODUCTION
Power system utilities have to increase the efficiency and effectiveness of
their communications and control systems to become faster, flexible and
more productive. The best solution to achieve this is to make more efficient
use of the information and telecommunication systems. The progressing
deregulation of the sector has been possessing new demands to integrate,
consolidate and disseminate real-time information quickly and accurately
within all kinds of company systems. Therefore, there is a further need to
utilise the data provided by the Substation Automation (SA) system to
extend it by enhancing information processing and management on system
level. Furthermore, coordination of the SA and network control levels is
also crucial.
Over the last years, there have been significant standardization
efforts in order to prepare standards for power system control equipments
and control systems. Interoperability between vendors and systems is the
key for standardization. The new IEC standard IEC 61850 is expected to
solve some of the problems. However, the existing standard remote control
protocols are not well suited for transmitting between the substations and
system control system. Hence, there is a need for improving the
coordination of communication protocols at station level and for remote
control, which will improve the vertical integration, and the cost benefit of
SA.
Heterogenous software systems, computing and communication
components form the basis of future computing platforms, which are also
subject to dynamic changes in resource availability. Distributed applications

such as the SA system applications have dynamic behaviours with respect
to their computation and communication needs. There is a need to create
software abstractions, tools and methods for building efficient component-
based applications for such feature platforms [40-43].
4.2 AUTOMATION SYSTEMS AND COMMUNICATION

NEEDS
Nowadays, the majority of the automation systems developed by
manufacturers have a similar architecture, except for a few differences. A
central computer is usually connected to decentralised computers and
protection relays as well as the synchronisation and communication
components. A Local Area Network is used, for system operation, with a
Human-Machine Interface (HMI) in order to control and monitor the system
and the processes. Such architecture is shown in Fig. 4.1, which shows the
architecture of substations in Europe. There are four different architecture
alternatives that could be chosen. The first two architectures differentiate
themselves from each other by means of a central computer. The former
has a central computer where the latter has no central computer and the
database is distributed. This section aims to investigate the specific points
of the communication networks in automation system.
Communication networks used in automation substation systems
can be used for adapting automation systems to the SCADA protocol and
HMI and storing data
Gateway
OI
Substation network
Lead Master
Controller
(optional)
Substation bus
Protection unit
Bay controller
Bay 1 Electrical Process Bay 2 Bay 3
FIGURE 4.1 Common substation architecture in Europe

MIDDLEWARE 281
to the various IED communication protocols. Nowadays, the comm-

unication network developments are mainly concerned with standards
used and particularly the interoperability between automation systems and
the equipments connected to them. Although some projects provide
interoperability that is far more extensive than that concerning IEDs, it is
commercially impossible for a central computer from a given manufacturer
to cohabit with the decentralised modules of another.
4.2.1 Types of Architectures
In general, the term ‘power system’ describes the collection of devices that
make up the physical systems that generate, transmit and distribute power.
The term ‘instrumentation and control (I & C) system’ refers to the collection
of devices that monitors, controls, and protects the power system [40-44].
4.2.1.1 Four Levels of the Instrumentation and Control System
The I & C system is composed of four levels as shown in Fig. 4.2. They are:
1. Process bus: The lowest level of I & C devices considered as the
process level, are physically connected to power system and are sensing
their current status. These include Current Transformers (CTs) to sense
current, Voltage Transformers (VTs) to sense voltage and Resistance
Thermal Detectors (RTDs) to sense temperature as well as various sensors.
Transducers are also the process level devices that convert the sensor output
of the above devices from one level to another.
2. Bay level: The next level is the bay level composed of IEDs that
collect the sensor data in order to create information from it and react to it.
PC To network
control Station
centre level
Interbay
bus
Integrated Bay control, Additional Bay

control and feeder products level
protection protection
Process
bus
Process
CD drive CTS/PTS Power transformers level
FIGURE 4.2 Modern substation

A bay refers to an area where a power system device such as feeder breaker,
and all of the I & C devices associated with it are located. These power
system IEDs include protective relays, meters, fault recorders, load tap
changers, VAR controllers, Remote Terminal Units (RTUs) and
Programmable Logic Controllers (PLCs).
3. Station level: Substation controller refers to devices that perform
data acquisition and control of IEDs and contain local I/O. They contain
data for the entire station. RTUs, PLCs, bay controllers, and Human
Machine Interface (HMI) software running on a personal computer are all
possible substation controllers.
4. Enterprise level: This is a generic term for all of the end users, or
clients of power system data inside and outside of the substation. These
applications acquire data from station level and unit-level devices.
For example, the three-fold purpose of a utility’s system integration
task can be summarised as follows. Transfer sensor measurements and
information created from this data among IEDs, between IEDs and a
substation controller and to end user clients directly from the IEDs and the
substation controller.
4.2.1.2 Example Communication Architectures of Power Systems
The communications architecture needs to be capable of data acquisition
and control to and from each IED in the substation. The following sections
detail the types of architectures used in substations.
4.2.1.2.1 Multidrop Network Architecture
The most common communication architecture used today is the multidrop
network or bus network shown in Fig. 4.3. As it can be seen from Fig. 4.2,
all devices are connected to the same physical wiring bus. Network master
is responsible for controlling the communications by issuing permission
commands to IEDs sharing the cable. An IED can only communicate when
it receives the virtual token and then passes the token when it is finished.
The most important advantage concerned with this kind of architecture is
that simple and fast peer-to-peer connections are possible. However, there
are major disadvantages as well such as the fact
Substation
FIGURE 4.3 Relays connected in the bus topology

MIDDLEWARE 283
that it does not allow simultaneous data polling of IEDs. The long-term
trend is away from multidrop networks and towards star networks.
4.2.1.2.2 Star Network Architecture
Many direct connections originating from one device is called a star network
topology as shown in Fig. 4.4. Any protocol, including those designed for
multidrop applications, can be used for direct communications in a star
topology. In this architecture, slow communicating devices can coexist with
more complex fast communication relays. Therefore star network supports
a wide range of IED capabilities.
Communications
FIGURE 4.4 Relays connected in the star topology
4.2.2 Wide Area Network Structures in Substation Automation

Wide area functions are used for the execution of protection and/or control
functions within a substation, which needs to make use of the information
coming from a more extensive area covered by the substation itself to
perform correctly. In the last decade, we have seen common incorporation
of LAN and WAN technologies in the Automation and Control of electric
power networks with devices and protocols within and outside the
substations being adapted to the use of Local and Wide Area Networks.
Ethernet, a famous computer network, is used, for communication between
IEDs offering high speed, high transfer capacity and versatility.
However, certain factors, nowadays still prevent the practical
implementation of such functions. These factors are mainly security,
availability, communication speed and response time. The performance
required is measured by the load imposed on the communication medium,
speed and functionality limitations imposed on communication by the
protocols and the suitability of the software applications in the devices to
make full use of the communication capabilities. Many applications aimed
at in wide area functions require on-line information at a speed only feasible
with very strict limitations and conditions of the communication network
[41-44].
When the conditions as mentioned are considered, it will be clear

that the incorporation of computer networks in power network automation,
control and protection represents a huge potential for new and more
efficient methods to be implemented with a huge potential of reducing
costs and increasing efficiency.
Applications can only be automated on a substation level when the
need for human interface is excluded or at least minimised, the reason
being the ongoing trend for unmanned substations. The exchange of data
requires the information to be made available to the network by IEDs
connected to the network and a data model to enable the processing of
these data in the desired applications. Recent developments aim not only
at control devices such as relays but on a broader range of IEDs such as
measurement devices and primary equipment. The feasibility of wide-area
applications will depend heavily on the realisation of standard
communication and standardized modelling of IEDs.
The UCATM 2.0 initiative in the USA is an attempt to meet these
requirements. IEC61850 takes these efforts one-step further; incorporating
the UCATM 2.0 works, and extending it towards the process level. The
harmonisation of data models between UCATM 2.0 and IEC 61850 will
eventually be an important step towards a worldwide-accepted standard.
Timing and performance together with the security and availability factors
are the important factors determining the feasibility of wide-area
automation, control and protection functions. Since the IEC 61850 is still
in the process of becoming a standard and the incorporation in devices is
not yet available in commercial products.
The UCATM 2.0 concept has been widely accepted in the USA with a
number of prototypes being introduced or planned to be introduced tested
extensively by the Utility Initiative, which is a consortium of utilities and
vendors.
4.2.3 Communication Standards and Applications

The strong technological development of large-scale integrated circuits
leading to the present availability of advanced, fast and powerful
microprocessors making substation automation possible, which resulted
in an evolution of substation secondary equipment. The evolution of
substation secondary equipment, in turn, has made it possible to implement
decentralised substation automation systems, using several IEDs to perform
the required functions such as remote monitoring and control. As a result,
the need arose for efficient communication among the IEDs, especially for
a standard protocol.
MIDDLEWARE 285
The vendor specific and hardware oriented structure of the market

has lead to a number of protocols for communication making
communication between IEDs of different manufacturers impossible or
only with disproportionate expenditure. The advancement of the
connectivity and interoperability of systems is only possible with
standardization since a reduction of variety in a relatively small market is
extremely beneficial for both the vendors and users.
The deregulation of the energy market has increased the
importance of data and a lot more data is necessary. Thus open and
standardised communication interfaces are needed. This fact is expected
to have enormous influence on rebuilding and renewing substations,
especially of control components and of the way the process bus is
connected to the control system. In the future, the primary equipment must
also be equipped with serial communication interfaces. The expected
changes are listed below:
• The importance of standardised hardware and software is
increasing.
• A maximum integration of functions within multi-function devices
is targeted since the experiences have pointed out that the
maximum allowed degree of integration has not yet been fully
realised. Therefore, considerable scope still exists for further
integration of functions up to the maximum allowed levels.
• The medium term target in substation automation will be towards
the combination of functionality’s to reduce the number of
hardware units and the achievement of an information system.
Implementation of interfacing modules between a particular
manufacturer process bus and legacy IEDs is seen to be the
intermediate step for the solution of problems in the short-term.
In the longer term, this is not advised since it adds further
installation costs. Besides, the possibility of a wireless substation
structure in the long-term means that the bay level devices will
be able to communicate directly with the primary plant by means
of a process bus [45-46].
4.3 MIDDLEWARE REQUIREMENTS FOR PROTECTION

APPLICATIONS
The need to integrate protection, control and data acquisition on the
substation Local Area Network (LAN) using a standardised communication
protocol has been recognised since the early 90s and many attempts have
been made since then to define the standardized communication protocol.
Many communities such as IEC, CIGRE and IEEE have focussed on

developing the specifications for peer-to-peer communication between
IEDs. In order to achieve interoperability between IEDs of different vendors,
i.e., to be able to communicate between IEDs built by different vendors,
the communication protocol must be constraint. This section reviews the
important constraints when choosing a protocol offering an option for
implementing more than one protocol within the substation. There are three
basic sets of requirements when choosing a communication protocol.
They are:
1. Performance: The most important performance requirement is
the protection commands over the LAN between IED applications within
4ms. This requirement then defines both a peer-to-peer communication
protocol and either a 100Mbps shared or 10Mbps-switched ethernet
implementation. The 4ms requirement between a sending application in
one slave IED and the receiving application in another slave IED cannot
be met with a framework implementing a master-slave communication
protocol or with a taken passing protocol since the latter has to wait for
the token to pass the data packet.
2. Interoperability: When the communication over the substation
LAN, which is used for control and monitoring only, is considered then
interoperability clearly becomes the dominant selection criteria. In cases
where protection does not need to be integrated with control and
monitoring over the substation LAN, then the selection of the
communication protocol does not strongly depend on the time required to
communicate a message from the sending IED application to the receiving
IED application. However multicast is still required for interlocking and
other control functions.
Interoperability is only achieved when the information exchanged
between IEDs is fully understood and unambiguous, which requires a well-
defined data model that specifies both the syntax and semantics of the
information exchanged. In other words, the protocol specification must
define the rules and building blocks for developing extensible objects that
will be communicated between IEDs. The interoperability requirement has
to be implemented in all future substations including the new ones as well
as the retrofitted ones. The protocol, which includes a well-defined set of
rules for implementing the data model, has to be selected with a project
notebook imposed on all vendors to ensure that every nuance of the build
out of the data model is understood and documented. The project
notebook will be governing document for defining the ‘as-built’
specification. Therefore, future changes to the substation automation system
should then be based on the project notebook [43-46].
MIDDLEWARE 287
3. Maturity: The most important criteria mitigating cost and schedule

risk for depolying selected communication architecture is maturity with
the most important input to the maturity analysis being the clarity of the
utility’s vision for substation automation. By comparing installed user bases
for a candidate protocol, maturity can be easily measured. However, care
must be taken since technology is rapidly changing and each installed user
base used in the comparison must be restricted to one version of protocol
specifications. The one version should include all the capabilities needed
for a specific utility’s vision of substation automation. The solution is not
to choose someone, who has built-in bias for a particular technology or
solution, for performing the maturity analysis. The person selected needs
to have a high degree of knowledge in the following areas:
• A good understanding of the candidate communication
technologies under consideration, which is best measured by
current participation in venues that are developing these
technologies.
• Modelling tools that can be used to develop a baseline, without
which the comparative analysis will be terribly confused and
useless to the decision-maker, from the utility’s vision for
substation automation.
• It is technically possible to implement more than one protocol in
substations by providing gateways for needed protocol
conversion, a technique that will always increase cost and reduce
performance and reliability of the communication architecture.
However, gateways are unavoidable in situations where the
required functional capabilities of substations can only be
achieved with the use of gateways. Gateways are unavoidable
when:
— IEDs within the substation operate with a different
communication protocol than those IEDs outside the
substation (external substation IEDs) or on the EMS/DMS
nodes.
— Substation LAN segments operate with different protocol,
which will probably be the case when migrating from legacy
systems to the communication protocol of choice.
— Communication segments, streaming data from the substation
yard into the substation control house, use different protocols
for their operation. If a ‘Process Bus’ as described in IEC 61850
is implemented, this could be the case since the process bus is
used to stream data from the yard and is not burdened with
substation LAN protocol overhead.
Provided that the LAN internal to the substation has sufficient

bandwidth, then more than one protocol can exist. However, it is crucial to
ensure that all required peer-to-peer communication operates correctly.
The best approach is performing integrated system testing under moderate
loads to determine the behaviour of each IED when it receives a message
communicated using a foreign protocol specified in the message header
[43-46].
4.4 MIDDLEWARE ARCHITECTURES

A new class of software, called middleware, has risen to address the
challenge of interoperability. Middleware software is a layer between the
networking code and the application code provided by the communication
processor. The function of the middleware is to insulate the application
programmer from the raw networking code thus providing an easier way
to communicate.
In this section we will discuss middleware requirements for remote
monitoring, protection and control applications. The requirements that are
not largely and sufficiently supported by present middleware platforms
are only outlined below.
The key requirements identified are as follows:
1. Event notification
2. Fine grained time synchronisation
3. Security
4. Naming and directory services
5. Mediation services
6. Unit and time conversation
7. Distribution configuration management
8. Dynamic substitution of computers
9. Debugging support for distributed systems
10. Real-time object request brokers
11. Performance enhancements in general.
The above requirements can be classified into more general
categories as follows:
• Basic services
• Security
• Distribution system management
• Performance.
MIDDLEWARE 289
4.4.1 Types of Middleware Architectures

The use of the appropriate middleware is crucial for achieving success in
substation automation. There are three main types of middleware
architectures, which are:
1. Point-to-point architectures.
2. Client-server architectures.
3. Publish-subscribe architectures.
4.4.1.1 Point-to-Point Architectures
This is simplest form of communication. The communication between the
Intelligent Electronic Devices (IED) within the substation is perhaps the
most familiar example of point-to-point communication. It can only be
used when the initiating IED or in other words the calling IED, knows the
address of the responding IED. They can have a two-way communication
dialog as soon as the connection between them has been established.
However, this type of connection is not useful in large substations where
one IED needs to be able to talk to several IEDs simultaneously. So it is
only designed to support one-to-one communications.
4.4.1.2 Client-Server Architectures
Client-Server architecture has a many-to-one architecture in which one
special ‘server’ node can connect simultaneously to many client nodes.
Client-server architectures are useful when all the nodes on the network
need to access centralised information. However, this type of architecture
is inefficient since they require all information, which is being generated
at multiple IEDs, to be sent to the server before it comes accessible to the
clients. Another disadvantage concerned with this type of architecture is
the unknown delay being added to the system since the receiving client
does not know when new information has been added to the server.
Substation database of configuration parameters and transaction processing
between two relay IEDs are two common examples of this type of
architecture [45-48].
4.4.1.3 Publish-Subscribe Architectures
With Publish-Subscribe architecture, an IED can perform two main tasks
enabling direct message exchange between the communicating IEDs. An
IED will either.
• Subscribe to data that it needs or
• Publish information that it produces.
Any authorised IED may add itself as a subscriber to a particular
publisher’s list. That subscribing IED will then receive the publications
directly from that publisher IED, as they become available. Publish-
Subscribe systems are useful since:
• They are good are quick distributors of large quantities of time-

critical information even when unreliable delivery mechanisms
are present.
• They can handle very complex data flow patterns.
One of the important properties of the publish-subscribe middleware
is that the application software in data sources and data sinks are kept
independent of each other. The most important of all is that it (middleware
layer) handles connections, failures and changes in the network eliminating
the need to handle exceptions. Middleware only delivers the data that has
been requested by the application software.
4.4.2 Real-Time Substation Applications and Requirements

Distributed real-time communications in the substation environment can
well be realised using the publish-subscribe architectures. Publish-subscribe
architectures make the most efficient use of the networking resources since:
• There is no need to request data.
• The data transfers are done directly between the publisher and
the subscriber.
In addition to making full use of the network resources, publish-
subscribe architectures also provide low latency delivery since the data
can be sent from the publisher to the subscriber as soon as it becomes
available. Moreover, they can take advantage of the new multicast
capabilities of modern communication networks.
Continuous sampled data from instrument transformers, group
communication of state changes (the famous GOOSE message) and reliable
status updates are the data transfer requirements of distributed real-time
systems that can be supported by the publish-subscribe architectures when
they are configured properly . However real-time systems have other needs,
which cannot be met by the publish-subscribe architectures. Some of these
are:
• The ability to trade off delivery reliability against delivery delay.
• The need for communications processors to handle the unique
behaviour of the real-time communication protocols.
4.4.2.1 Delivery Delay against Delivery Reliability
When communication protocols are designed for guaranteed delivery but
a less-than-reliable medium, then an important problem arises in the case
of failed transmissions. What happens is that the communication protocol
would be struck trying to transmit the failed transmissions wasting time
and destroying the timing determinism. For example, the process of
MIDDLEWARE 291
trading-off reliability of delivery for greater determinism is crucial for

multicasting GOOSE state change messages. Sending the most recent
GOOSE message is much more important than resending old updates,
which would probably be out-of-date when they are delivered anyway. In
this case, the best policy would be to send the latest update disregarding
the earlier updates [46-51].
In contrast, when a sequence of device operation commands is
required then a communications processor has to receive every step in the
command sequence properly, which can only be quarantined with reliable
delivery.
In some situations, protection applications require some intermediate
action, which is unfortunately not specified by any of the current
networking protocols. For example, Transaction and Control Protocol (TCP)
refuses to accept any subsequent packets for several minutes until a
dropped packet has successfully been delivered to the TCP. Hence, TCP
cannot provide deterministic timing such as retrying for 100 ms and then
moving on.
4.4.2.2 Synchronisation
In Electric Control and Protection applications, there is a need for a
synchronisation time of 10 ms. There is such a stringent requirement for
time synchronisation across the distributed system particularly due to the
phase differences among generators. An adaptive source rate control
mechanism can be developed to handle the changes in the effective
synchronization scheme, which can be used to compensate for long-term
delay variation caused by large-scale fading. Hence the synchronisation
will be maintained keeping the end-to-end delay in low values.
4.4.2.3 Event Notification Services
In any remote monitoring, protection and control systems, there is a clear
mechanism of one type for the notification of interesting events, when
remote sites notify the monitoring sites. Ideally, important events (such as
power system faults) need to be pre-defined so that the system can listen
for specific types of events to produce a synchronous, persistent and
multicast event notifications.
In a distributed system such as power system control and protection
system where system components (CVTs, relays, IEDs) can fail or in some
cases become unreachable, then we might need substantial technical
requirements for implementing such services. There are low-level event-
delivery protocols defined in Common Object Request Broker Architecture,
but however high-level event service behaviour still depends on the vendor-
specific products. What we need is to:
• To define and implement event service reliability and delivery
persistence capabilities.
• To identify and describe a significant subset of high level event

service functionality sufficiently well-described to allow multiple
implementations that exhibit interoperable behaviour.
4.4.2.4 Data Push Architectures
Data push architectures, when combined with publish-subscribe
architectures, enables users to distribute data to a large and variable set of
remote applications. The most important feature of publish-subscribe
architectures is that multiple applications, local to the publisher or at some
remote Internet site, can subscribe to data published by a single platform.
The most appropriate network protocol can be easily identified and used
for the network topology between the publisher and subscriber.
There are, however, problems when matching the pure, synchronous,
remote object method invocation model found in most ORBs to the push
model found in data push architectures. Individual data sources and control
objects are ideally mapped onto single software objects. When remote
applications are reading data out of the data sources, they make use of a
method called ‘remote method invocation’ on the data source’s software
object. Such a method is not really appropriate since the data push
architectures can only push data out at their own pace by performing object
method invocations at their own investigation. Another disadvantage is
that when data is collected from a variety of objects, it will be pushed out
as a single message ending up being treated in various special, ad hoc
ways. It has become necessary to bridge the gap between these two models.
If the data publishing protocol is formalised by making specific reference
to ORB interface repository entries that would tie elements of the data
message to specific method invocations on specific objects, then this
problem can be solved. When the elements of the data message are received,
they could easily be unmarshalled in accordance with interface repository
information [49–52].
4.4.2.5 Security Requirements
When monitoring and operating devices over the Internet, there are strict
security measures required for several reasons:
1. The monitored data can be stolen, corrupted and intentionally
falsified.
2. The devices can be used maliciously by impersonators.
3. The device can be used without unauthorisation.
4. The privacy of monitoring data has to be preserved for
commercial and national security.
MIDDLEWARE 293
Then there is clearly a need for new middleware functions in order

to enforce the security requirements detailed above. They are:
• Server authentication: It is needed to ensure operation on the
intended site.
• Client authentication: It is needed to ensure that an authorised
client/operator is operating the equipment.
• Confidentiality: It is needed for transferring data items in the
encrypted format preventing malicious and false operation and
eavesdropping.
• Non-repudiation of control operations to guarantee liability
[48-51].
4.4.3 New Features Required for Communication Protocols

There are many challenges, such as the ones outlined above, which need
to be solved. There is a clear need to design a real-time communication
protocol to run with a communication processor, which needs to have
features such as:
• Pre-emptive thread schedulers.
• Fixed memory managers.
• Predictable response times.
The required networking software can be effective running on the
communication processor only if it is capable of:
• Managing priorities.
• Control memory usage and timing.
• Restrict access to system resources.
For example, the networking software running on the communi-
cation processor must be completely event driven avoiding the latencies
and the inefficiencies of polling. Hence, it should not pause the execution
of a high-priority task just because that task issues a publication request.
Another important feature required is the re-entrancy of the software, which
allows for simultaneous accessing to networking devices due to the
possibility of many threads running at different priority levels.
4.4.4 Needed Communication Model

Clearly, there is a need to develop a formal communication model satisfying
the following features:
• It models time and timestamps each transaction.
• It allows the application software to trade-off timing against

reliable delivery.
• It controls and specifies memory usage.
• It allows for a adaptive synchronization scheme.
• It allows for event notification.
• It allows for the functions required to meet with the security
measures.
• It works in a real-time communication processor environment.
Only when the above conditions are met, then we can successfully
use publish-subscribe architecture in real time for protection applications.
Furthermore, each node in the network has to be able maintain a record of
its own internal subscribers and publication to which each subscribes; and,
its own internal publishers and the list of subscribers to which each sends
issues. TCP/IP protocol allows for the easy implementation of non-real-
time applications running over it. However, in order to implement real
time applications running over UDP/IP, the utilities require a network
data delivery service with small size and fast speed. UDP and IP are quite
simple and are reasonably fast. The network data delivery service
middleware, which needs to be implemented, should add only minimal
overhead to the underlying network communication stack and should be
much more efficient that TCP, DCOM, or CORBA. This distributed
approach not only simplifies the system configuration but also provides
for graceful degradation meaning that if any particular node is lost in the
network, this will not stop data transmission between unaffected nodes in
the rest of the system [41–46].
4.5 PUBLISH/SUBSCRIBE MIDDLEWARE

A Publish/Subscribe system, shown in Fig. 4.5, is a middleware
communication service supporting an asynchronous style of many-to-many
communication in contrast to the request/response type of synchronous
approach of object invocation. It relies on the preferences expressed by the
subscribers to deliver messages from one publisher to one or many
subscribers instead of the publisher relying on specific destination
addresses [41]. A publisher can also be referred as a producer, or sender.
Similarly, subscribers are most often referred to as consumers or receivers.
MIDDLEWARE 295
Publisher Event service Subscriber

1 1
Subscribe ( )
Publisher Subscriber
2 2
Publish Unsubscribe ( )
Publisher Push event ( ) Subscriber

3 3
FIGURE 4.5 Publish/subscribe communication model
Consumers make subscriptions using definitions of the information

they are particularly interested in. Producers create instances of
information, which will be forwarded to the subscribers of this information.
4.5.1 Different Subscription Mechanisms
There are three types of subscription mechanisms that consumers can
generally make us of when subscribing to information [41]. They are:
4.5.2 Channel-Based Subscription
This is the simplest form of subscriptions. Consumers can either subscribe
or listen to a channel. Channels will be sent copies of events in the
occurrence of new events, which in turn will be delivered to all subscribers
listening to that channel [41].
4.5.3 Subject-Based Subscriptions
This is the form of subscription mechanism where the idea of a channel
subscription has been extended with a more flexible addressing scheme.
The notification message include two different parts, which are:
• The subject attribute that determines the address.
• Followed by the event data.
The advantage of subject-based subscriptions is that consumers can
express interest in many channels or more than one subject. The subject of
the subscription will be assessed against the subject of the event message,
and those consumers with matching subscriptions will be forwarded with
relative events [41].
4.5.4 Context-based Subscriptions
The context-based subscription are an extended version of the subject-based
ones, where a consumer cannot only express interest in the subject of the
notification but also within the content of the notification. The main
advantage of the context-based subscriptions is that fact that the delivery
of uninteresting messages can be minimised or even avoided due to the
fact that the consumers’ capability of clearly expressing his interest has
been increased [41].
4.5.5 Routing Problem

The primary component of a Publish/Subscribe network is a routing
engine, which aims to make sure that each message will be delivered to its
potential subscribers, i.e. the consumers that have already notified the
Publish/Subscribe system of their interest [47]. The first dilemma when
facing the routing problem is to choose the most appropriate type of
messaging concept. There are three main messaging concepts, which are:
4.5.5.1 Unicast Messaging
Unicast messaging, shown in Fig. 4.6, requires the sender to send one
individual copy to each subscriber, which limits the number of available
subscribers as a result of the bandwidth limitations of the sender. Unicast
messaging is the least suitable concept when trying to achieve real-time
goals since it consumes bandwidth resources by creating a heavy load when
trying to achieve a number of point-to-point links. In addition, it is the
messaging concept in which message delivery delay is the least predictable
[48].
Sending host
Ethernet hub
Receiving Receiving Receiving

device device device
FIGURE 4.6 Unicast transmission
4.5.5.2 Multicast Messaging

Multicast messaging allows the sender to send a single copy to the data
stream, which will then be replicated and forwarded to the consumers,
which signaled their interest earlier on, by the network architecture.
MIDDLEWARE 297
Another subnet
Sending host
Another subnet
Ethernet hub
Receiver not interested in the

multicast does not receive it

FIGURE 4.7 Multicast transmission
Therefore, instead of sending thousands of copies, the sender sends a single

copy directed by the routers on the network to the consumers that have
indicated their interest in the message. Consumers usually indicate their
interest by joining a particular multicast session group. Thus, multicast
messaging reduces the amount to traffic over the network yielding an
increased efficiency for both the sender and the network with a number of
other performance improvements [48, 49, 50].
4.5.5.3 Broadcast Messaging
Broadcast messaging uses the concept of sending one copy of the message
to all nodes on the network (Fig. 4.8). With broadcast messaging, every
consumer on the network needs to process the message regardless of
whether or not the consumer is interested with the message. The main
problem concerned with this is the fact that many consumers might actually
be not interested with every message they receive increasing CPU usage,
which in turn reduces efficiency [48, 49, 50].
The problem of choosing the appropriate messaging for successful
routing is in general followed by the need to tackle the filtering problem,
which requires making sure that any subscriber does not receive more
message than what it has subscribed for. In the cases, where each publisher
is fully aware of its subscribers, then there is no need for filtering. This is
usually the case for multicast messaging. However, with broadcast
messaging where every consumer receives one copy of all messages, some
sort of filtering must take place in order to implement the Publish/Subscribe
model.
Another
subnet
Sending host
Another subnet
Receiver not interested

in the broadcast however
still receives it

FIGURE 4.8 Broadcast transmission
4.5.6 Filtering and Binding

Filtering is simply the decoding of the message against the subscriptions
in order to find out the relative subscribers that have shown interest in
that message. Therefore, the type of filtering mechanism to be chosen
depends on the subscription mechanism. For instance, when the concept
of context-based subscriptions is used where subscribers can express their
interest for the message contents, then the filtering mechanism needs to
evaluate the complete contents of each message, which adds a
non-predictable overhead on each node [47].
The relative overhead decreases when subject based subscription
concept is used. In this case, the message is usually tagged with its subject,
which has to be evaluated by the filtering process. The complexity of the
filtering will be reduced since the filter only needs to evaluate the subject
tag rather than all the message contents. A solution in Reformance [47] has
been proposed in order to make subject based addressing scheme more
efficient and usable in real time systems. In Reformance [47], this is achieved
by putting the subject in the address of the message instead of the message
contents, which is claimed to increase the efficiency of the filtering to a
great extent. Once the filtering problem has been tackled, the task of binding
has to be considered. Binding allows the system to find out the relative
addresses that it needs to use when forwarding messages. For context base
subscriptions, there is no need for any binding. However, for channel-based
and subject based subscription systems, some sort of binding needs to be
implemented. Binding problem is easily overcome in the event channel
based publish-subscribe systems such as CORBA event channel since
binding takes place when nodes are connecting to the event channel [47].
MIDDLEWARE 299
4.5.7 Publish-Subscribe and QoS

Although Publish-Subscribe communication model has many advantages
such as offering a weak coupling among the communication parties [51],
it also has a number of drawbacks such as the lack of support for the
negotiation or enforcement of Quality of Service (QoS) [51]. This is a
significant drawback because QoS features are nowadays very important
components of real time applications.
However, lately some significant research has been carried out in
order to incorporate the use of QoS features and techniques into publish-
subscribe systems [51, 52]. The subsection 4.5.7.1 aims to give reader some
background information about QoS features and techniques, while the
subsection 4.5.8 discusses some of the research work that has been done
with the intention of integrating QoS features and techniques into Publish-
Subscribe systems.
4.5.7.1 QoS Features and Techniques
Applications require certain network services to be delivered to at a certain
minimum performance level to be usable [53]. QoS refers to a network
system’s ability to sustain a given service at or above its required minimum
performance level [54]. Fig. 4.9 shows QoS architecture and components.
QoS techniques improves network performance level [55] by
Desired QoS
parameters
delay,
bandwidth,
reliability,
etc.
QoS signalling
Client
Host
Host
QoS techniques
signalling queuing, etc.
FIGURE 4.9 QoS architecture and components

• Supporting dedicated bandwidth

• Improving reliability characteristics
• Avoiding and managing network congestion by controlling jitter
and latency
• Setting traffic priorities across the network
• Shaping network traffic.
Delay is the time it takes a packet to travel from sender through the
network to receiver. As delay increases, the transport protocol becomes
less efficient. Jitter is the variation in the total end-to-end delay values of
different packets on the network. A high level of jitter is also undesirable
since it leads to the inefficiency of the transport protocol while at the same
time causing signal distortion. Bandwidth is the maximal data transfer
rate that can be sustained between two end points. Reliability can be
referred to as the average error rate of the medium. Poor reliability can
result in the transmission of packet in an order different than that of the
original transmission or even the loss of packets. Therefore, this situation
needs to be avoided [55].
When a network receives more packets than it can handle exceeding
its limitations, then congestion occurs, which results in a complete network
collapse where data packets are not transmitted at all [55]. Therefore, this
has to be avoided. Management tools help for providing QoS within a
single network element. The most popular congestion management
techniques are queuing, scheduling and traffic shaping and signaling. When
packets arrive at the transmitter interface faster than the transmitter can
transmit them, they will be queued until the interfaces are free to transmit
them. They will be scheduled for transmission according to their assigned
priority (Fig. 4.10) an end the type of the queuing algorithm configured
for the interface. Queuing and scheduling schemes together provide
predictable network service by
Low and medium priority High priority
dispense only when dispenses
high priority is empty until empty
Low Medium High

priority priority priority
FIGURE 4.10 Priority queuing

MIDDLEWARE 301
providing dedicated bandwidth, controlled jitter and latency and improved

packet loss characteristics [55]. The basic idea is to pre-allocate resources
(e.g., processor and buffer space) for sensitive data.
Each of the following schemes require customized configuration of
output interface queues:
1. Priority Queuing (PQ) assures that during congestion lower
priority traffic does not delay the highest priority data. However, as it can
be seen from Fig. 4.10, lower priority traffic can experience significant
delays. PQ is designed for environments that focus on mission critical data,
excluding or delaying less critical traffic during periods of congestion [55].
2. Custom Queuing (CQ) assigns a certain percentage of the
bandwidth to each queue to assure predictable output for other queues. It
is designed for environments that need to guarantee a minimal level of
service of all traffic [55].
3. Weighted Fair Queuing (WFQ) allocates a percentage of the
output bandwidth equal to the relative weight of each traffic class during
periods of congestion [55].
In addition to managing congestion, it is also important to try to
prevent congestion. Weighted Random Early Detection (WRED) algorithm
and is a congestion avoidance algorithm, which starts to drop low priority
packets to ensure the delivery of all mission critical traffic. This happens
only if WRED detects the possibility of future network congestion (Fig.
4.11). Therefore, it is very suitable for real time mission critical applications
(55).
4. Committed Access Rate mechanism (CAR) is a traffic shaping
mechanism, which defines a traffic contract in routed networks. CAR can
classify and set policies for handling traffic that exceeds a certain bandwidth
allocation. CAR can be also used to set IP precedence based on application,
incoming interface and type of service (TOS). It allows considerable
flexibility for precedence assignment [55].
4.5.8 Resource Reservation Protocol (RSVP)
Signaling is a form of network communication that enables network
elements to signal requests to its neighbours. Coordination between traffic
management and policing tools can be handled with the use of QoS
signaling. Reservation Protocol (RSVP) is a very complex signalling
protocol, which provides reservation set-up and control going far beyond
standard ‘best effort’ IP service to provide highest level to QoS in terms of
[55, 57]:
1. Service guarantees
2. Granularity of a resource allocation
3. Detail of feedback to QoS-enabled applications.
When the outgoing connection is congested,

RED discards lower priority and non-RSVP
packets randomly
FIGURE 4.11 Weighted random early detection [16]
RSVP can be summarized with a simple description [55]:

• Each sender defines the characteristics of traffic specification
(TSpeck) in term of upper and lower bounds of bandwidth, jitter,
and delay. The path (PATH) message originating from the sender
is sent to the receiver taking the traffic specification into
consideration is every step along the network.
• The reservation request (RESV) message from receiver to sender
not only specifies traffic.
• Each router receiving the RESV message allocates the needed
resources.
• The last router sends confirmation back to the receiver once it
receives the RESV message.
4.6 CORBA AND ITS FEATURES

The introduction of IEC 61850 and the Utility Communications Architecture
(UCA) has made it possible and justifiable to integrate station IEDs through
standardization. However, more advances are needed in order to establish
an open and standard working environment allowing for more and more
functions to be developed. The new forthcoming ideas need to follow the
MIDDLEWARE 303
path opened by the IEC 61850 in order to provide advanced complements

to this basic architecture. One of today’s most popular open and standard-
working environments is Common Object Request Broker Architecture
(CORBA) middleware with set of services, which also allows for
heterogeneous implementation of distribution applications. CORBA is an
object-oriented standard for distributed object systems, which is
implemented using the Object Request Broker (ORB) specification of the
Object Management Architecture (OMA). CORBA architecture consists a
client object, a server object, an Interface Definition Language (IDL) Stub,
an IDL Skeleton and an ORB. CORBA middleware makes the
communication between these various components possible through the
use of a method involving twelve steps of invocations.
The main emphasis in this section is the use of CORBA middleware
communications for power system communications. The section discusses
the advantages and disadvantages of the CORBA middleware architecture
for its use in substations and also the need for its harmonization with UCA.
The section aims to example how the standard CORBA middleware can be
exended to form an effective platform performance sensitive real-time
power system operations. It also deals with implementing the UCA protocol
over the newly designed CORBA application thus showing how the
interoperability requirements can be met while providing an effective
platform for performance sensitive substation automation applications.
4.6.1 Substation Automation Systems
The collection of devices that make up the physical systems that generate,
transmit and distribute power is usually referred to as the Power System.
Substation Automation (SA) is the use of the IED data within the substation
and control commands from remote users to control the power system
devices within the substation. A SA system is a distributed system dedicated
to the monitoring and protection of the primary equipment of such a
substation and its associated feeders [41].
When relays and IEDs are integrated together, they form a powerful,
economical Instrumentation and Control (I&C) system to support all aspects
of electric power protection, automation, control, monitoring and analysis.
Nowadays, the desire and the need to merge the communication
capabilities of all the relays and IEDs in a substation is clearly recognized,
which is capable of providing not only data gathering and setting capability,
but also remote control. Worldwide, electric utility deregulation is
expanding and creating demands to integrate, consolidate and disseminate
real-time information quickly and accurately within and with substations.
All these factors have lead to a strong customer drive for standardized
solutions [42].
The Utility Communications Architecture (UCA) specifies a suite of

internationally recognized open communication protocols, which meet the
requirements of the utilities industry, including electric, gas and water.
UCA should rather be regarded as an architecture than a simple protocol.
Although the specific object and operational models can be specific to any
one of the utilities mentioned above, the architecture remains applicable
to all with roughly the similar constraints.
Generic Object Models for Substation and Feeder Equipment
(GOMSFE) defines a set of object models for use within the UCA
architecture across a broad range of typical utility devices [43, 44].
The UCA Common Applications Service Models (CASM) provides
a common set of communications functions such as data access, data
reporting, data logging, and control functions, which are found in most
real-time utility field devices [43, 44].
Although CASM allows discrete devices to share data and services,
it is only an abstract application layer protocol without any real procedure
for sending and receiving data. It can only be usable when it is mapped, as
shown in Fig. 4.11, to a specific communication service such as
Manufacturing Message Specification (MMS) protocol, Distributed
Component Object Model (DCOM) or Common Object Request Broker
Architecture (CORBA) [43, 44].
At the application layer of this model, the MMS application layer
protocol is specified to provide the necessary messaging services for devices
in these environments. In addition, the UCA Station Management Protocol
is specified, which provides support for UCA Time Synchronization
services. However, in this book, we propose the use of CORBA to provide
the necessary communication services [43, 44].
CORBA is complementary to UCA since CORBA is concerned with
standard APIs while UCA focuses on the communications infrastructure.
Both worlds require the use of objects so they need to be harmonized. This
book investigates the mapping of CORBA services to CASM. The main
advantage of the use of CORBA comes from the fact that we can replace
MMS objects such as MMS-Events, Domains Download/Upload and
Semaphores while adding value to the communication system, e.g., Security,
Time Synchronization and Transaction.
4.6.2 CORBA Middleware
CORBA, shown in Fig. 4.12, is an object-oriented standard for distributed
object systems, which is implemented using the Object Request Broker
(ORB) specification of the Object Management Architecture (OMA). CORBA
aims at providing a uniform communication infrastructure for building
distributed applications. It provides the user with unifying mechanisms
capable of interoperating software components, operating on various
MIDDLEWARE 305
software platforms, and running under different operating systems

implemented in different programming languages [45]. In addition to all
of these, CORBA is capable of supporting heterogeneous, robust,
transparent real-time applications. Therefore, it is also quite suitable for
real-time substation communication and automation systems.
Client Server
object object
IDL IDL
stub skeleton
ORB ORB
Network
TCP/IP TCP/IP
FIGURE 4.12 Elements of CORBA
In addition to all these, CORBA services, add to the basic capabilities

of ORB. The most important of all is the Naming Service, which enables
CORBA objects to register themselves so that they can be located by name.
On the other hand, the security service describes how an ORB provides
secure communications and defines the different levels of security that
can be provided [46].
4.6.3 Mapping UCA Over CORBA
The use of CASM services within all UCA device models allows for the
expansion of the UCA protocol suite to other application protocols such as
CORBA reason being the fact that device models to be specified are
independent of the underlying protocol [44].
4.6.4 Basic UCA Models
Object Oriented Modelling (OOM) techniques are used to define the UCA
device models. UCA device models represent the behaviour of real devices
by defining standard classes and objects inherited and aggregated from the
basic class definitions. Objects accessible by clients often reside in remote
servers. CASM objects, which inherit from the CASM classes
LogicalDevice, DataObject and DataSet, usually reside within the Server
and they are directly accessible by the client through the network. CASM
objects can easily be translated into CORBA objects by giving them object
references. This is a straightforward process since both techniques use the
OOM approach as mentioned earlier.
4.6.5 CASM Communications

The communications mechanisms of the UCA also match with those of the
CORBA. There are four main communication mechanisms used in CASM.
They are [48]:
• Request/response
• Request/No response
• GOOSE message
• Sampled value.
For the first-two mechanisms, CORBA’s 12-step communication
mode, shown in Fig. 4.13, is used, which was implemented and described
earlier on by the previous chapter [49]. For more details, refer to the
references figure and other figures. For the remaining two mechanisms,
the developers are still in the process of designing a new Publish/Subscribe
CORBA Architecture, which will make it possible to transfer GOOSE state
change messages and sampled values over the network easily and with
high determinism.
4.6.6 Further Needs

UCA includes a definition for the relay-to-relay communication of binary
state data known as GOOSE, which has the packet format shown in
Fig. 4.14.
Client
1. Issue request 12. Return
2. Marshall arguments 11. Unmarshall results
3. Send request 10.Receive reply
TCP/IP
4. Receive request 9. Send reply
5. Unmarshall arguments 8. Marshall
6. Up-call 7. Return
Server
FIGURE 4.13 Twelve steps of CORBA

MIDDLEWARE 307
Source addrs Multicast

(48 bits) (1bit)
Relay name Server name
(32 bits) (32 bits)
Time of event Destination
(32 bits) (32 bits)
Next goose Req/res
(32 bits) (1 bit)
Dynamic network announcement

(64 bits)
Data
(12800 bits)
FIGURE 4.14 ‘GOOSE’ format

GOOSE messages are transferred over the network by a
communication mechanism called Publish-Subscribe architecture with
which, any IED can either [46] subscribe to data that it needs, or publish
information that it produces. Publish-subscribe architectures make the most
efficient use of the networking resources. However, real-time systems have
other needs, which cannot be met by the publish-subscribe architectures.
One of this is the ability to trade-off delivery reliability against delivery
delay [47].
The process of trading-off reliability of delivery for greater
determinism is crucial for multicasting GOOSE state change messages [47].
Sending the most recent GOOSE message is much more important than
resending old updates, which would probably be out-of-date when they
are delivered anyway. In contrast, when a sequence of device operation
commands is required then a communication processor has to receive every
step in the command sequence properly, which can only be quarantined
with reliable delivery. In some situations, protection applications require
some intermediate action, which is unfortunately not specified by any of
the current networking protocols. The developers are currently working
on a new publish-subscribe architecture in order to solve this problem that
exists within the substations. The simulation section further illustrates the
need for such an architecture.
4.6.7 Simulations
We will be running simulations on a point-to-point architecture to measure
the packet-end-to-end delay for two different configurations. The first
configuration uses TCP/IP as the transport protocol whereas the second
configuration uses UDP to transfer packets between the CORBA Client
object and the CORBA Server object over the network. The network model
used to simulate the point-to-point network architecture is shown in

Fig. 4.15.
APPL APPL
APPLICATION PROFILES
DEFINITION DEFINITION
Applicants Profiles
Atlantic Ocean
In
Corba_server IED1
FIGURE 4.15 Point-to-point network architecture
Fig. 4.16 shows the two components of the system reaction time for
the case when UDP is used as the transport protocol:
Communication Management
control interface
Communication Diagnostics
protocol stack and test
Real-
time
OS Application framework
Data transport Board support package

control
Software
Communication protocol Microprocessor peripheral

hardware hardware
FIGURE 4.16 Embedded communication device software architecture

MIDDLEWARE 309
• The end-to-end delay of a packet to travel from the IED to the

server, and
• The end-to-end delay of a control packet to travel from the server
back to the IED.
The transmission of trip signals including short but time-critical data
packets has a 4 ms-performance requirement, which is the most crucial
requirement. Under normal conditions such as this scenario, the delay going
from controller to IED is about 0.3 ms and the delay going in the opposite
path is about 0.35 ms. The sum of the delays in both directions gives us the
total end-to-end delay for a packet as 0.75 ms, which is well-below the
4 ms performance requirement.
However in the case of long and reliability-critical signals, TCP needs
to be used so that there will not be any loss of packets. In this case, 4 ms-
performance requirement cannot be met. We can clearly see from Fig. 4.15
that in the case of TCP transport protocol being used, the end-to-end packet
delay will be about 9 ms.
Substation Automation systems, responsible for operating
equipment in Electrical Substations, are prime examples of distributed
command and control systems. They were traditionally built as turnkey
systems based on proprietary technology but with the recent developments
in automation and networking, a new wave of interoperability
requirements has started. Standardization is a key issue offering
interoperability across vendor boundaries [57-60].
The IEC 61850 standard protocol, which will be a published standard
since year 2003, specifies a collection of elementary services such as breaker
operation, monitorization, reporting, etc. Different types of technologies
can be used for implementing these services, sometimes referred to as
logical nodes in the substation and then they can be distributed over a
network including various processors such as IEDS, RTUs and station
computers. Furthermore, such services can be bundled together to form
the global system behaviour. For example, a global synchronised switching
function can be implemented by integrating several logical nodes such as
synchronised switching, HMI and breaker control. The ACSI interface
designed by the IEC TC-57 group is used as the common abstract interface,
which supports interoperability. This abstract interface is mapped to
standardized communication protocols that provide the real substrate for
the implementation.
This section mainly deals with implementing the IEC 61850 over a
CORBA based middleware. This is not unique in the sense of implementing
the IEC 61850 standard over the CORBA middleware since some initial
work has already been carried out in this field since 1999. However, the
project can become unique in the sense that the IEC 61850 is not to be
implemented over a standard CORBA middleware but over an extended
version of the standard CORBA middleware. In other words, the main
phase of research comes in when the standard CORBA middleware is
extended to form an effective platform for performance sensitive real-time
power system operations. The main emphasis is on satisfying the
middleware requirements such as the ones detailed in Section 4.3.
Next step is to implement the IEC 61850 on the newly designed
middleware. Such an implementation is necessary since the designed
middleware will not deal with specific object functionality but with how
interfaces are specified, used and managed across the substation
automation network. Thus, we can provide an effective platform for
performance sensitive substation automation applications while also
meeting the interoperability requirements. Last step of the work can be
programming a programmable communications processor with the overall
code. The programmed communication processor can be referred to as
Universal Communications Processor (UCP) and FPGA technology could
well be used for this purpose [55-60].
CORBA enables running application analysis, monitorization and
simplifies system evolution. CORBA, nowadays, is the best suitable
platform for distributed systems construction due to its ability:
• To provide a good mixture of performance
• To provide resource consumption
• To provide a good support in the early phases of systems
engineering life cycles.
The research aims for making a standard middleware for substation
automation system by extending the standard CORBA middleware to meet
certain requirements as detailed earlier on. Thus, while keeping some of
the benefits it provides, the developers have tries to enhance it to end up
with a middleware platform speficifically designed to meet the
requirements of a substation automation network. Hence, the designed
communications processor is of universal type since it can include a
universal middle ware and the IEC 61850 universal standard.
4.7 COMMON ARCHITECTURES FOR COMMUNICATION

DEVICES
Many embedded communication devices have architectures with variations
on the same theme, containing a surprisingly similar set of key elements,
inspite of the variety of forms and diversity of functions. This section
examines those common architectural elements in some detail and touches
MIDDLEWARE 311
on the implementation of these elements. There are many developers of

embedded communication device architectures leading to many variations.
However, a common set of elements tends to appear in a great many
architectures. The basic software architecture of a hypothetical embedded
communication device is shown in Fig. 4.16, where usually only a couple
of elements in the architecture are more or less unique to each other while
the rest show very similar properties, form and fashion [58-60].
4.7.1 Communication Protocol Stack
The communication protocol stack is the heart of any communication
device. All types of communication are subject to some form of protocol.
A device can communicate with any other entity only if it knows at least
one communication protocol and if the targeted entity knows the same
protocol. Most standardized communication protocols roughly suit some
portion of the Open Systems Interconnect (OSI) model shown in Fig. 4.17.
Some portion of layers 2 and 3 are usually provided in software Transaction
and Control Protocol (TCP) of layer 4 is offering higher-level functionality.
Layer 7 Application
Layer 6 Presentation
Layer 5 Session
Layer 4 Transport
Layer 3 Network
Layer 2 Typical embedded

Data link
protocol layers
Layer 1 Physical
FIGURE 4.17 The OSI model
Physical layer is usually implemented in hardware due to prohibitive

processing and electrical signalling requirements. However in many cases,
we can see software and hardware interactions in layering. Some special
communication processors can perform physical layer functions in
software, while upper-layer functions can be performed in hardware
[59-60].
For example, it is common for a Digital Signal Processor (DSP) to
perform some physical layer functions in software. On the other hand, a
Network Processing Unit (NPU) may perform portions of all seven layers
in hardware.
The lower layers in a protocol stack are primarily used to establish

and maintain a physical, logical or virtual connection suitable for reliable
transmission of the information in the upper layers.
Whereas the primary function of the upper layers is to transport
information relevant to the application of the transporting entity. Various
communication-oriented activities can occur at any layer. These are mainly
activities such as flow control, error detection and security checking and
etc. An important implementation strategy of the protocol layers is to use
event-driven state machines since in many cases a protocol layer require
some event to occur in order to proceed to the next state in the protocol.
Fig. 4.18 shows an extremely simple layer two state machine [56-60].
Idle
Layer 3 Layer 1
Correct request Release confirm
Layer 1
Failure indicate
Resources
pending
Layer 1
Resource confirm Release
pending
Layer 1
Failure indicate
Link Layer 1
pending Release indicate
Layer 1 Layer 1
Link confirm Link Release request
established
Layer 3
Data request
FIGURE 4.18 A simple layer two state machine
The state machine, shown in Fig. 4.18 is used for:

• Establishing a connection
• Forwarding data
• Releasing the connection based on indicators and directives from
layers 1 and 3, respectively.
The state machine approach is seen to be an effective way of
implementing the protocol behaviour based on the sequential nature of
the control protocol shown in the Fig. 4.18.
4.7.2 Communication Control
The application or control functionality of the communication device is
contained in the control element. It may be
MIDDLEWARE 313
• An intelligent application, which uses the protocol stack to

communicate with a peer entity on another device.
• A bridge between the protocol stack and another protocol stack.
• An interface between the protocol stack and another application
or device.
The communication control element is effectively the top-most layer
of the protocol stack, a service provider to the control element, regardless
of its function. Similarly, event-driven state machine approach can be used
for implementing a communication control element.
4.7.3 Data Transport Control
Any communication device is responsible for transforming information to
and from peer devices. In some cases, the volume of information allows
the software to construct every bit of the information sent and to analyse
every bit of the information received. On the other hand, in most cases,
devices must transfer massive quantities of information in real-time.
General-purpose microprocessors are typically ill equipped to handle the
raw processing requirements associated with such high data rates. The
solution for accommodating the need for additional data bandwidth is to
perform most of the protocol functions in communication-specific
hardware, where maximum output can be attained. The use of
communication-specific processors, which are specially equipped to handle
the activities of a protocol or a family of protocols, is also one of the
solutions.
Therefore, the task of the communication devices software is to
control the data transport hardware logic, rather than performing the data
transport directly. This task usually includes setting up (programming)
the hardware to transport data, and having the hardware to notify the
software when something of interest occurs. The appearance of an
important piece of data such as a device address, the occurrence of an
error and the crossing of a threshold are among those that might be
considered as interesting information. Thus, this ‘set-up-and-notify’
approach enables the communication device to transport massive volumes
of data with minimal intervention from the software. The data transport is
typically initiated and controlled by the protocol stack as it walks through
its various states. It is quite common to implement the data transport control
as a hardware device driver since the data transport is generally hardware
assisted. The driver provides hardware access functions as well as hardware
interrupt service routines [40, 58-60]. Fig. 4.19 shows a general driver model,
where the access functions are used for programming and controlling the
data transport hardware, such as:
• Configuring the hardware to perform data transport in a specific

way
• Setting up interrupts
• Initiating the flow of data.
Protocol stack
Interrupt Interrupt
Access Access
service service
function function
routine routine
Hardware
FIGURE 4.19 A general driver model
The interrupts from the data transport hardware are processed by

the interrupt service routines, such as those indicating the receipt of
interesting data or the occurrence of error conditions or threshold crossings.
4.7.4 Management Application Interface

The ability to adjust system characteristics of a device in response to
changing conditions is often ignored in the world of embedded
communication devices.
There is a clear need for a management interface, which would allow
a device to be adjusted in response to changing system conditions. Some
of the typical management activities are:
• Configuration-setting device parameters that affect the device’s
run-time behaviour.
• Performance monitoring-collecting information concerning the
performance of the device, which may prompt reconfiguration.
• Usage monitoring-collecting information concerning the usage
of the device.
• Fault monitoring-collecting information concerning the faults,
errors, and warnings related to device operation.
• Diagnostic and test-prompting the device to perform self-
diagnostic and tests, and reporting the information to the
management source.
The management interface can be implemented as a group of
sub-elements as shown in Fig. 4.20.
MIDDLEWARE 315
Protocol Management protocol access

stack
Diag.
Configuration Performance and test
manager monitor manager
Usage Fault
monitor monitor
Diag./Test
Objects and/or Data stores application
FIGURE 4.20 Management interface
The sub-elements of the management interface are:

1. Management protocol agent: A management protocol has to be
executed so that the management interface communicates with the
external management system. The function of this sub-element is to
act as the protocol agent for the other management sub-elements
hiding the details of the management protocol from them. The event-
driven state machine approach can well be used when implementing
the management protocol agent.
2. Configuration manager: This sub-element maintains the
configuration data in a data store and/or passes the data to the
appropriate elements of the communication device.
3. Performance monitor, usage monitor and fault monitor: These sub-
elements access information from other elements of the
communication device reporting this information through the
management protocol agent to the external management system
[40, 57-60].
The elements can usually be implemented as functions or methods.
However, in cases of high complexity of their interactions with other
elements in the communication device, state machine behaviour may be
needed.
CHAPTER
5
Information Embedded
Power Systems
5.1 OVERVIEW OF COMMUNICATION NETWORKING

REQUIREMENTS
In every parts of the world, society has entered into a new era of economics
and social experience driven by digitally based technologies. Our world is
more interconnected than at any time in history, utterly dependent on the
integrity of complex networks including the internet, telecommunication
and electric power system. Existing information management systems
could not satisfy the new challenges as the demand for more and faster
information increase by many players either in developed or developing
countries. An equal step should be taken by developing countries in par
with the developed ones to upgrade the power system communications
infrastructure. The rapid developments of the internet and distributed
computing have opened the door for feasible and cost-effective solutions.
In this chapter, an overview of communication networking requirements
is discussed briefly. An emerging power system communication technology
to be used in the future electric utility industry called IEPS-LAN/WAN is
presented. The benefits of employing the IPES-LAN/WAN have also been
discussed.
In recent years, telecommunication systems have undergone radical
changes prompted mainly by the desire to increase system performance.
New technologies are providing huge increases in performance at lower
unit cost. At the same time, deregulation and privatisation of the electricity
industry coupled with the liberalization of the telecommunication market
have imposed new requirements on the electrical power communication
network. Key drivers for change of traditional SCADA networking to a
INFORMATION EMBEDDED POWER SYSTEMS 317
common communication networking (LAN/WAN) include: cheaper

availability of optical fiber installations, efficient transport control protocol
over internet protocol (TCP/IP) networking and emergence of quality of
service (QoS) over LAN/WAN technology [61, 62]. Traditional SCADA
networking is based on fixed voice grade circuits and modem thus making
sure that a communication path is there when required and that
transmission delay and variations are very small. This technology is
becoming obsolete and unsuited to the increasing demands of today’s
power system operations.
5.2 INFORMATION EMBEDDED POWER SYSTEM VIA

LAN/WAN
Information embedded power system is an extension of traditional power
systems with added monitoring, control and telecommunication
capabilities. A simplified illustration of an information embedded power
system is shown in Fig. 5.1. This system consists of: (i) power system
hardware, (ii) the measurement system (RTUs), (iii) the communication
system and (iv) the electric utility control centre. In this system, the RTU
computers record power system measurements and send them in real time
over a computer network (LAN/WAN) to the power control centre.
NAT NAT
REG REG REG REG
PROV PROV PROV PROV PROV
RTU RTU RTU RTU RTU RTU RTU RTU RTU RTU
RTU = Remote terminal unit, PROV = Provincial control centres

REG = Regional control centres, NAT = National control centres
FIGURE 5.1 Computer network controlling the electric network with a tree topology
Data communications have always played a great role in the

operation and control of power utility systems. Applications of data
communications in power systems range from relay-communications to

‘inter-control centre’ data sharing with direct link computer networks used
to deliver real-time measurements from RTU computers to an energy
control centre as well as other business data. Popular communication
network protocols employed in IEPS are direct link networks and end-to-
end network.
Detailed information on power system hardware, communication
system, the measurement system and the electric utility control centre can
be found in references [61, 63, 64, 65, 66].
5.2.1 Deregulated Power
Deregulation has served to complicate the operation of power systems. In
the new deregulated environment, the pattern of power flows in the
network is less predictable than it is in the vertically integrated systems,
in view of the new possibilities associated with open access and the
operation of the transmission network under energy market rules. The
goal of modern power utilities, in the presence of new competitive markets,
is to provide services to customers aiming at high reliability with the lowest
cost. Before the days of deregulation, utilities performed both power
network and marketing functions but were not motivated to use tools that
required accurate real-time network models such as optimal power flows
and available transfer capability determination. These practices are starting
to change in the emerging competitive environment. Worldwide, electric
utility deregulation is expanding and creating demands to integrate,
consolidate and disseminate information quickly and accurately between
and within utilities. Utilities spend an ever-increasing amount - estimated
$ 2 billion to $ 5 billion dollars a year in the USA only-for voice and data
communication. There are strong ways to find ways of reducing operating
costs to improve utility earnings. In the deregulated power industry, it is
necessary to have global vision of the network situation. That is, the
measurements acquired locally in the RTUs should be transmitted to a
provincial control centre. The information from these provisional control
centres is transmitted to a control centre of higher level such as regional in
which a more global vision of electric network can be obtained. In a similar
way, the information from the regional control centres can be transmitted
to a national control centres in which one obtains a general vision of the
network. This results in a hierarchy of control centres with several levels,
from the RTU until the general (national) network control centre. In
addition, information is frequently exchanged among control centres of
the same hierarchical level or different levels as shown in Fig. 5.1. The
increasing incorporation of digital devices throughout the utility enterprise

as well as the forces of deregulation are driving utility communication
into new realms with new requirements and paradigms. Deregulation place
new requirement on the communication of data and information
throughout the utility enterprise. With the current advancement in
Information Technology (IT), utility can meet the present data sharing in
broader perspective. The time has come to fully employ WAN technology
in the power system industry.
5.2.2 LAN/WAN
5.2.2.1 Migration to WAN
Conventional SCADA network designs rely on the predictable nature of
connection-oriented services using fixed audio bandwidth links, analogue
modems and specific protocols. Setting up and maintaining these networks
require specialised skills. Reconfigurations involve hardware rewiring, are
time consuming and costly. Bandwidth is limited to 3 kHz, which is
adequate for current RTUs but potentially limiting business move towards
the use of substation automation and remote management. As the world
moves to digital communications, the support of analogue modems is
becoming increasingly difficult [67-69, 72].
Changing to utilitizing WAN technology will enable the manage-
ment of SCADA networks to be integrated into a system common to the
corporate data network. Reconfigurations will be simplified to keyboard
commands rather than rewiring at multiple points. Bandwidth can
be allocated as required and RTUs themselves remotely managed. In
addition, the advantages of WAN networking include: worldwide adop-
tion, very well developed hardware and software market, simplicity and
choice of application layer protocols, inherent resilience of the IP routing
concept and strong network management, including remote control and
monitoring. Furthermore, WAN presents the opportunity to migrate to a
single network for both operational and non-operational requirements
[70-72]. Applications will include SCADA data, business data and video
monitoring, which are integrated with Network Integrated System (NIS),
Energy Management System (EMS) and Human machine Interface (HMI)
as shown in Fig. 5.2.
9600 baud RTU

PBX management &
EMS provisioning
GW
Host
NIS Site
Video
Monitoring
Wide Area
Network
VOIP Video VOIP Video VOIP Video
NIS NIS NIS
HMI HMI HMI

RTUs RTUs RTUs
Alarms Alarms Alarms
FIGURE 5.2 Integrated WAN communication network
5.2.2.2 Information Embedded Power Systems via WAN (IEPS-W)

Due to significant changes in the power system industry, Information
Embedded Power System via Wide Area Network (IEPS-W) is essential
to accurately and effectively monitor, control and use telecommunication
facilities in a broader perspective. Therefore, a wider flexible system is
desired so that RTU computers can record power system measurements
and send them in real time over Wide Area Network (WAN) to the power
control centre efficiently together with other non-SCADA data.
An IEPS-W is an extension of traditional power systems with added
monitoring, control and telecommunication capabilities. A simplified
illustration of an IEPS-W is shown in Fig. 5.3. This system consists of:
(i) power system hardware; (ii) the measurement system (represented by
three remote terminal computers – RTUs); (iii) the communication system
(WNA) and (iv) the electric utility control center. In this system, the RTU
computers record power system measurements and send them in real time
over a Wide Area Network (WAN) to the power control centre. Control
centres are also capable of sending messages back to the RTUs to perform
control actions such as opening/closing breakers, transformer tap changing,
generation control, etc.
Power utility
Power system control centre
RTU 1 RTU 2
I2
V1
Wide area
Measurement network Display
data
P3
Mainframe
RTU 3
FIGURE 5.3 Illustration of information embedded power system
This IEPS-W model should be based on high speed monitoring of

measurement points, ‘concentration’ of these measurement, and generation
of displays based on these measurements. By constantly monitoring
conditions throughout a wide-area-network, IEPS-W can detect abnormal
system conditions as they arise. Expansion of this capability is crucial for
implementation of an integrated wholesale power market. Reliable, real-
time gathering of a range of power system parameters will enable power
delivery system operators to detect and counteract abnormality over a wide
geographic area, thus enabling the power delivery system to operate safely
to its inherent limits. Broader implementation of IEPS-W like system will
provide the real-time information needed for integrated control of a large
highly interconnected transmission networks. Obviously, this will add
intelligent components to conventional controls to learn and make decisions
quickly, process imprecise information, provide high level of adaptation.
One of the challenges will be to meet human resources needed (technical
as well business skill). It will require well-trained personal, highly skilled
and well-informed people [62, 67-69].
5.3 THE BENEFITS OF USING IEPS-LAN/WAN TECHNOLOGY

IEPS-W looks very promising as shown in Fig. 5.2 integrating all required
parameters for the modern power system communication. However, an
extensive research is required to study how the random characteristics of
the computer network can affect the accuracy of the measurements sent
from the RTUs to the control centre especially when data are transmitted
via WAN environment when other data will be sharing the same network
as well. Large amount of computer network traffic may result in large

measurement errors and temporarily render parts of the power system
unobservable.
To maintain power system observability, the computer network must
meet two main criteria: bounded time delay and guaranteed transmission;
that is, a measurement should be transmitted successfully within a bounded
time delay (4 ms). Unsuccessfully transmitted or large time-delay messages
from the RTUs to the control centre may cause several buses in the power
system to become unobservable.
In addition, it is crucial to study the random characteristics of the
computer network as the traditional power system observability methods
does not consider measurement errors due to delays in delivering the
measurement. In other words, traditional power system monitoring
methods assume that the state of the power system remains unchanged
during the time it takes to deliver a newly recorded set of measurements
to a control centre.
In this new proposed model, delay could be higher when data are
transmitted via WAN and further research is required to minimize the
system delay since data are very sensitive for utility industry. The
characteristics of the measurement delays associated with these types of
networks will be much more complex due to the added complexities of
routing and switching. Before employing this sophisticated and realistic
IEPS-W model, utility must ensure through experimental analysis that this
model meets the communication delay acceptable by electrical power
industry [71-72].
CHAPTER
6
Fiber Optic Network
Infrastructure as Next Generation
Power System Communications
6.1 BACKGROUND
Power system communication has from high-speed substation control and
protection data communication to wide area power system monitoring
and measurement data transmission, the increasing incorporation of
computer network throughout the utility as well as the forces of
deregulation are compelling power system communications into new
realms with new requirements and challenges. Expanding network services
such as real time wide area control and Flexible AC Transmission System
(FACTS) device coordination are also driving the need for evermore
bandwidth in the network backbone.
These need will grow further as new real-time service, protection
and control applications become more feasible and pervasive. Electric
utilities often employ several types of communication media for different
functions. With more and more bandwidth required by the power system
data communication, the current transmission media cannot meet all the
high capacity and quality of service requirement. This chapter reviews
current power system communication media and discusses fiber optic
network infrastructure for the next generation power system
communications.
The power system is using several media for its protection, control,
and information sharing function. The most common ones include: Power
Line Carrier (PLC), microwave, pilot wire and wireless.
PLC operates by transmitting radio band of frequency signals
between 10 kHz to 490 kHz over the transmission lines. PLC with power
output of order 150 W can be used up to 240 km. Normally, PLC carriers
only one channel of 4 kHz bandwidth. The frequency range is limited by
government regulations. The PLC is the most common communication
media used in the USA. However, it has some disadvantage such as
bandwidth limit. It is subject to lightening switching surges, and networks
reconfiguration.
Another media is microwave. Microwave operates in the 150 MHz
to 20 GHz frequency range. This bandwidth can carry a lot of
communication channels for a variety of information. The disadvantages
of microwave is that the transmission length is limited to sight path between
antennas. Microwave is subject to atmospheric attenuation and distortion.
The combined latency using modem plus analog microwave is around 100
milliseconds between two adjacent antennas.
Pilot wire is normally a telephone wire either owned by utility
companies or leased from telephone companies. This type of
communication has a bandwidth from 0-4 kHz. Overhead lines may
experience interference from power lines while the underground is subject
to damages for many obvious reasons.
Wireless is one of the modern methods of communication. Low orbit
satellite communication system provides an existing option to transmit
information covering a very large range. The delay is a problem, which
depends on the distance. For example, the latency for low orbit satellite at
10 km above the earth is about 300 ms one-way. Another disadvantage is
the cost of installation.
All the above media may be using different communication networks
such as circuit-switched networks, packet-switched networks, and cell-
switched networks [73–80].
6.2 CURRENT POWER SYSTEM DATA COMMUNICATION

MEDIA
Real time monitoring is an expanding network service that drives the need
in the network backbone for ever wider bandwidth. As new remote real
time protection and control applications become more flexible and
pervasive these needs go further for increased bandwith. Conventionally
electric utilities made avail of several types of communication media for
different functions. [81–86]. Extensively used communication media in the
power system are highlighted in following sections. Fibre Optic will be
the ideal choice for future communication infrastructure, with more and
more bandwidth required.
FIBER OPTIC NETWORK INFRASTRUCTURE AS NEXT GENERATION...... 325
6.2.1 Power Line Carrier

Power Line Carrier (PLC) [81] [85] [100] does not offer a reliable solution
for wide area data transmission. Communication with remote sites cannot
be maintained during a disturbance. Therefore, PLC’s effectiveness for wide
area data transmission is limited.
6.2.2 Dedicated Links

In order to communicate between Control centre and substation RTUs;
dedicated links [81] [85] [86] [87] are employed by many SCADA systems.
However, capability to provide high data rates is the foremost advantage
of dedicated links. The main disadvantage in dedicated links occurs in
remote areas, due to lack of connectivity. Dedicated links have no practical
application for controlling medium voltage grids.
6.2.3 Radio Systems
Wide area data transmission can facilitate Conventional Radio (CR),
Trunked Radio (TR) or Spread Spectrum Radio (SSR). CR, TR or SSR are
based on licensed channels or over non-licensed frequencies. Many
countries have limited frequencies in the VHF/UHF bands (Very High
Frequency/Ultra High Frequency) bands. Overutilisation of unlicensed
frequencies by mass consumer applications causes questions in the
reliability of VHF/UHF for commercial and industrial use. Poor utilization
of air time and unreliable communication resulted when using line
protocols over radio.
6.2.4 Microwave
The radio signal operating in the 150 MHz to 20 GHz frequency range is
Microwave [81] [85] [87] [93]. The main disadvantage of the microwave is
that the transmission length is limited to a line of sight path between
antennas, is subject to distortion and atmospheric attenuation. Using
modem plus analogue microwave, the combined latency is around 100
millisecond between two adjacent antennas.
From Table 6.1, it is evident that none of the above currently used
transmission media in the power system, can meet the real time
measurement requirement. Time latency requirements by fast control and
protection function cannot be met by transmission media in Table 6.1.
Therefore, the above media in Table 6.1 is not suitable to meet ‘Quality of
Service’ requirements.
Table 6.1 Currently used transmission media in the power system
Transmission Media Data Rate

T1 1 Mbps. Effective bandwidth considering
network traffic, data collision etc 125 kbps
Frame Relay 280 kbps
ISDN 140 kbps
T1 Fractional 62.5 kbps
56k leased line 56 kbps (Effective bandwidth lower than this)
Internet Effective rate 40 kbps depends on network traffic
Radio Frequency 9.6 kbps
Power Line Carrier 1.2 kbps
6.3 NETWORKS AND INFORMATION TECHNOLOGY

The problems mentioned in the previous sections pose the challenge of
making use of more advanced technology for future power system
communications infrastructure design.
6.3.1 Fiber Optic and its Enabling Technologies
Fiber optic system [81] [131] is the most suitable data transmission medium
for power system control, protection and monitoring function. The
particular characteristics [81] [132] [133] of the optical fibers that make
them so useful are: low attenuation; high bandwidth; electromagnetic
interface immunity and security.
Low attenuation: The wavelength of the light signal in use has a
direct effect in the attenuation of the Optical Fiber. Attenuation of 0.35 and
0.2 dB/km are achieved by Wavelength of 1300 nm and 1550 nm. 0.2 dB/km
attenuation value allows 100 km between repeaters or amplifiers; given
that the achievable net loss is 20 dB.
High bandwidth: Potential bandwidth of 20Tbps which could
accommodate 312 million voice channels (64 kbps) can be facilitated by a
single optic fibre operating at 1300 nm or 1500 nm of wavelength. Electro
Optic devices and receivers or the Electronic interface to these devices,
mostly determine, the bandwidth limits of fiber Optic transmission.
Small physical cross-section: Small physical cross-section, is an
advantage provided by Fiber Optic Systems. By merely installing
Wavelength Division Multiplexing (WDM) or higher Speed Electronics the
Optic Fiber can be upgraded for more capability.
Electromagnetic interface immunity: High powered transmitter

or very sensitive receivers are unable to cause interference in the form of
noise or cross talk on optic fibers, making them immune to Electro Magnetic
Interface Interference.
Security: It is impossible to non-invasively eavesdrop of the signal,
as optic fibers radiate no energy. The advancement in key component
technologies [81, 131–136] such as fiber, amplifiers, lasers, filters and
switching devices has driven to the rise of optical networking such as WDM.
Switching components and Optical linking components are the two
classifications of Fiber Optical linking components. WDM Multiplexers/
DeMultiplexers and WDM passive star couplers form the make up of optical
linking components. OADMs/OXCs and tunable transmitter/receivers are
the make up of switching components.
The following features are seen widely used optic fibers:
1. OADMs: Programmable devices configured to add or drop
different wavelengths.
2. Amplifier technology: Act as wavelength routers or wavelength
cross connects.
3. Transmitters: Which send the optical signal to fibers. Fibers are
two types—Tunable or fixed.
4. Amplifier technology: Advances in amplifier technology have
increased the distance between signal generators.
5. Amplifiers: Two basic amplifiers have been proposed.
Semiconductor Optical Amplifier (SOA) can be integrated with
other silicon components for improved packaging. Erbium
Doped Fiber Amplifier (EDFA) design can typically achieve high
gain.
6. Optical packet switches: These are nodes that have optical
buffering capability and perform the packet header processing
function required of packet switches. [81, 91–103].
6.3.2 IP Over Optic Network Topology
WDM employments are point-to-point at present. In order to interface to
higher layer of protocol stacks, WDM uses SDH/SONET as the standard
layer. Different protocol stacks provide different functionality in terms of
bandwidth overhead, rate scalability, traffic management and Q.O.S.
(quality of service).
Segmentation and reassembly of data with class of service and setting
up of connections from source to destination is the function of the ATM
layer. The function of SONET/SDH layer are interfaces with the electrical
to optical layer, delivers highly reliable ring based topologies, performing
mapping of Time Division Multiplexing (TDM) slots from digital hierarchial

levels. The Wave Division Multiplexing layer multiplexes electrical signals
onto specific wavelengths in a point-to-point topology and constructs the
power system communication backbone. [81, 104–115].
The multi-layer topology as shown in Fig. 6.1 forms the basis of
current optic fiber networks. Time delay and function overlay are problems
with multi-layer topology. IP over WDM will be the better combination
[81, 116–135] based on results of studies conducted, because of the studies
conducted.
IP/ATM/SONET
layering
Packet Over SONET
IP/MPLS layering
IP Over WDM
ATM IP/MPLS layering
SONET/SDH SONET/SDH IP/MPLS
Optical WDM
FIGURE 6.1 IP over optic network technology
Each wavelength can be considered as a dedicated connection, as IP

over WDM architecture brings in the property of virtual fibers. For
performing control operations, the signals need not to be converted onto
an electrical domain. The Latency in the IP/WDM system is smaller
compared to that encountered in the SONET system. By using IP over
WDM, the transport capabilities of SONET/SDH are being absorbed by
the optical layer and the ATM function of traffic engineering is being
absorbed into IP. Therefore, the multiplayer architecture converges to two
layers.
6.3.3 The Need of Information Management Methodology
IT will play an increasingly important role to cater for the great needs of
information exchange in power system. By the redesign of the system for
information exchange, integration, utilities have progressed from their
initial efforts to re-consider their information needs. A methodology for
efficient information consolation, exchange and sharing is required. As the
volume of information holding increases, this need will become increasingly
more critical [81, 137]. In different application such as information accessing
and sharing [81, 138, 139] IT technology has been employed in power
systems.
Issues such as multiple data formats, compatibility problem and lack
of metadata standardization complicate the exchange of data among
different users. Although many standards have been developed already
for information management, they are often incompatible with one another,
as they tend to overlap. Isolation between power system applications also

hinders the development of a universal information infrastructure in power
system utilities. A great waste of resources has resulted because of the lack
of standards among the disparate systems. The development of a universal
information infrastructure in power system utilities has hindered the
isolation between power system applications. Future power industry may
benefit from the use of universal information architectures, such as a
standard data exchange model and communication network that can
support the different data requirements, transfer rates and qualities of data
flow among various systems.
To address the critical need of a universal data interchange problem,
the more recent XML technology can be used. For easy information
exchange between disparate power systems and application, XML
technology is used. Information exchange can be accomplished with
minimum modifications necessary on existing applications, using XML.
While still being able to exchange power system data with other
applications, utilities can continue to use their respective proprietary data
formats internally XML [142] has been widely used in networking [143],
earth science [145], e-commerce [144], simulation [146] for data exchange.
XML has also been introduced and discussed in power systems for power
market and information exchange [149] with the development of power
industry deregulation [81].
6.4 WIDE AREA COMMUNICATION INFRASTRUCTURE

IP over WDM network design for future power system is proposed through
the above discussion. One of the key advantages of WDM is that it offers
support for multi-protocols, allowing multiple independent network
protocols to coexist on the same filter network. It is extremely important
to cooperate with the existing multi-protocol network in power systems.
Increasing operational costs are required to deploy fiber rings with
physical topologies. There is a great desire to deploy WDM technology
and further optimize this with wavelength routing, as shown in Fig. 6.2.
The overall fiber optic network will be mesh architecture [81] [147] [149].
The key nodes will be the big utilities or control centres. Together they set
up the core optical network. Data transmission will set up on this network
over IP protocol.
FIGURE 6.2 IP over WDM fiber optic network
Fiber optic technologies have developed rapidly over the last 15-20
years. All-Optical cross connects and All-Optical Add-drop multiplexers
enable the evolution from simple point-to-point WDM links to full
networks. Fig. 6.3 shows detailed fiber optic network architecture.
Consisting of N-full duplex ports, where each of which can connect to any
other device, an OXC is a large photonic switch. An OADM is a 2 × 2
degenerate form of the NN × OXC. An OADM extracts and reinserts certain
light paths for local use and routes the others through.
End user
OXC
OXC Router
OXC OXC End user
Long distance OEO
OXC OXC
mesh network
(100s–100s km) OXC Router End user
Metro distance
network OXC
(10s–100s km) OXC Router
OXC OXC OADM

OADM GigE
Local area network GigE
(1s–10s km)
GigE
Local area network Router
End user
FIGURE 6.3 Detailed fiber optic network architecture
Physical Layer
The layer at which signals are exchanged is known as the physical layer.
Based on fiber optic along the distribution feeders or single mode fiber
installed in the substation, is the feature of the physical layer. While bit
rates can be OC-48 (2.5 Gbs)/OC–192 (10 Gbps) or higher, transmitting
options are based on laser.
Data Link Layer

Responsible for delimiting data fields, the link layer, above the network
layer, makes acknowledgement of receipt of data and error control and
the capabilities. Receipt of information that passes the error check is
acknowledged to the sending station, in most of the communication
systems. When two devices of different speeds try to communicate; the
data link layer, may communicate; the data containing a flow control
mechanism. By using flow control mechanism, message first time
transmitted, can be ensured to all traffic in fiber optic network. In
combination with the antibody algorithm, fiber optic network can help
guarantee minimum network congestion.
Application Layer
The application layer will be allocated to power system communication.
For information exchange between utilities and substations, the wide area
fiber optic network can be used. Information such as rate schedules,
operating constraints, available transmission capacity can be shared
between different users of power system.
6.5 LOCAL AREA SUBSTATION NETWORK DESIGN

Due to implementation of advanced computer and network technologies,
at present, substation automation tends to be more complicated. For the
proper protection and control operations, real time information acquired
by IED must be transmitted under few milliseconds. Real time, high speed
communication links between station unit is in demand. A large number
of power equipment, supported by an efficient communication system, is
required for this type of substation system. The optic fiber choice as
transmission medium, guarantee the robustness against EMI (Electrical
Magnetic Interference).
6.5.1 Substation Communication Network Requirement
The automation system has the following needs:
• Data acquisition: Analog and digital information are included in
the data, from the equipment in the substation. With the
provisioning of consolidated metering, alarm and status
information, local operations are facilitated.
• Control and monitoring: A monitoring system is needed for
substation level control. Between the control centre, local
substation and IEDs a control hierarchy can be implemented.
• High availability and redundancy: The “no single point of
failure” communication criterion, system, must be guaranteed.
• Capability for future expansion: Easy future communication

expansion, should be allowed by the system.
6.5.2 Substation Communication Network Architecture

As shown in Fig. 6.4, the proposed all-fiber substation communication
network architecture can be divided into three levels. IEDs or PLCs
(Programmable Logic Controllers) connected to substation equipment is
incorporated in level 1. These PLCs or IEDs are classified either as control
or protection IEDs. The purpose of protection IEDs is to report the
equipment status information and implement the protection algorithms.
The purpose of control IEDs are to act as gateway between substation server
and protection IED. Substation server makes up level 2. All control and
monitoring operations can be performed anytime. Communication between
substation server and control centre server takes place, for information
exchange. The utility control server makes up level 3. The entire substation
system is monitored and controlled by the utility control server. Various
LAN topologies such as FDDI (Fiber Distributed Data Interface) or Giga
Byte Ethernet, can be used for substation automation system design.
Scada server Level 3

Scada server Scada server
Control centre
Switch
Substation
server Level 2
Substation
Switch
Level 1
Control IED Control IED Substation
Control IED
Protection IED Protection IED Protection IED
FIGURE 6.4 Substation communication network
6.6 TIME DATA COMMUNICATION AND EXCHANGE

Applications based on the suite of transmission control protocol/user
diagram protocol/ real time protocol/internet protocol/HTTP (TCP/UDP/
RTP/IP/HTTP) protocols [81] [148] [150] [151], or commonly referred to
as IP protocol form the largest segment of traffic. IP protocol is clearly the

convergence layer in today’s data communication network. In the new
future, it is no doubt, that IP protocols, will expand its service to multi-
service networks. Transportability over a broad variety of data link layer
protocols and underlying network frequencies, is a features of IP based
data. For real time information transmission through the network an IP
based protocol such as TCP/IP can be used.
6.6.1 ISO-OSI Network Architecture
The four-layered internet architecture is commonly used instead of the
generic seven layered model and has been illustrated in Fig. 6.5. [81] [148]
[150]
• Network layer: Raw data transmission is performed in this layer.
This layer is implemented by a combination of hardware (NIC-
Network Interface Card) and software (NIC driver). FDDI
Ethernet are commonly used networks defined in this layer.
• IP layer: Internet Protocol fits within the content of the IP layer.
IP supports multiple interconnected networks into one logical
network.
• TCP/UDP: In-order to provide logical channels to application
programs TCP and UDP are utilized. Both IP/TCP are most
commonly used IP based protocols.
• Application layer: User defined applications are supported by
application layer.
User User
SMTP/FTP... application TFTP/NFS... application
TCP UDP
Internet protocol
Network layer
FIGURE 6.5 Internet architecture
6.6.2 IP, TCP and UDP

Operating as a network layer protocol, Internet Protocol (IP) [81] [148] [150]
[151] is responsible for routing addressing and package delivery. Internet
Protocol (IP) does not handle assured delivery, package division,
sequencing or error correction.
UDP [146] [148] [149] does not add much service to the underlying
IP. A mechanism of sending packets called datagrams is simply provided
by IP. The arrival of data neither the order of arrival nor important are
guaranteed. For large volume and non-critical data transmission like stream
audio and video UDP is normally used.
The TCP [146] [148] [149] offers connection-oriented byte-stream
service to the data transmission. The features are very important for power
system data transmission, since when data sent to the control centre or a
command is issued through the network, each bit will be very critical for
the correct information interpretation.
Reliability facilities to the IP protocol such as error detection and
correction, flow control, resequencing and duplicated segments
management, are added by TCP [146] [148] [149].
CHAPTER
Conclusions
CONCLUSION TO CHAPTER THREE ON PROTOCOLS

Substation communications and ongoing communication standardization
are addressed in this chapter. Substation Integration and Automation are
becoming the tools that can help utility to achieve reduced installation,
maintenance and operation costs. This is possible because of the integration
of microprocessor based IEDs/relays into Substation or even Power System
Integration Systems.
In this chapter, the IEC61850 and the Utility Communications
Architecture (UCA) protocols have been discussed. By providing
innovative, simple to use, robust technologies for power system protection,
automation, control and monitoring; power providers are focussed on
increasing productivity and making electric power safer, more reliable and
more economical.
The transmission of data along the utility powerline network is called
Powerline Communication. Powerline Communication eliminates the
need to rewire houses and buildings with separate communication links.
The fact that the physical network is already installed over a wide area is
the main advantage of powerline communications.
In substation environments the following main protocols have been
found widespread. The MODBUS, Distributed Network Protocol (DNP),
IEC-870-5-101 and UCA in particular have extensive use.
The DNP-2 protocol is implemented in 0.8 µm CMOS process, with
a target of 50 MHz clock rate. This chip features a power dissipation of 2
Watts at 50 MHz and has 299 pins. The silicon size measures 11.5 × 11.5 mm2
and integrates 60,000 gates excluding on chip memory of DNP-2. The DNP
has a peak performance of around 50 MCPS.
Its use of object models of devices and device components is the
main difference in UCA from the previously designed and used protocols.
The common data formats; identifiers, controls for substation and feeder
devices can be defined with the use of object models of devices and device
components. The models specify standardized behaviour for most common
device functions and allow for significant vendor specialization for future
innovation.
The IEC 61850 protocol identifies all the known functions in a
substation automation system and splits them into sub-functions or
so-called logical nodes. A logical node is a sub-function located in a physical
node, which exchanges data with other separate logical entities. In IEC
61850, all logical nodes have been grouped according to their most common
applications area, a short textual description of the functionally, a device
function number if applicable and the relationship between logical nodes
and functions. The introduction of UCA and IEC61850 has made it
justifiable and possible to integrate station IEDs through standardization.
Utility Engineers can eliminate, many expensive stand-alone devices and
use the sophisticated functionality and the available data to their full extent,
using the standardized high-speed communications between IEDs.
OVERALL CONCLUSION TO IEC61850 AND DNP-3

Recent developments with communication media and protocols,
networking technology, computing devices and substation equipment have
presented major new opportunities for utilities to improve their electric
systems, operations, and business process automations.
One area of significant development is with applications that can
involve communications to substations such as:
• SCADA
• Interfacing to a power line carrier AMR and/or load management
system
• Video (for physical security or equipment observation)
• Mobile data hotspots
• Voice over IP
• Relay, control, and meter configuration-related functions
• Retrieval of disturbance information
• Communication network management
• Physical access security such as card readers
CONCLUSIONS 337
• Communications link to down-line devices for distribution

automation and,
• Remote access from the substation to central application servers.
Each of these applications impact the communications network,
cyber security needs, plans and designs for substation equipment, and
head-end software applications. There are numerous details that go into
navigating this new technology “landscape”. The focus of this presentation
is to look at the communication protocol situation for substations given
this wide range of applications.
It was not that long ago when the applications using communications
at the substation might have been limited to SCADA, AMR data backhaul,
and a dial-up phone for voice communication or possible ad hoc data
retrieval from a meter or set of relays. In terms of protocol choice, the major
question might have been what to user for SCADA.
Today, a substation can have numerous communication protocols
supporting a wide variety of applications. This situation developed partly
as a result to the board acceptance and support of IP-based communications
to substations. This support comes in the form of numerous substation
devices supporting TCP/IP communications and perhaps hundreds of
different wireless communication products. In addition, through small on
a percentage basis, there are a growing number of substations with access
to fiber optics for communications.
Communication messages for an application typically involve
multiple layered protocols such as a file transfer application using FTP,
connection application using TCP/IP, or a connection less application using
UDP/IP, all over Ethernet. Modern communication networks to substations
will often have multiple application layer protocols running ‘‘on top of’’
TCP/IP and the lower-level data link and physical layer protocols.
Some examples of water communications to a substation could
involve today include:
• TCP, UDP (user datagram protocol), IP, Ethernet
• DNP3 and DNP3 over IP
• IEC 61850-compliant standards
• XML
• SNMP
• FTP
• COMTRADE file formats (with protocol having file transport
capability)
• HTTP
• Vendor-specific protocols to run their intelligent electronic device

(IED) software packages remotely
• Radio diagnostic protocols
• Video protocols
• VoIP
• Security card reader protocols and,
• Mobile application protocols.
For cyber security purposes, there may be more than one IP network
or the set of applications supported may be limited to the most critical
substations. The bottom line is that today’s technology allows utilities to
run multiple protocols. Two important communication standards that have
unique positions in the industry are IEC 61850 and DNP3.
Some utilities have already jumped into implementing IEC 61850.
Give the continual expectation and move to improve information
technology to achieve better results, it seems reasonable to assume that, in
time, most all utilities will move to an object-oriented, standardized
approach to substation automation and other application area that can
benefit from the same technology. The different question is probably more
when it will make sense for a particular utility’s situation give the state of
their existing systems, their objective and the state of technology that might
meet the needs.
While IEC 61850 has gained some momentum with completion of
the standard and the availability of more vendor products, other technology
and standard as well as IEC 61850 will continue to evolve, much like DNP3’s
move to incorporate some of the IEC 61850 self-description capability. There
is still no silver bullet answer or solution, but there are great new
opportunities.
When moving to take advantage of opportunities, it will continue
to be important for any utility to understand the quality and
interoperability of the technology they are looking at regardless of the
standards it adheres to. Due diligence with requirements definition, design
and implementation will continue to be essential for a successful project.
CONCLUSION TO CHAPTER FOUR ON MIDDLEWARE

Open standardized communication interfaces are needed as deregulation
of the energy market has increased the importance of data, in a scenario
where a lot more data has been necessary.
Since the early 90s many attempts have been made to define the
standardized communication protocol and the need to integrate protection,
control and data acquisition on the substation Local Area Network (LAN).
CONCLUSIONS 339
The three basic sets of requirements when choosing a communication

protocol are performance, interoperability and maturity.
In order to address the challenge of interoperability, a new class of
software middleware has risen. Provided by the communications processor,
middleware software, is a layer between the networking code and the
application code. The function of the middleware is to insulate the
application programmer from the raw networking code thus providing an
easier way to communicate.
In this chapter, middleware requirements for remote monitoring,
protection and control applications have been discussed.
For achieving success in substation automation, the use of the
appropriate middleware is crucial. The three main types of middleware
architectures, namely, Point-to-Point Architectures, Client-Server
Architectures and Publish-Subscribe Architectures have each played a key
role in substation automation.
Distributed-real-time communications in the substation environment
can be realized using the publish-subscribe architectures.
The process of trading-off reliability of delivery for greater
determinism is crucial for multicasting GOOSE state change message.
Sending the most recent GOOSE message is much more important than
resending old updates, which would probably be out-of-date when they
are delivered anyway. When communication protocols are designed for
guaranteed delivery but less-than-reliable medium, than an important
problem arises in the case of failed transmissions. What happens is that
the communication protocol would be stuck trying to transmit the failed
tranmissions wasting time and destroying the timing determinism. In this
case, the best policy would be to send the latest update disregarding the
earlier updates.
In electric Control and Protection applications, there is a need for a
synchronisation time of 10ms. There is such a stringent requirement for
time synchronization across the distributed system, particularly due to
the phase differences among generators.
In any remote monitoring, protection and control systems, there is a
clear mechanism of one type for the notification of interesting events, when
remote sites notify the monitoring sites.
Ideally, important events (such as power system faults) need to be
pre-defined so that the system can listen to specific types of events to
produce asynchronous, persistant and multicast event notification.
When combined with publish-subscribe architectures, data push
architectures, enables users to distribute data to a large and variable set of
remote applications.
There is clearly a need for new middleware functions in order to

enforce the security requirements of server authentication, client
authentication, confidentiality and non-repudation.
Only if it is capable of managing priorities, controlling memory usage
and restricted access to system resources can the effective running on the
communication processor and the network software be effective on the
communications processor.
The needed communication model time stamps each transaction,
allows the application software to trade-off timing against reliable delivery,
controls and specifies memory usage, allows for adaptive synchronization
scheme, event notification and the functions required to meet with the
security measures and works in a real-time communications processor
environment.
The three types of subscription mechanisms that consumers can
generally make use of when subscribing to information are channel-based
subscription, subject-based subscriptions and context-based subscriptions.
There are three messaging concepts which are Unicast messaging,
Multicast messaging and Broadcast messaging.
CORBA aims at providing uniform communication infrastructure
for building distributed applications. CORBA enables running application
analysis, monitorization and simplifies system evolution. CORBA,
nowadays, is the best suitable platform for distributed systems construction
due to its ability to provide good mixture of performance, resource
consumption and good support in the early phases of system engineering
life cycles.
CONCLUSION TO CHAPTER FIVE ON INFORMATION

EMBEDDED POWER SYSTEM
Information embedded power system is an extension of traditional power
systems. When a traditional power system has added monitoring, control
and telecommunication capabilities, then it is known as an information
embedded power system.
The system consists of power system hardware, the measurement
system (RTUs), the communication system and the electric utility control
centre. In this system, the RTU computers record power system
measurements and send them in real time over a computer network (LAN/
WAN) to the power control centre.
Demands to integrate, consolidate and disseminate information
quickly and accurately between and within utilities is expanding and
creating electric utility deregulation. It is necessary to have a global vision
CONCLUSIONS 341
of the network situation in the deregulated power industry. That means,

the measures acquired locally in the RTUs should be transmitted to a
provincial control centre. The information from these provisional control
centres is transmitted to a control centre of higher level such as regional, in
which a more global vision of electric network can be obtained.
World-wide adoption and very well developed hardware are some
of the main advantages of WAN networking. The simplicity and choice of
application layer protocols, inherent resilience of the IP routing concept
and strong network management, including remote control and
management have been very well developed. For both operational and
non-operational requirements, WAN represents the opportunity to migrate
to a single network.
Reliable, real time gathering of a range of power system parameters
will enable power delivery system operators to detect and counteract
abnormal over a wide georgraphical area, thus enabling the power delivery
system to operate safely to its inherent limits.
Real-time information needed for integrated control of a large highly
interconnected transmission network, will be provided, by a broader
implementation of IEPS-W like system.
Integrating all required parameters for the modern power system
communication, looks very promising and is known as IEPS-W.
Temporarily rendering parts of the power system unobservable, large
amounts of computer network traffic may result in large measurement
errors.
Bounded time delay and guaranteed transmission are two main
criteria, that must be met by computer networks.
Several buses in the power system will become unobservable,
because of unsuccessfully transmitted large time-delay messages from the
RTUs to control centre.
Traditional power system monitoring methods assume that the state
of the power system remains unchanged during the time it takes to deliver
a newly recorded set of measurements to a control centre.
Before employing sophisticated and realistic IEPS-W model, utility
must ensure through experimental analysis that this model meets the
communication delay acceptable be electrical power industry.
CONCLUSION TO CHAPTER SIX ON FIBER OPTIC NETWORK

INFRASTRUCTURE AS NEXT GENERATION POWER SYSTEM
COMMUNICATIONS
In this chapter, Fiber Optic Network Infrastructure has been discussed.
Fiber Optics is the next generation power system communication medium.
For protection, control and information sharing functions Power Line

Carrier (PLC), microwave, pilot wire and wireless are some of the different
media that have been used.
The most common media in the U.S.A. is PLC. Limitations in
bandwidth is a disadvantage of PLC. PLC is subject to lightening, switching
surges and network reconfiguration.
Microwave has disadvantages such as transmission length being
limited to sight path between antennas, atmospheric attenuation and
distortion. Pilot wire overhead lines may experience interference from
power lines while the underground is subject to damages. Wireless
communication has to deal with delays in propotion to distance and the
cost of installation.
With more and more bandwidth required, Fiber Optic will be the
ideal choice for future communications infrastructure. Fiber Optic system
is the most suitable data transmission medium for power system control,
protection and monitoring functions. The particular characteristics of the
optic fibers that contribute to its advantages are: low attenuation, high
bandwidth, electromagnetic interface immunity and security.
Widely used features in optic fibers are OADMs, Amplifier
technology, Transmitters, Amplifier Technology, Amplifiers and Optical
packet switches. The multi-layer topology forms the basis of current optic
fiber networks.
Without the need of an information management methodology and
wide area communication architecture, fiber optics will not effectively
contribute to power system communication.
To facilitate an effective substation communication network, the
automation system has the following needs of data acquisition, control
and monitoring, high availability and redundancy as well as capability for
future expansion.
For substation automation system design, various LAN topologies
such FDDI (Fiber Distributed Data Interface) or Giga Byte Ethernet, can be
used.
For the purpose of supporting the Fiber network infrastructure, a
four (4) layered internet architecture has been adopted instead of the seven
(7) layered model and has been both illustrated and discussed in the
network, layer, IP layer, TCP/UDP and Application layer.
A functional discussion, IP, TCP and UDP have been made.
In this chapter, the vital features of Chapter 3 on Protocols, Chapter
4 on Middleware, Chapter 5 on Information Embedded Power Systems
and Chapter 6 on Fiber Optic Network Infrastructure as Next Generation
Power Systems Communications has been discussed succinctly.
References
1. Udren, E., Kunsman, S. and Dolezilek, D.J., ‘‘Significant Substation

Communications Standardisation Developments,’’ SEL 2000.
2. Apostolov, A.P., ‘‘Configuration requirements for UCA-based
Intelligent Electronic Devices for Protection and Control,’’ CIGRE
2001 SC #4 Colliquium, Sibiu.
3. Dolezilek, D.J., ‘‘Choosing Between Communications Processors
RTUs and PLCs as Substation Automation Controllers,’’ SEL 2000.
4. Zimmerman, K. and Schweitzer, E.O., ‘‘Substation Communications:
When Should I Use EIA-232, EIA-485 and Optical Fiber,’’ Western
Protective Relay Conference, Spokane, Washington, 1996.
5. Behrendt, K.C. and Dood, M.J., “Substation Relay Data
Communications. Schweitzer Engineering Laboratories”, Pullman,
Washington, 2000. [Online] Available: http://www.selinc.com.
6. Dolezilek, D.J. and Klas, D.A., “Using Information From Relays to
Improve Protection”. Schweitzer Engineering Laboratories. Pullman,
Washington, 1999, pp. 1–7. [Online] Available. http://
www.selinc.com/techpprs/6080.pdf
7. Dolezilek, D.J., “Understanding, Predicting and Enhancing the
Power System through Equipment Monitoring and Analysis”.
Schweitzer Engineering Laboratories, Pullman, Washington, 1998,
pp.1–6. [Online] Available: http://www.selinc.com/techpprs/
6104.pdf.
8. Dolezilek, D.J., “Power System Automation”. Schweitzer
Engineering Laboratories. Pullman, Washington, 2000. [Online]
Available: http://www.selinc.com/techpprs/6091.pdf.
9. Apostolov, A. Gregory, P., “Network Interface Modules for
Microprocessor Relays Integration into a Substation Automation
System”. Proceedings of the Sixth International Conference on
Developments in Power System Protection, University of

Nottingham, UK, 1997.
10. Adamiak, M., Redfern, M., “Communication Systems for Protective
Relaying”. IEEE Computer Applications in Power, Vol. 3, No. 3,
pp. 14–22, July 1998.
11. Adamiak, M., Patterson, R., and Melcher, J., “Inter and Intra
Substation Communications: Requirements and Solutions”. GE
Protection and Control, Malvern, PA, Tech. Ref. APC–951, November
1998. [Online] Available: http://www.geindustrial.com/pm/notes/
apc–951.pdf.
12. Udren, E., Kunsman, S. and Dolezilek, S.J., “Significant Substation
Communications Standardization Developments”. Schweitzer
Engineering Laboratories, Pullman, Washington, 2000. [Online]
Available: http://www. Selinc.com/techpprs/6105.pdf.
13. Aldinger, G. and Koch, G., “Expanding the Scope of Protective Relays
Across Traditional Boundaries”. Proc. of the 2001 IEE Conf.,
pp. 323–327, 2001.
14. Ericssion, G.N., “Classification of Power Systems Communications
Needs and Requirements: Experiences from Case Studies at Swedish
National Grid.” IEEE Trans. On Power Delivery, Vol. 17 and No. 2,
pp. 345–46, April 2002.
15. Lai, R., Jirachiefpattana, A., “Communication Protocol Specification
and Verification”. Kluwer Academic Publishers, USA, 1998,
pp. 5–25.
16. Driscoll, F.F., “Data Communications”. International ed., Ed. Florida:
Harcourt Brace Jovanovich Publishers, 1992, pp. 233–35.
17. Holzmann, Gz., “Design and Validation of Computer Protocols”,
2nd ed., Ed., New Jersey: Prentice Hall, 1991, pp. 27–30.
18. Dolezilek, D.J., “Choosing Between Communications Processors,
RTUs and PLCs as Substation Automation Controllers”. Schweitzer
Engineering Laboratories, Pullman, Washington, October 2000.
[Online] Available: http://www.selinc.com/techpprs/6112.pdf.
19. CIGRE Study Committee 34. The Automation of New and Existing
Substations: Why and How. Report No. 3, CIGRE, Paris, France,
February 2002.
20. CIGRE Study Committee B5. The Automation of New and Existing
Substations: Why and How. Final Report, CIGRE, Paris, France,
November 2002.
21. Skeie, T., Johannessen, S. Brunner, C. “Ethernet in Substation
Automation. ABB Corporate Research,” Norway, 2000. [Online]
Available: http://www.simula.no/photo/ethernetinsa–col.pdf.
REFERENCES 345
22. Ozansoy, C., Zayegh, A., Kalam, A., “Modelling of a Substation

Communications System Using OPNET”. Proceedings of the 4th
International Modelling and Simulation Conference, Melbourne,
Australia, November 2002, pp. 255–260.
23. Proudfoot, D., UCA and 61850 for Dummies. Siemens Power
Transmission and Distribution 2002. [Online].
Available:http://www.nettedautomation.com/downolad/UCA%
20 and% 2061850% 20 for % 20 dummies % 20V12.pdf.
24. Adamiak, M., Baigent, D., UCA 2.0 The GE Experience. GE Power
Management, Markham, Ontario. [Online] Available:http://
www.geindustrial.com/pm/notes/ucaprot.pdf.
25. Janssen, M.C., Koreman, C.G.A., “Substation Components Plug and
Play Instead of Plug and Pray” The http:/www.
Nettedautomation.com/standardization/IEC –TC57/WG10 – 12/
index.html impact of IEC 61850. Kema T&D Power, Netherlands.
[Online] Avail Amsterdam.
26. Schwarz, K., “IEEE, UCA and IEC 61850 Applied in Digital
Substations,’’ Schwarz Consulting Company, SCC, Karlsruhe,
Germany. [Online]. Available://http:/www. Nettedautomation
.com/standardization/IEC–TC57/WG10–12/index.html.
27. Electric Power Research Institute, “Utility Communications
Architecture Version 2.0,’’ EPRI, California, USA, 1998.
28. Apostolov, A.P., “Configuration Requirements for UCA based IEDs
for Protection and Control”. CIGRE 2001 SC 34 Colloquium, Sibiu,
Sept. 2001.
29. Preiss, O., Wegmann, A., “Towards a Composition Model Problem
Based on IEC 61850.ABB Corporate Research Ltd.,”
Switzerland.[online]Available:http://www.sei.cmu.edu/pacc/
CBSE4–papers/Preiss Wegmann-CBSE4-4.pdf.
30. Steinhauser, F., Gupta, M., “Communication Technology in
Substations—Actual Developments from the View of Testing,”
Omicron Electronic GmbH, Austria. [Online] Available:http://
www.sei.cmu.edu/pacc/CBSE4–papers/Preiss Wegmann-CBSE4-
4.pdf.
31. IEC-TC57-WG10/11/12. IEC61850: “Communications Networks and
Systems in Substations.” International, Electrotechnical Commission,
Draft IEC61850, 1999.
32. IEC870-4. “Telecontrol Equipment and Systems—Part 4:
Performance Requirements.” International Electrotechnical
Commission, Geneva, International Standard, March 1990.
33. Kunsman, S.A., Kleman, M.C., “High Speed Communications for

Protective Relaying”. Proceedings of the 24th Annual Western Protective
Relaying Conference, Spokane, WA, 1997.
34. Apostolov, A., Gregory, P., “Network Interface Modules for
Microprocessor Relays Integration into a Substation Automation
System”. Proceedings of the Sixth International Conference on
Developments in Power System Protection, University of Nottingham,
UK, 1997.
35. Comparison of IEEE 1525 and IEC 61850, IEEE Standards Draft,
September, 1999.
36. Kalam, A., Spicer, A., Coulter, R., Klebanowski, A., Biasizzo, C. and
McDonald, H., “Power System Protection”. 6th ed.,Victoria
University, 2002, Chapter 21, pp. 1–22.
37. Electric Power Research Institute, “Communications on Power Lines:
Changing Business and Technology Ramifications”. EPRI, California,
USA, 2002.
38. Montoya, L.F., “Power Line Communications: Performance
Overview of the Physical Layer of Available Protocols”. University
of Florida, Gainesville, Florida, USA, 1998.
39. Kim, J. M., Song, Y., Kim, M. W., ‘‘Digital Neural Processor for Parallel
Structure Architecture and Cost Performance Comparison of the
Architectures’’, 1995. Proceedings IEEE International Conference, Vol. 3,
Nov. 27 – Dec. 1 1995, pp. 1474–77, Vol. 3.
40. Ozansoy, C., Zayegh, A., Kalam, A., ‘‘Communications for Substation
Automation and Integration,’’ presented at the 2002 AUPEC
Conference, Melbourne, Australia, 2002.
41. Sahingöz, Ö. K., Erdogan, N., ‘‘Agvent Agent Systems’’, Proc. of 4th
Int. ICSC Symp. On Engineering of Intelligent Systems (EIS 2004),
Madeira, Portugal, 2004, pp. 548-54. [Online] Available: http://
www.cs.itu.edu.tr/~erdogan/sahingoz-erdogan-EIS2004–Agv.pdf.
42. Bolton, F. 2001. Pure CORBA. Sams Publishing, Indianapolis, USA.
43. Object Management Group. ‘‘Real-Time CORBA Specification.’’
Version 1.1, OMG, August 2002.
44. Henning, M. and Vinoski, “S. 2000, Advanced CORBA Programming
with C++”. Addison Wesley, Boston, USA.
45. TC-57. “Communication Networks and Systems in Substations”—
Part 7-1: Basic Communication Structure for Substation and Feeder
Equipment—Principles and Models. Raft Standard 61850-7-1, IEC 2001.
46. Rajkumar, R., Gagliardi, M. and Sha, L., ‘‘The Real-time Publisher/
Subscriber Inter-process Communication Model for Distributed Real-time
REFERENCES 347
Systems: Design and Implementation,’’ in IEEE Real-time

Technology and Applications Symposium, 1995.
47. Kaiser, J., Mock, M., ‘‘Implementing the Real-time Publisher/
Subscriber Model on the Controller Area Network (CAN)’’,
Proceedings of the 2nd Int. Symp. On Object-oriented Real-time
distributed Computing (ISORC99), Saint-Malo, France, May 1999.
48. River Stone Networks, ‘‘Introduction to Multicast Routing’’. [Online]
Available: http://www.riverstonenet.com/support/multicast/
index.shtml.
49. CISCO, ‘‘Multicast Routing’’. [Online] Available: http://
www.cisco.com/warp/public/614/17.html.
50. Blumenfeld, S. M., ‘‘Multicast’’, Broadcast Engineering, Feb. 1, 2001.
[Online] Available: http://advertisers.broadcastengineering.com/
ar/broadcasting–multicast.
51. Araújo, F., Rodrigues, L., ‘‘Quality of Service in Indirect
Communication System’’. [Online] Available: http://
citeseer.ist.psu.edu/442144.html.
52. Araújo, F., Rodrigues, L., ‘‘On QoS—Aware Publish-Subscribe’’,
Proceedings of the International Workshop on Distributed Event-Based
Systems, Vienna, Austria, July, 2002.
53. CISCO, “Quality of Service Networking”, [Online] Available: http:
//www.cisco.com/univercd/cc/td/doc/cisintwk/ito_doc/
qos.htm
54. LOOSELY COUPLED, “QoS”. [Online] Available: http://
www.looselycoupled.com/glossary/QoS.
55. What is Quality of Service? [Online] Available: http://
www2.rad.com/networks/2001/qos/main.htm.
56. CISCO, “Quality of Service. End to End with CISCO IOS”, [Online]
Available: http://www.scd.ucar.edu/hps/CISCO/sld001.htm.
57. Ed. Braden, R., Zhang, L., Berson, S., Herzog, S., and Jamin, S.,
“Resource Reservation Protocol (RSVP)—Version 1, Functional
Specification”, September 1997. Request for Comments 2205.
58. Holstein, D.K., “Communication Architectures for Substation
Automation”, OPUS Publishing, 1999. [Online] Available: http://
www.opusss.com/Communication Architectures.htm.
59. Apostolov, A., Brunner, C., Clinard, K., “Use of IEC 61850 Object
Models for Power System Quality/Security Data Exchange.”
Proceeding of the CIGRE/IEEE PES International Symposium,
8–10 Oct. 2003, pp. 155–64.
60. Ozansoy, C., Zayegh, A., Kalam, A., “Modelling of a Substation
Communication System Using OPNET.” Proceeding of the 4th
International Modelling and Simulation Conference, Melbourne,

Australia, November 2002, pp. 255–60.
61. Amanullah Maung Than Oo, Kalam, A. and Zayegh, A.,
“Information Embedded Power System: The Effective
Communication System of the 21st Century Power System Industry.”
AUPEC 04, September 26–29, Brisbane, Australia.
62. Kwok-Hong Mak, Barry Holland, “Migrating Electrical Power
Network SCADA System to TCP/IP and Ethernet Networking”, IEE
Power Engineering Journal, pp. 305–11, December 2002.
63. Wood, A.J. and Wollenburg, B.F., “Power Generation, Operation and
Control”, John Wiley and Sons Inc., New York, NY, 1996.
64. IEEE Standard, “Definition, Specification and Analysis of System
Control”, Publication ANSI/IEEE C37.1.
65. Smit, H.L. and Block, W.R., “RTU’s Slave for Supervisory Systems”,
IEEE Computer Application in Power, Vol. 5, No. 1, January 1993,
pp. 27–32.
66. Goran, N., Ericsson, “Communication Requirements—Basis for
Investment in Utility Wide Area Netwrk”, IEEE Transactions on Power
Delivery, Vol. 19, No. 1, January 2004.
67. Allen Morgan, F., “Data on Demand”, IEEE Computer Application in
Power, January 1993, pp. 39–42.
68. Krumpholz, G.R., Clements, K.A. and Davis, P.W. “Power System
Observability: A Practical Algorithm Using Topology”, IEEE
Transactions on Power Systems, Vol. PAS-99, July 1980, pp. 1534–42.
69. Trevor, G., Alvey, Kwok, Cheung, W., “Deregulation Impacts and
Outcomes in Competitive Electicity Markets” Proceeding of the 5th
International Conference on Advances in Power System Control, Operation
and Management, APSCOM 2000, pp. 561–565, Hong Kong, October
2000.
70. Bin Qui, Hoay Beng Gooi, Yilu Liu; Eng Kiat Chan, “Internet-Based
SCADA Display System,” IEEE Computer Application in Power,
pp. 14–19, Jan. 2002.
71. Wang Zhaojia, Wang Yingtao, “PMU-Based Wide Area Monitoring
in Power System”, Power System and Communication Infrastructures
for the Future, Beijing, September 2002.
72. Luque, J., Escudero, J.I., and Perez, F., “Analytic Model of the
Measurement Errors Caused by Communications Delay”, IEEE
Transactions on Power Delivery, Vol. 17, No. 2, April 2002, pp. 334–37.
73. Gowar, John, “Optical Communication Systems”, Prentice Hall, 1995.
REFERENCES 349
74. Wook Hyun Kwon, Beon Jin Chung, “Real-Time Fiber Optic Network
for an Integrated Digital Protection and Control System”, IEEE
Transaction on Power Delivery, Vol. 7, No. 1. January 1992,
pp. 160–66.
75. Veeraraghavan, M., Karri, R., “Architecture and Protocols that enable
New Applications on Optical Network”, IEEE Communication
Magazine, Vol. 39, Issue 3, March 2001, pp. 118–27.
76. Tellez, J., Meriem, T.B., “Management Solutions for WDM
Networking”, IEEE International Conference on Network (ICON 2000)
Proceedings, 2000, pp. 120–24.
77. Bala, Rajagopalan, Pendarakis, D., “IP Over Optical Network:
Architecture Aspect”, IEEE Communication Magazine, Vol. 38, Issue
9, September 2000, pp. 94–102.
78. Harold Kirkham, Alan, R. Johnson, “Design Consideration for a Fiber
Optic Communications Networks for Power Systems”, IEEE
Transaction on Power Delivery, Vol. 9, No. 1, January 1994, pp. 510–18.
79. K. Washburn, J.T. Evans, “TCP/IP—Running Successful Network”,
Addison Wesley Publishing Company, 1998.
80. Amanullah Maung Than Oo, Kalam, A. and Zayegh, A., “Fiber Optic
Network Infrastructure.” As Next Generation Power System
Communications, JIEEE, November 2005, Jordan.
81. Qui, Bin, “Next Generation Information Communication
Infrastructure End Case Studies for Future Power Systems,” Ph.D.
Dessertation, Electrical and Computer Engineering, Virginia
Polytechnic Institute or State University, April 2002.
82. Hauer, J., Hughes, F.J., Trudenowski, D., et al., “A Dynamic
Information Manager for Networked Monitoring of Large Power
Systems”, EPRI Report WO 8813-01, October, 1998.
83. Liu, C.C., Heydt, G.T., Phadke, A.G., et al., “The Strategic Power
Infrastructure Defense (SPID) Systems”, IEEE Control System
Magazine, Vol. 20, Issue 4, August 2000, pp. 40–52.
84. Adamiak, M., Premerlani, W., “The Role of Utility Communications
in a Deregulated Environment”, Proceeding of the Hawaii’s International
Conference on System Sciences, Maui, Hawaii, January 1999, pp. 1–8.
85. “SCADA Communications”, Training Course, Institution of
Engineers, Australia, Western Australia Divison, 1996.
86. Newbury, John, Miller, William, “Potential Metering Communication
Services Using the Public Internet”, IEEE Transaction on Power
Delivery, Vol. 14, No. 2, October 1999, pp. 1202–07.
87. Mak, Siok, Radford, Denny, “Communication System Requirements

for Impementation of Large Scale Demand Side Management and
Distribution Automation”, IEEE Transaction Power Delivery, Vol. 11,
No. 2, April 1996, pp. 683–89.
88. Hauer, J., Trudnowski, D., “Keeping an Eye on Power System
Dynamics”, IEEE Computer Application in Power, Vol. 10, No. 4,
October 1997, pp. 50–54.
89. Heydt, G.T., Liu, C.C., Phadke, A.G., Vittal, V., “Solution for the Crisis
in Electric Power Supply”, IEEE Computer Applications in Power, Vol.
14, Issue 3, July 2001, pp. 22–30.
90. Adamiak, M., Redfern, Miles, “Communications Systems for
Protective Relaying”, IEEE Computer Applications in Power, Vol. 11,
July 1998, pp. 14–18.
91. Future Research Directions for Complex Interactive Electric
Networks, NSF/DOE/EPRI/DOD Sponsored Workshop, November
2000.
92. Adamiak, M., Premerlani, W., “Data Communications in a
Deregulated Environment”, IEEE Computer Application in Power,
Vol. 12, Issue 2, July 1999, pp. 33–39.
93. Chung, Sheng-Luen, Yang, Wen-Fa, “Data Acquisition and
Integration in Heterogeneous Computing Environment”, 1995
International IEEE/IAS Conference on Industrial Automation and Control:
Emerging Technologies, IEEE 1995, pp. 598–603.
94. Khatib, A.R., Dong, Z., Qiu, B., Liu, Y., “Thoughts on Future Internet-
based Power System Information Network Architecture”, IEEE,
Power Engineering Society Summer Meeting, Vol. 1, 2000, pp. 155–60.
95. Guilfoyle, D., Connolly, E., “Distributed SCADA Systems for
Electricity Distribution Control”, Power Technology International,
1994, pp. 169–172.
96. “System Adequacy and Security. Disturbance Monitoring,” NERC
Planning Standards. Approved by Board of Trustees, September
1997.
97. “Coordination of Control of Multiple FACTS/HVDC Links in the
Same System”, Electra. No. 187, December 1999, pp. 133–37.
98. Thomas, L. Baldwin, “Real Time Phasor Measurements for Improved
Monitoring and Control of Power System Stability”, Virginia Tech.
Ph.D. Dissertation, May 1993.
99. Adapa, R., Edris, A.A., “The Use of Real-time Phasor Measurement
in Power System Delivery”, Proceedings of EPRI Conference, 1996,
pp. 516–22.
REFERENCES 351
100. Mittelstadt, W.A., Krause, P.E., “The DOE Wide Area Measurement
System (WAMS) Project-Demonstration of Dynamic Information
Technology for the Future Power System”, EPRI Conference on the
Future of Power Delivery, April 1996.
101. Pao-Hsiang His, Shi-Lin Chen, “Distribution Automation
Communication Infrastructure”, IEEE Transactions on Power Delivery,
Vol. 13, No. 3, July 1998, pp. 728–34.
102. Qiu, B., Gooi, H.B., “Internet-based SCADA Display Systems (WSDS)
for Access via Internet.” IEEE Transaction on Power System, Vol. 15,
Issue 2, May 2000, pp. 681–86.
103. Fromm, W., Halinka, A., “Accurate Measurement of Wide-Range
Power System Frequency Changes for Generation Protection”, IEE
Conference Publication, No. 434, 1997, pp. 53–57.
104. Hauer, J.F., Cresap, R.L., “Measurement and Modelling of Pacific
AC Intertie Response to Random Load Switching”, IEEE Transactions
on Power Apparatus and Systems, Vol. PAS-100, No. 1, January 1981,
pp. 353–57.
105. Novosel, D., Vu, K.T., Eric Udren, Hart, “Practical Protection and
Control Strategies during Large Power System Disturbance”,
Transmission and Distribution Conference, 1996, pp. 560–65.
106. Martinez, J., Dortolina, C., “Dynamic Simulation Studies on Electric
Industrial System for Designing and Adjusting Load Shedding
Schemes”, Industrial and Commercial Power System Technical Conference,
Annual Meeting, 1994, pp. 23–29.
107. Chuvychin, V.N., Gurov, N.S., Venkata, S.S., Brown, R.E., “An
Adaptive Approach to Load Shedding and Spinning Reserve Control
During Underfrequency Conditions”, IEEE Transactions on Power
Systems, Vol. 11, Issue 4, November 1996, pp. 1805–10.
108. Moya, O., “Power System Computer Controlled Load Shedding”,
Electric Power System Research, April 1996, pp. 165–71.
109. Thang, W.Y., Boussion, J.Y., Peruzzo, B., Hubner, R., ”An Approach
for an Open Control System for Substations”, 14th International
Conference and Exhibition on Electricity Distribution, Part 1:
Contirbutions, IEE, Vol. 4, 1997, pp. 8/1–5.
110. Jay Murphy, R., “Power System Disturbance Monitoring”, IEEE
Proceeding of the International Symposium of Signal Processing and its
Applications, ISSPA, Vol. 1, Piscataway, NJ, 1996, pp. 282–85.
111. ACTLEM, “Power Quality Monitoring Systems for Utilities and
Industry”, http://www.actlem.com/
112. Advancements in Microprocessor-based Protection and

Communication, IEEE Tutorial Course 97-TP120-0.
113. Computer Aided Coordination of Line Protection Schemes, IEEE
PWR Report 90TH0285-7 PWR.
114. Carl, E., Grund, George Sweezy, “Dynamic System Monitoring for
HVDC Modulation Control”, IEEE Transactions on Power Delivery,
Vol. 8, No. 3, July 1993, pp. 853–58.
115. Report of a Panel Discussion, “Power System Disturbance
Monitoring Utility Experiences”, IEEE Transactions on Power Systems,
Vol. 3, No. 1, February 1988, pp. 134–48.
116. Menz, M.J., Payne, B., “Servers in SCADA Applications”, IEEE
Transactions on Industry Applications, September 1997,
pp. 1295–99.
117. Ma, T.K., Liu, T.M., Wu, L.F., “New Energy Management System
Architectural Design and Intranet/Internet Applications to Power
Systems”, 1998. International Conference on Energy Management and
Power Delivery, Vol. 1, 1998, pp. 207–12.
118. Abdel Rahman Amin Khatib, ”Internet-based Wide Area
Measurement Application in Deregulated Power Systems”, Virginia
Tech., Ph.D. Dissertation, 2002.
119. James, B. Bassich, Gerald W. Lester, “Open Systems and Freeware:
A Costeffective Approach”, Proceedings Conference XIX AM/FM
International. Thriving in an Age of Competition. AM/FM Int., 1996,
pp. 407–16.
120. “Photogrammetric Engineering Co. Ltd., GIS-An Overview”, http:/
/www.agis.co.il/overview.htm.
121. Smallworld, “Smallworld’s SRP Solutions”, http://
www.smallworldus.com/Products/utilities.
122. Blaha, M. and Premerlani, W., “Object-Oriented Modelling and
Design for Database Applications”, Prentice-Hall 1998.
123. Mark Wald, “Data Warehousing and the Internet”, IEEE PICA 99,
May 1999.
124. Marmiroli, H., Suzuki, H., “Web-based Framework for Electricity
Market, Electric Utility Deregrulation and Restructuring and Power
Technologies”, DRPT, 2000, pp. 471–75.
125. FERC Order No. 889, “Open Same-Time Information System and
Standards of Conduct”, Docket #RM95-9-000, April 1996.
126. Tian, Y., Gross, G., “OASISNET: An OASIS Network Simulator”,
IEEE Transactions on Power Systems, Vol. 13, No. 4, November 1998,
pp. 1251–58.
REFERENCES 353
127. Timothy, L., Skvarenina, “Use of a World Wide Web Page in An

Introductory Electrical Power and Controls Course”, Frontiers in
Education Conference, 1997, pp. 1051–55.
128. Yegappan, S.A. Chandrasekaran, “Development of ATP Tutorial
Using Java”, Proceedings of Thirtieth Southeastern Symposium on System
Theory, IEEE, 1998, pp. 20–24.
129. Data and Data Base Structures for Protection Engineering, IEEE PES
Report TP-129-0.
130. Fault and Disturbance Data Requirement for Automated Computer
Analysis, IEEE PES Report 95-TP 107.
131. John Gowar, Optical Communication Systems, Prentice Hall, 1995.
132. Wook Hyun Kwon, Beon Jin Chung, “Real-Time Fiber Optic Network
for an Integrated Digital Protection and Control System”, IEEE
Transaction on Power Delivery, Vol. 7, No. 1. January 1992,
pp. 160–66.
133. Veeraraghavan, M., Karri, R., “Architecture and Protocols that
Enables New Applications on Optical Network”, IEEE
Communication Magazine, Vol. 39, Issue 3, March 2001, pp. 118–27.
134. Awduche, A., Rekhter, Y., “Multiprotocol Lambda Switching:
Combining MPLS Traffic Engineering Control with Optical
Crossconnects”, IEEE Communication Magazine, Vol. 39, Issue 3,
March 2001, pp. 111–16.
135. Tellez, J., Meriem, J.T.B., “Management Solutions for WDM
Networking”, IEEE International Conference on Networks (ICON 2000)
Proceedings, 2000, pp. 120–24.
136. Rajagopalan, Bala, Pendarakis, D., “IP Over Optical Network:
Architecture Aspect”, IEEE Communication Magazine, Vol. 38, Issue
9, September 2000, pp. 94–102.
137. Ciancarini, P., Vitali, F., Mascolo, C., “Managing Complex Documents
Over the WWW: A Case Study for XML, Knowledge and Data
Engineering”, IEEE Transactions on Power System, Vol. 11, Issue 4,
July/August 1999, pp. 629–38.
138. Frances, M., Cleveland, G., “Information Exchange Modelling (IEM)
and Extensible Markup Language (XML) Technologies”, IEEE Power
Engineering Society Winter Meeting, Vol. 1, 2000, pp. 592–95.
139. Singh, J., “XML for Power Market Data Exchange”, IEEE Power
Engineering Society Winter Meeting, Vol. 2, 2001, pp. 755–56.
140. Utility Communication Architecture Manual, 1991.
141. www.lantronix.com/support/docs/pdf/cobox-e1m_t1m.pdf.
142. W3C Consortium, http://www.w3.org/XML.
143. Pokorny, J., “XML Functionally”, Database Engineering and

Applications Symposium, 2000, pp. 266–74.
144. Rollins, S., Sundaresan, N., “A Framework for Creating Customized
Multi-modal Interfaces for XML Documents”, IEEE International
Conference on Multimedia and Expo, Vol. 2, 2000, pp. 933–36.
145. Suresh, R., Shukla, P., Schwenke, G., “XML-based Data Systems for
Earth Science Applications”, IEEE 2000 International Symposium on
Geosciences and Remote Sensing (IGRASS) Proceedings, Vol. 3, 2000,
pp. 1214–16.
146. Buchner, A.G., Baumgarten, M., “Data Mining and XML: Current
and Future Issues”, Web Information Systems Engineering, Vol. 2,
2000, pp. 131–35.
147. Harold Kirkham, Alan, Johnson, R., “Design Consideration for a
Fiber Optic Communications Network for Power Systems”, IEEE
Transaction on Power Delivery, Vol. 9, No. 1, January 1994,
pp. 510–18.
148. Sabir, S., Mahoney, H., “Building a Backbone for Integrated Business
Communications”, IEEE Computer Applications in Power, Vol. 9,
Issue 1, February 1996, pp. 38–41.
149. Washburn, K., Evans, J.T., “TCP/IP – Running Successful Network”,
Addison-Wesley Publishing Company, 1998.
150. Larry, L. Peterson and Bruce, S.D., Computer Network: A System
Approach, Morgan Kaufmann, 2000.
151. IEEE Communication Protocol Tutorial, February 1995.
152. Mohammad Shahidepur, Yaoyu Wang, “Communication and
Control in Electric Power Systems” IEEE Press, Wiley Interscience,
A. John Wiley & Sons, Inc., Publication, 2003.
153. Kothari, D.P., and Nagrath, I.J., “Modern Power System Analysis”,
McGraw-Hill, New York, 2006.
154. Kothari, D.P. and Nagrath, I.J., “Power System Engineering”, 2nd
edn. Tata McGraw-Hill, New Delhi, 2007.
155. Triangle Micro Works Inc, ‘‘DNP 3 Overview’’, 2002, [Online]
Available: http://www.TriangleMicroWorks.com.
156. Lebakken, T.M. and Orlando, D.R., ‘‘Substation Automation and
Communication Standard: IEC 61850 and DNP3’’, 2009.
Index
Aliasing 203 Confidentiality 293

Amplitude and phase comparison Context-based 295
173 Conventional fusing current 162
Amplitude comparator 176 Conventional non-fusing current
Application layer 331 162
Average comparator 192 CORBA 340
Core saturation 94
‘B’ type fuses 137 Current transformers 168
Custom queuing 301
Back to back bridge 186
Back up relaying 14, 15
Balanced voltage scheme 216 ‘D’ type fuselink 136
Biased differential 212 Data link layer 331
Binding 298 Data push architectures 292
Blind spot 224, 225 Data transport control 313
Block-average comparator 190 Datagrams 334
Breaking capacity 163 Dedicated links 325
Broadcast messaging 297, 340 Derived burden 95
Dial-out 249
Capacitor divider 84
Differential protection 209
Channel-based 295
Digital relay 198
Circulating current 216
Directional relay 178
Client-server architectures 289, 339
Discrimination 20
Committed access rate mechanism
Distance relay 169
301
Distance protection 167
Common multiple earthed
Distributed network protocol 265
neutral 65
Double switching 10
Communication protocol stack 311
Duplicate primary relaying 16 Liquid fuse 141

Duplicated protection 75 Local back up 15, 225, 226
Low pass filter 195
Earth fault 183 Low speed schemes 92
Earth fault current limitation 65 Low voltage cartridge fuse 128
Earth fault protection 122
Earth potential rise code 60 Magnetising impedance 211
Eavesdrop communications 250 Major fuse 157
Erbium doped fiber amplifier 327 Management application interface
Expulsion fuses 142 314
Mann-Morrison algorithm 204
Fault current paths 73 Master-slave communication 286
Fault resistance 200 Maximum prospective current 164
FDDI 342 Memory circuit 194
Ferro-resonance 83 Metcalf 130
Filtering 298 mho 171
Fixed time reference comparator 189 Microwave 325
Footing resistance 200 Middleware 288
Fulgurite 146 Miniature type 128
Fully cross-polarised 193 Minimum fusing current 127, 151
Fuseholder 126 Minor fuse 157
Fuselinks 134 Multicast messaging 296, 340
Fusion factor 127
Negative sequence impedances 31
GOOSE 339 Non Unit schemes 18, 219
Grading by current 113
Grading margin 121 OADMs 330, 342
Normal operation 3
High fidelity CVT 83 Overcurrent protection 167
Heat sink effect 130 Overreach relay 179
High speed schemes 92
High voltage cartridge fuse 128 Peak arc voltage 146
Peer-to-Peer direct connection 254
Impedance measurement 172 Phase comparator 176
Impedance relay 177 Phase faults 184
Instrument transformers 167 Physical layer 330
Inter-control centre 318 Pilot wire 324
Inverse definite minimum time 102 Plug fuse 140
Inverse overcurrent relay 211 Plug setting multiplier 103
IP routing 341 Point-to-Point architectures 289, 339
Positive sequence impedances 31
Knee-point e.m.f. 87 Power line carrier 325
INDEX 357
Power transformer 213 Station earthing 58

Powerline communication 335 Subject-based 295
Pre-arcing time 150, 155 Subtransient impedance 6
Prevention of electrical failure 3 Summation transformer 216
Primary relaying 14 Switched distance 170
Priority queuing 301 Symmetrical components 31
Process bus 287 Synchronisation 291
Prospective voltage 214 Synchronous impedance 6
Protocol gateway 248 System earthing 54
Publish-subscribe architectures
289, 339 Time grading 197
Publish/Subscribe middleware 294 Time multiplier setting 103
Pulse type phase comparator 188 Time synchronization broadcast 250
Tower-footing resistance 4
Radio systems 325 Transactor 192
Reactance relay 178 Transducers 84
Remote back up 224 Transient impedance 6
Remote monitoring 339
Remote terminal unit 261 Underreach relay 179
Resistance 179 Unicast messaging 296, 340
Resource reservation protocol 301 Unit protection 19
Restricted earth fault 221 Unit schemes 219
Router 249
RTUs 340 Voltage transformers 168
Voltage-controlled voltage source
Safety earthing 55 (VCVS) 195
Safety margin 121
Sallen and Kelly 195 WAN 340
Self-polarised 193 Warrington’s formula 200
Semi-enclosed fuse 133 Weighted fair queuing 301
Semiconductor optical amplifier 327 Wide area communication 329
Servers 249
Single switching 9
Zero sequence impedances 31

Kalam, Akhtar - Kothari, Dwarkadas Pralhaddas - Power System Protection and Communications-New Age Science LTD (2010)

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

Kalam, Akhtar - Kothari, Dwarkadas Pralhaddas - Power System Protection and Communications-New Age Science LTD (2010)

Uploaded by

Copyright:

Available Formats

Power System

Electrical protection is as fundamental to the generation, transmission and

started adding other features like fault locators, measurements and

RTU — Remote Terminal Units

GOOSE — Generic Object Oriented Substation Event

XML — Extensible Markup Language

Chapter 2 Network Analysis and Fault Calculations 22–50

Chapter 3 Earth Fault and Interferences 51–79

Chapter 4 Relaying Transducers 80–100

Chapter 5 Overcurrent Protection 101–125

Chapter 6 Fuses 126–166

Chapter 7 Distance/Impedance Protection 167–208

7.4 Relay Types and Their Application 176

Chapter 8 Differential Protection 209–218

Chapter 9 Unit, Remote and Back Up Protection 219–239

Chapter 1 Introduction 243–245

Chapter 2 Communication Principle 246–259

Chapter 3 Protocols 260–278

ESAA COMMITTEE NO. 2.14 – PROTECTION

Term Definition Remarks

Backup Protection A protection which is intended See also Local Backup

Balanced Fault A fault giving rise to steady

Balanced Voltage A differential protective system See also Longitudinal

Biased Relay An electrical measuring relay Similar to IEC 50(16).

Blinder A relay whose characteristic, Usually applied to long,

Bolted Fault A balanced fault of virtually Typically a fault arising as a

Breaker Failure A specific form of local backup Frequently by means of a

Also, a relay which by its A current check relay in a local

Check Zone Term applied to the non-

Circulating A differential protective system The classical Merz-Price

Comparator A device which compares two In most distance relays, either

Component Test A test to determine the BS 3950.

Conjunctive Test A test on a protective system, BS 3950.

Cross-country Simultaneous flashover to Common on unearthed

DC/DC Converter A power supply employing

Dead Zone See ‘Blind Spot’. Not recommended

Dependent Time A measuring relay having an

Differential A unit protective system in Similar to BS 3950 definition.

Digital Relay A relay which processes the

Direct Intertrip The signal initiated by local

Directional A unit protection system

Directional A non-unit protection, the

Discrimination The quality whereby a BS 3950.

Discriminative Term applied to the selective

Distance A non-unit protection system, IEC 50 Section 448-03.

Distance A Distance Protection System

Distance A Distance Protection System

Electro-mechanical An electrical relay or relay

Fault Clearance The time interval between

Fault Detector An element of a protection Alternative term for Starting

Fault Level The rms value of the fault

Fault Setting The limiting value of the May be referred either to

Fault Throwing A method in which operation IEC 50(16).

Field Failure A protective system which

Final Tripping A tripping command issued by

Flag Indicator An indicating device which Preferred to the term ‘target’.

Flip-Flop A bistable logic device with

Frame Leakage Protection in which the Similar to IEC 50(16).

High Frequency A type test intended for AS 2481-1981.

High-Impedance A single-input relay having a Frequently used in differential

High-speed A protective system whereby Preferred to the term

Impedance Relay A specific type of distance

Impulse Starter A starting element in a Often used to discriminate